From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 11:13:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 13 11:13:32 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 11:13:32 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0001.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 13 15:14:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:14:02 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0002.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 15:21:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0003.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 16:00:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 13 20:01:58 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0004.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 20:01:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 20:01:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 13 20:01:59 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0005.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 08:02:33 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0006.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 12:00:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0007.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 16:00:53 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0008.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 14 20:00:54 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0009.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 08:02:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0010.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 12:01:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0011.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 16:00:53 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0012.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 15 20:00:52 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0013.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 08:02:45 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0014.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 12:00:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0015.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 16:00:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0016.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 16 20:00:54 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0017.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 08:02:48 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0018.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 12:01:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0019.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 16:01:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0020.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 17 20:01:00 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0021.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 08:04:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 18 08:04:35 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 08:04:35 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0022.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 12:01:16 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 18 16:01:48 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0023.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 16:01:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 16:01:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 18 16:01:48 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 16:01:48 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 16:01:49 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0024.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 18 20:01:07 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 19 08:04:55 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0025.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 08:04:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0026.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 12:01:12 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 19 12:01:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 12:01:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0027.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 16:03:42 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 19 16:03:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 16:03:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0028.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 19 20:00:58 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 20 08:07:55 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0029.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 08:07:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0030.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 12:02:30 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 20 16:01:02 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0031.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 16:01:02 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 16:01:02 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 16:01:03 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0032.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 20 20:00:59 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0033.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 08:04:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0034.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 12:01:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0035.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 16:01:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0036.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 21 20:00:59 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0037.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 08:03:04 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0038.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 12:01:02 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0039.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 16:01:39 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0040.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 22 20:01:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0041.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 08:03:01 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0042.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 12:01:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0043.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 16:02:14 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 23 16:02:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 16:02:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 23 16:02:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 16:02:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0044.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 20:01:03 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 23 20:01:04 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 20:01:04 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 23 20:01:04 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 23 20:01:04 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0045.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 08:03:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 24 12:03:44 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0046.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 12:03:44 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 12:03:44 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 24 12:03:44 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 12:03:44 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 12:03:45 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0047.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 16:01:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0048.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 24 20:02:10 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0049.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 08:02:25 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0050.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 12:01:20 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0051.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 16:01:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0052.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Jan 25 20:01:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 26 08:06:50 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0053.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 08:06:50 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 08:06:50 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 26 08:06:50 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 08:06:51 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0054.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 12:01:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0055.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 16:01:05 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0056.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Jan 26 20:00:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0057.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 08:07:46 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 27 12:02:26 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0058.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 12:02:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0059.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 16:01:20 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0060.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Jan 27 20:01:10 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 20:01:12 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Jan 27 20:01:12 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 20:01:12 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Jan 27 20:01:12 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Jan 27 20:01:12 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 28 08:06:26 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0061.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 08:06:26 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 08:06:26 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 28 08:06:26 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 08:06:26 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 08:06:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0062.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 12:02:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 28 12:02:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 12:02:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0063.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 16:01:07 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0064.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Jan 28 20:01:06 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 29 08:07:55 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0065.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 08:07:55 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 08:07:55 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 29 08:07:55 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 08:07:55 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 08:07:56 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0066.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 12:01:17 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0067.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 16:01:06 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0068.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Jan 29 20:01:08 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0069.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 08:10:41 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 30 08:10:42 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 08:10:42 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 30 08:10:42 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 08:10:42 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0070.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 12:04:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0071.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 16:04:42 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 30 16:04:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 16:04:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0072.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Jan 30 20:02:36 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 31 08:08:38 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0073.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 08:08:39 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0074.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 31 12:04:50 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 12:04:51 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 31 12:04:51 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 12:04:51 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 31 12:04:51 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 12:04:51 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0075.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 16:02:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0076.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Jan 31 20:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 1 08:10:57 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0077.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 08:10:58 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0078.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 12:03:09 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0079.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 16:01:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0080.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 1 20:01:16 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 2 08:14:54 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0081.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 08:14:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 08:14:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 2 08:14:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 08:14:56 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 08:14:56 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 2 08:14:56 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 08:14:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 2 08:14:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 08:14:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 2 08:14:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 08:14:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0082.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 12:03:17 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0083.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 16:01:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0084.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 2 20:01:18 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 3 08:14:06 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0085.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 08:14:07 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0086.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 12:03:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0087.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 16:01:18 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0088.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 3 20:01:23 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 4 08:09:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0089.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 08:09:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 08:09:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 4 08:09:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 08:09:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 08:09:35 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0090.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 4 12:03:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 12:03:14 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0091.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 16:01:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 4 20:01:20 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0092.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 20:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 20:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 4 20:01:20 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 4 20:01:21 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0093.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 08:08:24 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0094.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 12:03:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0095.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 5 16:01:24 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 16:01:25 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0096.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 5 20:01:17 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 6 08:17:11 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0097.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 08:17:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0098.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 12:03:18 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0099.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 16:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 6 20:01:19 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0100.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 20:01:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 20:01:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 6 20:01:19 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 6 20:01:20 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 7 08:27:30 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0101.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 08:27:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 7 08:27:32 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 08:27:32 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0102.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 12:03:51 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0103.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 16:01:29 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0104.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 7 20:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 8 08:26:35 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0105.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 08:26:36 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0106.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 12:04:04 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0107.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 16:01:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 8 20:01:21 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0108.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 20:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 20:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 8 20:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 20:01:21 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 8 20:01:21 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 8 20:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 20:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 8 20:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 20:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 8 20:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 8 20:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0109.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 08:10:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 9 08:10:28 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 08:10:28 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 9 08:10:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 08:10:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 9 12:03:47 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0110.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 12:03:47 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 12:03:48 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0111.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 16:01:34 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 9 20:01:30 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0112.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 20:01:30 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 20:01:30 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 9 20:01:30 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 9 20:01:31 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 10 08:26:10 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0113.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 08:26:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0114.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 12:04:15 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0115.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 10 16:01:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 16:01:28 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0116.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Feb 10 20:01:25 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 11 08:14:08 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0117.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 08:14:09 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 08:14:09 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 11 08:14:10 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 08:14:10 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 08:14:10 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 11 08:14:10 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 08:14:10 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 11 08:14:10 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 08:14:11 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 11 08:14:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 08:14:11 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 11 12:03:12 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0118.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 12:03:12 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 12:03:12 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 11 12:03:12 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 12:03:12 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 12:03:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0119.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 16:01:22 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sat Feb 11 20:01:13 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0120.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 20:01:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 20:01:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sat Feb 11 20:01:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 20:01:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sat Feb 11 20:01:14 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0121.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 08:10:46 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0122.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 12:03:19 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0123.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 16:01:27 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0124.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Sun Feb 12 20:01:19 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 13 08:19:35 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0125.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 08:19:36 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 13 12:03:40 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0126.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 12:03:43 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0127.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 16:01:45 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0128.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Mon Feb 13 20:01:24 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 14 08:21:48 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0129.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 08:21:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 08:21:48 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 08:21:49 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0130.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 12:03:57 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 14 16:01:59 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0131.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 16:02:00 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Tue Feb 14 20:01:29 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0132.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 20:01:29 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 20:01:29 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Tue Feb 14 20:01:30 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 15 08:33:27 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0133.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 08:33:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 08:33:28 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 15 08:33:28 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 08:33:28 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 08:33:29 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 15 08:33:29 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 08:33:30 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 15 08:33:30 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 08:33:30 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 15 08:33:31 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 08:33:33 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0134.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 12:06:13 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0135.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 16:01:52 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 15 19:28:31 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0136.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 19:28:32 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0137.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Wed Feb 15 20:15:31 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Thu Feb 16 08:50:34 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0138.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 16 08:50:35 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 16 08:50:35 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Thu Feb 16 08:50:35 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 16 08:50:36 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Thu Feb 16 08:50:37 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Thu Feb 16 08:50:37 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 16 08:50:37 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Thu Feb 16 08:50:37 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 16 08:50:37 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Thu Feb 16 08:50:38 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Thu Feb 16 08:50:38 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From jeff.oneill at gmail.com Wed Sep 7 10:53:53 2005 From: jeff.oneill at gmail.com (Jeffrey O'Neill) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] bugs in mx.Number.Rational Message-ID: <80211807050907065375312954@mail.gmail.com> Hello, I recently installed your mx.Number package in order to use the Rational number type in a software package I am working on: http://stv.sf.net. I do fairly intensive computations and I would like to convert my floating point calculations to rational number calculations. Overall it seems to be working well, but I'm wondering if the following are two bugs: (1) from math import * a = Rational(pi) * Rational(e) b = Rational(pi) / Rational(e) a > b returns True as expected a < b returns False as expected a == b returns RuntimeWarning tp_compare didn't return -1, 0 or 1 (2) a = Rational(1, 2) a < 0.1 incorrectly returns True Any thoughts would be greatly appreciated. best regards, Jeff -------------- next part -------------- An HTML attachment was scrubbed... URL: /mailman-archives/egenix-users/attachments/20050907/b568544f/attachment-0139.htm From mick at portofhamptonroads.com Mon Sep 12 16:14:30 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Hello! I don't suppose this hotfix would work for 1.0.8 as well? I'm having the exact same problem after upgrading to Zope 2.8.1 (Python 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 release? I would have downloaded the 1.0.9 pre-release but did not see a binary package of it for the Windows platform. I'm running two instances with mxODBCZopeDA commercial licenses that I can't finish migrating until I get a fix for this. Any help would be appreciated. Thanks! M >Daniel W. Adair wrote: >> Thanks so much! >> >> Hotfix worked, everything fine again. >> I don't know what else to say, this is just fantastic response time and >> quality! > > Thanks :-) > > We've put the hotfix up online until we have the 1.0.9 final > release ready: > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease -py2.3-Hotfix-2005-08-16.zip > > This is for Python 2.3 and the 1.0.9 pre-release only. Simply > unzip the above file in the Zope directory. It provides an > updated ZopeDA.pyc file which includes the fix. From mal at egenix.com Tue Sep 13 11:39:29 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 In-Reply-To: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> References: <97ECE86F93743340AA667B28DD13EBA22156C3@mail.intranet.hrma.portofhamptonroads.com> Message-ID: <43269041.1070707@egenix.com> Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not see > a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final >>release ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelease > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply >>unzip the above file in the Zope directory. It provides an >>updated ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From rune.froysa at usit.uio.no Tue Sep 13 12:32:37 2005 From: rune.froysa at usit.uio.no (Rune Froysa) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py Message-ID: I needed a lockfile implementation to prevent two instances of a script from running at the same time, and the FileLock module seems to fit my needs perfectly. I'm a bit surprised that this bug (TypeError: not all arguments converted) hasn't been encountered earlier. diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 @@ -156,7 +156,7 @@ continue raise Error,\ 'file "%s" is locked by process %s:%i' % \ - (self.filename,host,locking_pid,hostname) + (self.filename,host,locking_pid) self.locked = 1 def unlock(self,sleeptime=0.0001, /Rune From mal at egenix.com Tue Sep 13 13:03:20 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Bug in {BeeBase,Misc}/FileLock.py In-Reply-To: References: Message-ID: <4326A3E8.7070402@egenix.com> Rune Froysa wrote: > I needed a lockfile implementation to prevent two instances of a > script from running at the same time, and the FileLock module seems to > fit my needs perfectly. I'm a bit surprised that this bug > (TypeError: not all arguments converted) hasn't been encountered > earlier. Thanks. Indeed, it looks as if people are not really using mxBeeBase a lot :-) We've fixed it some time ago in the (non-final) 2.1 version. Here's the current snapshot: http://www.egenix.com/files/python/egenix-mx-base-2.1.0-2005-05-01.zip > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/BeeBase/FileLock.py 2001-06-24 13:54:25.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/BeeBase/FileLock.py 2005-09-13 11:23:39.229964215 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > Only in egenix-mx-base-2.0.6/mx/BeeBase: FileLock.py~ > diff -u --recursive egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py egenix-mx-base-2.0.6/mx/Misc/FileLock.py > --- egenix-mx-base-2.0.6.orig/mx/Misc/FileLock.py 2001-06-24 13:54:29.000000000 +0200 > +++ egenix-mx-base-2.0.6/mx/Misc/FileLock.py 2005-09-13 11:23:47.715743293 +0200 > @@ -156,7 +156,7 @@ > continue > raise Error,\ > 'file "%s" is locked by process %s:%i' % \ > - (self.filename,host,locking_pid,hostname) > + (self.filename,host,locking_pid) > self.locked = 1 > > def unlock(self,sleeptime=0.0001, > > /Rune > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From mick at portofhamptonroads.com Tue Sep 13 10:15:04 2005 From: mick at portofhamptonroads.com (Mick Drevyanko) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Message-ID: <97ECE86F93743340AA667B28DD13EBA22156CA@mail.intranet.hrma.portofhamptonroads.com> I tested the hotfix for 1.0.9 in my environment and it seems to be working well so far. Thanks! -----Original Message----- From: M.-A. Lemburg [mailto:mal@egenix.com] Sent: Tuesday, September 13, 2005 4:39 AM To: Mick Drevyanko Cc: egenix-users@egenix.com Subject: Re: [egenix-users] Re: Help! - mxODBCZopeDA 1.0.9 breaks my Zope 2.8.1 Mick Drevyanko wrote: > Hello! > > I don't suppose this hotfix would work for 1.0.8 as well? The hotfix should also work for the 1.0.8 version, provided you use Python 2.3 to run Zope. The ZIP file only contains the ZopeDA.pyc file which should be portable across platforms. However, we haven't tested it with the mxODBC interface included in mxODBC Zope DA 1.0.8, so there may be problems. In any case, we're going to release 1.0.9 final next week which will include the hotfix for Zope 2.8 on all platforms. > I'm having the exact same problem after upgrading to Zope 2.8.1 > (Python > 2.3.5) in a Windows environment. Is a hotfix available for the 1.0.8 > release? I would have downloaded the 1.0.9 pre-release but did not > see a binary package of it for the Windows platform. > > I'm running two instances with mxODBCZopeDA commercial licenses that I > can't finish migrating until I get a fix for this. Any help would be > appreciated. > > Thanks! > > M > > >>Daniel W. Adair wrote: >> >>>Thanks so much! >>> >>>Hotfix worked, everything fine again. >>>I don't know what else to say, this is just fantastic response time > > and > >>>quality! >> >>Thanks :-) >> >>We've put the hotfix up online until we have the 1.0.9 final release >>ready: >> >> > > http://www.egenix.com/files/python/egenix-mxodbc-zopeda-1.0.9-prerelea > se > -py2.3-Hotfix-2005-08-16.zip > >>This is for Python 2.3 and the 1.0.9 pre-release only. Simply unzip >>the above file in the Zope directory. It provides an updated >>ZopeDA.pyc file which includes the fix. > > > > _______________________________________________________________________ > eGenix.com User Mailing List http://www.egenix.com/ > https://www.egenix.com/mailman/listinfo/egenix-users -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 13 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From arso at libero.it Mon Sep 19 23:50:54 2005 From: arso at libero.it (Arso - Gentoo) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module Message-ID: <432F24AE.7000400@libero.it> This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') occurs the error File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, in ParseDateTimeGMT return DateTime.DateTime(year,month,day,hour,minute,second) - offset TypeError: an integer is required this becouse after the split in date items at line 133 litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() the missing century is handled by calling DateTime.add_century(year) while year is string, so the add_century does not work. I suggest tu use instead on line 135 DateTime.add_century(atoi(year)) The same on line 67 and 98. -- ====================================================== Arso - Italy Gentoo 2004-02 Stage 1 Armada e500 ====================================================== From mal at egenix.com Tue Sep 20 12:00:52 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] Incorrect handle of 2 digit year in ARPA module In-Reply-To: <432F24AE.7000400@libero.it> References: <432F24AE.7000400@libero.it> Message-ID: <432FCFC4.5010609@egenix.com> Arso - Gentoo wrote: > This happen on Li nux Gentoo using package egenix-mx-base-2.0.5. > > Parsing a date in format 'Wed, 14 Sep 05 11:28:48 +0100' with > DateTime.ARPA.ParseDateTimeUTC('Wed, 14 Sep 05 11:28:48 +0100') > occurs the error > > File "/usr/lib/python2.3/site-packages/mx/DateTime/ARPA.py", line 155, > in ParseDateTimeGMT > return DateTime.DateTime(year,month,day,hour,minute,second) - offset > TypeError: an integer is required > > this becouse after the split in date items at line 133 > litday,day,litmonth,month,year,hour,minute,second,zone = date.groups() > the missing century is handled by calling > DateTime.add_century(year) > while year is string, so the add_century does not work. > > I suggest tu use instead on line 135 > DateTime.add_century(atoi(year)) > > The same on line 67 and 98. Thanks. While I'll add support for 2-digit years, I'd like to encourage everyone to use the 4-digit year representation instead. This is also the format defined in the documentation and the one used throughout the Internet nowadays: http://rfc.sunsite.dk/rfc/rfc2822.html -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 20 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From cariaso at yahoo.com Fri Sep 23 17:03:31 2005 From: cariaso at yahoo.com (Mike Cariaso) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable Message-ID: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Marc-Andre, I've seen your reply to this message. http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html For a long time I've done as you've advised, and ignored the issue. However the most recent version of py2exe provides the ability to bundle everything into a single executable. For me this is a very compelling new feature. However the py2exe code now treats this issue as fatal. While I don't use the egenix libraries directly, I'm dependent on too many projects which are. What updated advice can you offer? Mike Cariaso From mal at egenix.com Sat Sep 24 12:33:22 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] ['mxDateTime.__version__'] no longer ignorable In-Reply-To: <20050923230331.33155.qmail@web52713.mail.yahoo.com> References: <20050923230331.33155.qmail@web52713.mail.yahoo.com> Message-ID: <43351D62.5000809@egenix.com> Mike Cariaso wrote: > Marc-Andre, > > I've seen your reply to this message. > http://www.egenix.com/mailman-archives/egenix-users/2004-November/000549.html > > For a long time I've done as you've advised, and > ignored the issue. However the most recent version of > py2exe provides the ability to bundle everything into > a single executable. > > For me this is a very compelling new feature. However > the py2exe code now treats this issue as fatal. > > While I don't use the egenix libraries directly, I'm > dependent on too many projects which are. > > What updated advice can you offer? I have no idea why py2exe thinks that mxDateTime.__version__ is a module. It actually is a symbol that is being imported from the C extension. I'd suggest you ask Thomas Heller about the problem and why py2exe thinks it needs a module with that name. Thanks, -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 24 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! :::: From misa at redhat.com Fri Sep 30 21:20:47 2005 From: misa at redhat.com (Mihai Ibanescu) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning Message-ID: <20050930202159.GC15179@abulafia.devel.redhat.com> Hello, Cleaning up my bug list... https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 A more juicy one: https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 IIRC, on 64-bit systems there are lots of warnings about unsafe pointer manipulation. Are these known issues? Thanks! Mihai From mal at egenix.com Fri Sep 30 22:20:32 2005 From: mal at egenix.com (M.-A. Lemburg) Date: Fri Mar 31 16:34:16 2006 Subject: [egenix-users] comparisons like X<=Y<=Z do not have their mathematical meaning In-Reply-To: <20050930202159.GC15179@abulafia.devel.redhat.com> References: <20050930202159.GC15179@abulafia.devel.redhat.com> Message-ID: <433DABE5.8070502@egenix.com> Mihai Ibanescu wrote: > Hello, > > Cleaning up my bug list... > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=162375 ... if (0 < number_len < sizeof(buffer)) { ... Has never caused problems in the past, but we'll fix it anyway :-) > A more juicy one: > > https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=116880 That bug report has next to no usable content. > IIRC, on 64-bit systems there are lots of warnings about unsafe pointer > manipulation. > > Are these known issues? egenix-mx-base 2.0.x does have some issues on x86_64 which is the only 64-bit platform we currently support. Most of these should be fixed in the 2.1.x version branch which we'll be releasing sometime in the next few months (we still have some work to do on the documentation). -- Marc-Andre Lemburg eGenix.com Professional Python Services directly from the Source (#1, Sep 30 2005) >>> Python/Zope Consulting and Support ... http://www.egenix.com/ >>> mxODBC.Zope.Database.Adapter ... http://zope.egenix.com/ >>> mxODBC, mxDateTime, mxTextTools ... http://python.egenix.com/ ________________________________________________________________________ ::: Try mxODBC.Zope.DA for Windows,Linux,Solaris,FreeBSD for free ! ::::