From chris at simplistix.co.uk Fri Dec 2 19:30:07 2011 From: chris at simplistix.co.uk (Chris Withers) Date: Fri, 02 Dec 2011 18:30:07 +0000 Subject: [Catalog-sig] Attempts to hack pypi? In-Reply-To: <20111202062506.A5EE84F8107@server1.simplistix.co.uk> References: <20111202062506.A5EE84F8107@server1.simplistix.co.uk> Message-ID: <4ED9192F.2000403@simplistix.co.uk> Hi All, I got these from a logwatch warning on my server: /pypi?:action=/../../../../../../../../../../etc/passwd%00&digest=517d7c7014d2b581f321d500eed4305e HTTP Response 302 /pypi?%3Aaction=../../../../../../../../../../etc/passwd HTTP Response 302 /pypi?%3Aaction=../../../../../../../../../../proc/self/environ HTTP Response 302 /pypi?%3Aaction=../../../../../../../../../../etc/passwd%00 HTTP Response 302 /pypi?:action=../../../../../../../../../../proc/self/environ&provider=myOpenID HTTP Response 302 /pypi?:action=show_md5&digest=/../../../../../../../../../../proc/self/environ%00 HTTP Response 302 /pypi?:action=show_md5&digest=../../../../../../../../../../proc/self/environ HTTP Response 302 /pypi?:action=../../../../../../../../../../etc/passwd%00&digest=517d7c7014d2b581f321d500eed4305e HTTP Response 302 /pypi?%3Aaction=/../../../../../../../../../../etc/passwd%00 HTTP Response 302 /pypi?:action=../../../../../../../../../../proc/self/environ&digest=517d7c7014d2b581f321d500eed4305e HTTP Response 302 /pypi?:action=../../../../../../../../../../etc/passwd&provider=myOpenID HTTP Response 302 /pypi?:action=/../../../../../../../../../../etc/passwd&provider=myOpenID HTTP Response 302 /pypi?%3Aaction=/../../../../../../../../../../proc/self/environ%00 HTTP Response 302 /pypi?%3Aaction=/../../../../../../../../../../etc/passwd HTTP Response 302 /pypi?:action=../../../../../../../../../../etc/passwd&digest=517d7c7014d2b581f321d500eed4305e HTTP Response 302 /pypi?:action=/../../../../../../../../../../etc/passwd&digest=517d7c7014d2b581f321d500eed4305e HTTP Response 302 /pypi?:action=show_md5&digest=/etc/passwd HTTP Response 302 /pypi?:action=/../../../../../../../../../../proc/self/environ%00&provider=myOpenID HTTP Response 302 /pypi?:action=../../../../../../../../../../etc/passwd%00&provider=myOpenID HTTP Response 302 /pypi?%3Aaction=/etc/passwd HTTP Response 302 /pypi?:action=/../../../../../../../../../../proc/self/environ%00&digest=517d7c7014d2b581f321d500eed4305e HTTP Response 302 ...which looks a lot like someone trying to hack information from the PyPI server, so thought I'd let you know... cheers, Chris From martin at v.loewis.de Fri Dec 2 23:42:45 2011 From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Fri, 02 Dec 2011 23:42:45 +0100 Subject: [Catalog-sig] Attempts to hack pypi? In-Reply-To: <4ED9192F.2000403@simplistix.co.uk> References: <20111202062506.A5EE84F8107@server1.simplistix.co.uk> <4ED9192F.2000403@simplistix.co.uk> Message-ID: <4ED95465.1080006@v.loewis.de> > ...which looks a lot like someone trying to hack information from the > PyPI server, so thought I'd let you know... People are encouraged to review the code, but I'm fairly sure that these attempts are futile. ISTM that this is a generic attack to try to fill out parameters that remotely look like file names with what the attacker thinks might also be valid filenames. You can find the :action processing in inner_run of https://svn.python.org/packages/trunk/pypi/webui.py Regards, Martin From martin at v.loewis.de Wed Dec 7 09:29:59 2011 From: martin at v.loewis.de (=?ISO-8859-15?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Wed, 07 Dec 2011 09:29:59 +0100 Subject: [Catalog-sig] PyPI operational changes Message-ID: <4EDF2407.4020301@v.loewis.de> Over the last few weeks, I did a number of changes to the PyPI installation, namely - replace Apache with nginx - replace FastCGI with uwsgi - full vacuum of postgres, and activate of autovacuum - introduce a separate uwsgi logging daemon Together, these changes seem to have a positive effect on stability of PyPI. Peak load average is down from 500 to about 10: http://pypi.python.org/munin/localdomain/localhost.localdomain/load.html I believe this is mainly due to switching from Apache to nginx. Apache would spawn hundreds of worker threads in an overload situation, which made things worse, not better. Memory consumption is down. Application memory would fluctuate up to 3.5G, and is now at 750M. Committed memory would increase up to 20G, and is now below 2G. Swap might did use up to 3G, and is now practically unused (7M). http://pypi.python.org/munin/localdomain/localhost.localdomain/memory.html Peak usage is again probably reduced to the change in process model between Apache and nginx; in addition, the rejuvenation features of uwsgi (replace worker process after 1000 requests) prevent Python processes from growing too much unused memory. Postgres response time is improved. There had been occasional transactions taking 1700s, and occasional queries taking 870s. This is now down to 45s/30s for the last day: http://pypi.python.org/munin/localdomain/localhost.localdomain/postgres_querylength_ALL.html There are two factors that likely cause this reduction. On the one hand, the postgres database wasn't vacuumed: http://pypi.python.org/munin/localdomain/localhost.localdomain/postgres_size_ALL.html The reason for the failure to autovacuum probably was that it was successively upgraded from 7.x release which didn't do autovacuum, and Debian at some point dropping the cron job that did the manual vacuum. Tables and indices now better fit into the address space, improving performance. Performing the full vacuum caused an outage of about 20 min two weeks ago. In addition, I set the uwsgi harakiri timeout to 60s, causing any query taking longer to be aborted. I believe such queries still occasionally happen; it's not clear to me what HTTP requests are triggering such long-running transactions. While I'm mostly happy with the current setup, one issue is that uwsgi doesn't support proper logrotation; in particular, it is unwilling to close-then-reopen the log files. Debian tries to use the copytruncate approach of logrotate, but that apparently didn't work too well (log space would constantly increase). I have now written a UDP server which supports proper log rotation and configured uwsgi to send log records to that UDP port. Regards, Martin From fred at fdrake.net Wed Dec 7 12:07:02 2011 From: fred at fdrake.net (Fred Drake) Date: Wed, 7 Dec 2011 06:07:02 -0500 Subject: [Catalog-sig] PyPI operational changes In-Reply-To: <4EDF2407.4020301@v.loewis.de> References: <4EDF2407.4020301@v.loewis.de> Message-ID: On Wed, Dec 7, 2011 at 3:29 AM, "Martin v. L?wis" wrote: > Together, these changes seem to have a positive effect on stability > of PyPI. Thanks, Martin! ? -Fred -- Fred L. Drake, Jr.? ? "A storm broke loose in my mind."? --Albert Einstein From sdouche at gmail.com Wed Dec 7 12:51:00 2011 From: sdouche at gmail.com (Sebastien Douche) Date: Wed, 7 Dec 2011 12:51:00 +0100 Subject: [Catalog-sig] PyPI operational changes In-Reply-To: <4EDF2407.4020301@v.loewis.de> References: <4EDF2407.4020301@v.loewis.de> Message-ID: On Wed, Dec 7, 2011 at 09:29, "Martin v. L?wis" wrote: > Together, these changes seem to have a positive effect on stability > of PyPI. Interesting feedback. Thanks Martin. -- Sebastien Douche Twitter: @sdouche / G+: +sdouche From chrism at plope.com Sun Dec 11 05:44:12 2011 From: chrism at plope.com (Chris McDonough) Date: Sat, 10 Dec 2011 23:44:12 -0500 Subject: [Catalog-sig] "python 2 only" classifier Message-ID: <1323578652.2710.295.camel@thinko> Hi, It'd be useful to have a Trove classifier that signified "the authors have no current intention to port this code to Python 3". Such a classifier could be used by e.g. http://python3wos.appspot.com/ to present a more accurate picture of the current state of porting efforts. Currently automated reports like that are bad PR because there's no way to communicate the intent. Perhaps: Programming Language :: Python :: 2 :: Only - C From robertc at robertcollins.net Sun Dec 11 18:41:23 2011 From: robertc at robertcollins.net (Robert Collins) Date: Mon, 12 Dec 2011 06:41:23 +1300 Subject: [Catalog-sig] trove - LGPL v3 not recognised? In-Reply-To: <20111115222849.GN2861@unaka.lan> References: <4EC035BC.2040905@v.loewis.de> <20111114003747.GG2861@unaka.lan> <20111115012828.GK2861@unaka.lan> <4EC20ADE.80500@v.loewis.de> <20111115222849.GN2861@unaka.lan> Message-ID: These are the ones I would like to see: > >License :: OSI Approved :: GNU Lesser General Public License v2 (LGPLv2) > >License :: OSI Approved :: GNU Lesser General Public License v3 (LGPLv3) > >License :: OSI Approved :: GNU Lesser General Public License v2 or later (LGPLv2+) > >License :: OSI Approved :: GNU Lesser General Public License v3 or later (LGPLv3+) > >License :: OSI Approved :: GNU General Public License v2 (GPLv2) > >License :: OSI Approved :: GNU General Public License v3 (GPLv3) > >License :: OSI Approved :: GNU General Public License v2 or later (GPLv2+) > >License :: OSI Approved :: GNU General Public License v3 or later (GPLv3+) Doing a 4th level would raise a bunch of interactions between legal intent - the above list seems like a good compromise. We've just had a user turn up confused around this again, so - do we have sufficient consensus to add these? -Rob From noah at coderanger.net Sun Dec 11 20:12:47 2011 From: noah at coderanger.net (Noah Kantrowitz) Date: Sun, 11 Dec 2011 11:12:47 -0800 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323578652.2710.295.camel@thinko> References: <1323578652.2710.295.camel@thinko> Message-ID: <56757B3A-27B3-445B-958B-B236DCA0D36F@coderanger.net> Is there a reason this needs to be a trove classifier? Just make some standard tag and use that :-) --Noah On Dec 10, 2011, at 8:44 PM, Chris McDonough wrote: > Hi, > > It'd be useful to have a Trove classifier that signified "the authors > have no current intention to port this code to Python 3". Such a > classifier could be used by e.g. http://python3wos.appspot.com/ to > present a more accurate picture of the current state of porting efforts. > Currently automated reports like that are bad PR because there's no way > to communicate the intent. > > Perhaps: > > Programming Language :: Python :: 2 :: Only > > - C > > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 163 bytes Desc: Message signed with OpenPGP using GPGMail URL: From chrism at plope.com Sun Dec 11 20:16:31 2011 From: chrism at plope.com (Chris McDonough) Date: Sun, 11 Dec 2011 14:16:31 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <56757B3A-27B3-445B-958B-B236DCA0D36F@coderanger.net> References: <1323578652.2710.295.camel@thinko> <56757B3A-27B3-445B-958B-B236DCA0D36F@coderanger.net> Message-ID: <1323630991.2710.303.camel@thinko> On Sun, 2011-12-11 at 11:12 -0800, Noah Kantrowitz wrote: > Is there a reason this needs to be a trove classifier? Just make some standard tag and use that :-) Please define "tag". - C > > --Noah > > On Dec 10, 2011, at 8:44 PM, Chris McDonough wrote: > > > Hi, > > > > It'd be useful to have a Trove classifier that signified "the authors > > have no current intention to port this code to Python 3". Such a > > classifier could be used by e.g. http://python3wos.appspot.com/ to > > present a more accurate picture of the current state of porting efforts. > > Currently automated reports like that are bad PR because there's no way > > to communicate the intent. > > > > Perhaps: > > > > Programming Language :: Python :: 2 :: Only > > > > - C > > > > > > _______________________________________________ > > Catalog-SIG mailing list > > Catalog-SIG at python.org > > http://mail.python.org/mailman/listinfo/catalog-sig > From noah at coderanger.net Sun Dec 11 22:01:07 2011 From: noah at coderanger.net (Noah Kantrowitz) Date: Sun, 11 Dec 2011 13:01:07 -0800 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323630991.2710.303.camel@thinko> References: <1323578652.2710.295.camel@thinko> <56757B3A-27B3-445B-958B-B236DCA0D36F@coderanger.net> <1323630991.2710.303.camel@thinko> Message-ID: <44CBC6FF-8E4F-4CB7-929A-33240822A4B4@coderanger.net> The "keywords" argument to setup(), which can then be searched via the XML-RPC API. --Noah On Dec 11, 2011, at 11:16 AM, Chris McDonough wrote: > On Sun, 2011-12-11 at 11:12 -0800, Noah Kantrowitz wrote: >> Is there a reason this needs to be a trove classifier? Just make some standard tag and use that :-) > > Please define "tag". > > - C > > >> >> --Noah >> >> On Dec 10, 2011, at 8:44 PM, Chris McDonough wrote: >> >>> Hi, >>> >>> It'd be useful to have a Trove classifier that signified "the authors >>> have no current intention to port this code to Python 3". Such a >>> classifier could be used by e.g. http://python3wos.appspot.com/ to >>> present a more accurate picture of the current state of porting efforts. >>> Currently automated reports like that are bad PR because there's no way >>> to communicate the intent. >>> >>> Perhaps: >>> >>> Programming Language :: Python :: 2 :: Only >>> >>> - C >>> >>> >>> _______________________________________________ >>> Catalog-SIG mailing list >>> Catalog-SIG at python.org >>> http://mail.python.org/mailman/listinfo/catalog-sig >> > > -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 163 bytes Desc: Message signed with OpenPGP using GPGMail URL: From chrism at plope.com Sun Dec 11 22:42:10 2011 From: chrism at plope.com (Chris McDonough) Date: Sun, 11 Dec 2011 16:42:10 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <44CBC6FF-8E4F-4CB7-929A-33240822A4B4@coderanger.net> References: <1323578652.2710.295.camel@thinko> <56757B3A-27B3-445B-958B-B236DCA0D36F@coderanger.net> <1323630991.2710.303.camel@thinko> <44CBC6FF-8E4F-4CB7-929A-33240822A4B4@coderanger.net> Message-ID: <1323639730.2710.314.camel@thinko> On Sun, 2011-12-11 at 13:01 -0800, Noah Kantrowitz wrote: > The "keywords" argument to setup(), which can then be searched via the XML-RPC API. > > --Noah The intent is to allow a site like http://python3wos.appspot.com/ to ignore distributions that signify an intent never to port to Python 3. To this extent, it has to be a marker that is widely understood and gains usage. It seems weird to say "ok yeah, tell us which versions of Python you support in the Trove classifiers here.. and yeah, over here too, except not in the Trove classifiers." I don't understand. - C > > On Dec 11, 2011, at 11:16 AM, Chris McDonough wrote: > > > On Sun, 2011-12-11 at 11:12 -0800, Noah Kantrowitz wrote: > >> Is there a reason this needs to be a trove classifier? Just make some standard tag and use that :-) > > > > Please define "tag". > > > > - C > > > > > >> > >> --Noah > >> > >> On Dec 10, 2011, at 8:44 PM, Chris McDonough wrote: > >> > >>> Hi, > >>> > >>> It'd be useful to have a Trove classifier that signified "the authors > >>> have no current intention to port this code to Python 3". Such a > >>> classifier could be used by e.g. http://python3wos.appspot.com/ to > >>> present a more accurate picture of the current state of porting efforts. > >>> Currently automated reports like that are bad PR because there's no way > >>> to communicate the intent. > >>> > >>> Perhaps: > >>> > >>> Programming Language :: Python :: 2 :: Only > >>> > >>> - C > >>> > >>> > >>> _______________________________________________ > >>> Catalog-SIG mailing list > >>> Catalog-SIG at python.org > >>> http://mail.python.org/mailman/listinfo/catalog-sig > >> > > > > > From martin at v.loewis.de Sun Dec 11 23:36:26 2011 From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Sun, 11 Dec 2011 23:36:26 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323578652.2710.295.camel@thinko> References: <1323578652.2710.295.camel@thinko> Message-ID: <4EE5306A.9050200@v.loewis.de> > It'd be useful to have a Trove classifier that signified "the authors > have no current intention to port this code to Python 3". That classifier already exists: "Programming Language :: Python :: 2" indicates support for Python 2, and absence of "Programming Language :: Python :: 3" then indicates that this version lacks Python 3 support. I don't think classifiers should be used to indicate some intention of package authors, in particular since the intention may change, but past releases will not. If you think that this is still different from what you are asking: What specific packages would be tagged with that classifier (I need two at least), and did that package authors agree to add the classifier to their package if it was available? Which specific classifier do you propose to add? Regards, Martin From richard at python.org Mon Dec 12 00:01:12 2011 From: richard at python.org (Richard Jones) Date: Mon, 12 Dec 2011 10:01:12 +1100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323578652.2710.295.camel@thinko> References: <1323578652.2710.295.camel@thinko> Message-ID: On 11 December 2011 15:44, Chris McDonough wrote: > It'd be useful to have a Trove classifier that signified "the authors > have no current intention to port this code to Python 3". ?Such a > classifier could be used by e.g. http://python3wos.appspot.com/ to > present a more accurate picture of the current state of porting efforts. > Currently automated reports like that are bad PR because there's no way > to communicate the intent. > > Perhaps: > > ?Programming Language :: Python :: 2 :: Only I think it could be a useful flag. Added. Richard From richard at python.org Mon Dec 12 00:02:42 2011 From: richard at python.org (Richard Jones) Date: Mon, 12 Dec 2011 10:02:42 +1100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <4EE5306A.9050200@v.loewis.de> References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> Message-ID: On 12 December 2011 09:36, "Martin v. L?wis" wrote: > If you think that this is still different from what you are asking: > What specific packages would be tagged with that classifier (I need > two at least), and did that package authors agree to add the classifier > to their package if it was available? Which specific classifier do you > propose to add? I believe the unittest2 package is a good example of a package that would be tagged with "python 2 only" as it represents a backport of the python 3 unittest module to python 2. Richard From fuzzyman at gmail.com Mon Dec 12 01:07:06 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Mon, 12 Dec 2011 00:07:06 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> Message-ID: On 11 December 2011 23:02, Richard Jones wrote: > On 12 December 2011 09:36, "Martin v. L?wis" wrote: > > If you think that this is still different from what you are asking: > > What specific packages would be tagged with that classifier (I need > > two at least), and did that package authors agree to add the classifier > > to their package if it was available? Which specific classifier do you > > propose to add? > > I believe the unittest2 package is a good example of a package that > would be tagged with "python 2 only" as it represents a backport of > the python 3 unittest module to python 2. > > Although there is a Python 3 compatible distribution of unittest2, which is currently called unittest2py3k. (It is maintained as a set of patches against CPython head - hence separate distributions for the same package names). Python 2 Only is still applicable to unittest2, until / unless I decide to use a single distribution for both. My only worry is that the classifier would seem to indicate that there *isn't* a Python 3 distribution available, when in fact there is. All the best, Michael > > Richard > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From a.badger at gmail.com Mon Dec 12 07:57:55 2011 From: a.badger at gmail.com (Toshio Kuratomi) Date: Sun, 11 Dec 2011 22:57:55 -0800 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <4EE5306A.9050200@v.loewis.de> References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> Message-ID: <20111212065755.GA8294@unaka.lan> On Sun, Dec 11, 2011 at 11:36:26PM +0100, "Martin v. L?wis" wrote: > > It'd be useful to have a Trove classifier that signified "the authors > > have no current intention to port this code to Python 3". > > That classifier already exists: "Programming Language :: Python :: 2" > indicates support for Python 2, and absence of "Programming Language :: > Python :: 3" then indicates that this version lacks Python 3 support. > I don't think classifiers should be used to indicate some intention of > package authors, in particular since the intention may change, but > past releases will not. > +1 Additionally, I'm not certain of the use case cited. Isn't the utility of a site like http://python3wos.appspot.com/ in seeing which popular or widely depended upon packages have no python3 version? With that in mind, the listing on that page wouldn't seem to depend on whether a package's author intends to port to python3. It depends more on whether the package has some well-known python3 counterpart that dependent code can be (easily?) ported to use instead. I'm thinking here of things like: try: from simplejson import loads except ImportError: from json import loads Seems kind of a miscategorizaion to use the Programming Language :: classifier for that.... Some sort of classifier that indicates that the software is only relevant for a certain version of python seems better. Looking at the existing catgories, maybe something like Environment :: Python :: 2 would be appropriate. (Although, ugh,I think that someone jumped the gun and added Programming Language :: Python :: 2 :: Only without waiting for the results of this discussion to finish. Not sure if that can be reverted or if it's already too late.) -Toshio -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 198 bytes Desc: not available URL: From chrism at plope.com Mon Dec 12 08:28:13 2011 From: chrism at plope.com (Chris McDonough) Date: Mon, 12 Dec 2011 02:28:13 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <4EE5306A.9050200@v.loewis.de> References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> Message-ID: <1323674893.2710.343.camel@thinko> On Sun, 2011-12-11 at 23:36 +0100, "Martin v. L?wis" wrote: > > It'd be useful to have a Trove classifier that signified "the authors > > have no current intention to port this code to Python 3". > > That classifier already exists: "Programming Language :: Python :: 2" > indicates support for Python 2, and absence of "Programming Language :: > Python :: 3" then indicates that this version lacks Python 3 support. > I don't think classifiers should be used to indicate some intention of > package authors, in particular since the intention may change, but > past releases will not. Thanks to Richard for adding the classifier. Point taken about past releases; lots of packages on PyPI have already seen their last release. But if the current party line in python-dev that goes something like "if you don't want to port to Python 3, you don't have to" is to be credible, then ongoing maintainers who choose not to port to Python 3 should not be subject to pressure exhibited by sites like the "wall of shame". Being able to express the no-port intent explicitly relieves the stupidest of those pressures. I'm sympathetic to the opinion "the wall of shame should not make a difference". But we have a PR problem, because the wall of shame is linked endlessly from comments about Python 3 stores in every news aggregator (Reddit, Hacker News, etc). People tend to use it as a stick to beat the helpless with. > If you think that this is still different from what you are asking: > What specific packages would be tagged with that classifier (I need > two at least), and did that package authors agree to add the classifier > to their package if it was available? Which specific classifier do you > propose to add? The ones that I have some direct responsibility for are: - Supervisor - meld3 The ones that I am one step removed from and can venture a very educated guess about because I participate on the related maillists and occasionally contribute are: - Routes - Pylons - zdaemon - ZConfig - most zope.* packages that aren't already py3 compat - The Zope2 package - The Plone package - The Acquisition package (Zope-related) - most zc.* packages - all five.* packages (Zope-related) - all plone.* packages - borg.localrole (Plone) - all archetypes.* packages (Plone) - all kss.* packages (Plone) - all Products.* packages (Zope2 products) - RestrictedPython All of these packages have > 90K downloads and are featured prominently on the wall of shame (they actually form a large portion of the "unported" stuff; 65 packages or so out of about 120). To the extent that the above packages will see any new release, I suspect their maintainers would be happy to mark them "Python 2 only". - C From fuzzyman at gmail.com Mon Dec 12 09:48:08 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Mon, 12 Dec 2011 08:48:08 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323674893.2710.343.camel@thinko> References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> <1323674893.2710.343.camel@thinko> Message-ID: On 12 December 2011 07:28, Chris McDonough wrote: > On Sun, 2011-12-11 at 23:36 +0100, "Martin v. L?wis" wrote: > > > It'd be useful to have a Trove classifier that signified "the authors > > > have no current intention to port this code to Python 3". > > > > That classifier already exists: "Programming Language :: Python :: 2" > > indicates support for Python 2, and absence of "Programming Language :: > > Python :: 3" then indicates that this version lacks Python 3 support. > > I don't think classifiers should be used to indicate some intention of > > package authors, in particular since the intention may change, but > > past releases will not. > > Thanks to Richard for adding the classifier. > > Point taken about past releases; lots of packages on PyPI have already > seen their last release. > > But if the current party line in python-dev that goes something like "if > you don't want to port to Python 3, you don't have to" is to be > credible, then ongoing maintainers who choose not to port to Python 3 > should not be subject to pressure exhibited by sites like the "wall of > shame". Being able to express the no-port intent explicitly relieves > the stupidest of those pressures. > > I'm sympathetic to the opinion "the wall of shame should not make a > difference". But we have a PR problem, because the wall of shame is > linked endlessly from comments about Python 3 stores in every news > aggregator (Reddit, Hacker News, etc). People tend to use it as a stick > to beat the helpless with. > > Well sure - but python-dev has absolutely nothing to do with the wall-of-shame. In addition whoever is responsible for it, is very unresponsive and apparently not interested in fixing factual errors in the list. So I doubt the new classifier will have any effect on that website I'm afraid. :-( Michael Foord > > If you think that this is still different from what you are asking: > > What specific packages would be tagged with that classifier (I need > > two at least), and did that package authors agree to add the classifier > > to their package if it was available? Which specific classifier do you > > propose to add? > > The ones that I have some direct responsibility for are: > > - Supervisor > - meld3 > > The ones that I am one step removed from and can venture a very educated > guess about because I participate on the related maillists and > occasionally contribute are: > > - Routes > - Pylons > - zdaemon > - ZConfig > - most zope.* packages that aren't already py3 compat > - The Zope2 package > - The Plone package > - The Acquisition package (Zope-related) > - most zc.* packages > - all five.* packages (Zope-related) > - all plone.* packages > - borg.localrole (Plone) > - all archetypes.* packages (Plone) > - all kss.* packages (Plone) > - all Products.* packages (Zope2 products) > - RestrictedPython > > All of these packages have > 90K downloads and are featured prominently > on the wall of shame (they actually form a large portion of the > "unported" stuff; 65 packages or so out of about 120). To the extent > that the above packages will see any new release, I suspect their > maintainers would be happy to mark them "Python 2 only". > > - C > > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrism at plope.com Mon Dec 12 09:51:10 2011 From: chrism at plope.com (Chris McDonough) Date: Mon, 12 Dec 2011 03:51:10 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> <1323674893.2710.343.camel@thinko> Message-ID: <1323679870.2710.352.camel@thinko> On Mon, 2011-12-12 at 08:48 +0000, Michael Foord wrote: > I'm sympathetic to the opinion "the wall of shame should not > make a > difference". But we have a PR problem, because the wall of > shame is > linked endlessly from comments about Python 3 stores in every > news > aggregator (Reddit, Hacker News, etc). People tend to use it > as a stick > to beat the helpless with. > Well sure - but python-dev has absolutely nothing to do with the > wall-of-shame. In addition whoever is responsible for it, is very > unresponsive and apparently not interested in fixing factual errors in > the list. So I doubt the new classifier will have any effect on that > website I'm afraid. :-( Jesse seemed to think he was reachable and receptive. I'll try to get a hold of him; if he doesn't respond or responds in the negative, I won't advertise the classifier. - C From fuzzyman at gmail.com Mon Dec 12 09:56:27 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Mon, 12 Dec 2011 08:56:27 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323679870.2710.352.camel@thinko> References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> <1323674893.2710.343.camel@thinko> <1323679870.2710.352.camel@thinko> Message-ID: On 12 December 2011 08:51, Chris McDonough wrote: > On Mon, 2011-12-12 at 08:48 +0000, Michael Foord wrote: > > > I'm sympathetic to the opinion "the wall of shame should not > > make a > > difference". But we have a PR problem, because the wall of > > shame is > > linked endlessly from comments about Python 3 stores in every > > news > > aggregator (Reddit, Hacker News, etc). People tend to use it > > as a stick > > to beat the helpless with. > > > Well sure - but python-dev has absolutely nothing to do with the > > wall-of-shame. In addition whoever is responsible for it, is very > > unresponsive and apparently not interested in fixing factual errors in > > the list. So I doubt the new classifier will have any effect on that > > website I'm afraid. :-( > > Jesse seemed to think he was reachable and receptive. I'll try to get a > hold of him; if he doesn't respond or responds in the negative, I won't > advertise the classifier. > > Ah, ok. Sounds good. I've tried contacting the maintainer in the past about errors and not had any response - you and Jesse may have better luck. Michael - C > > > > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From tjreedy at udel.edu Mon Dec 12 10:17:28 2011 From: tjreedy at udel.edu (Terry Reedy) Date: Mon, 12 Dec 2011 04:17:28 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323674893.2710.343.camel@thinko> References: <1323578652.2710.295.camel@thinko> <4EE5306A.9050200@v.loewis.de> <1323674893.2710.343.camel@thinko> Message-ID: On 12/12/2011 2:28 AM, Chris McDonough wrote: > But if the current party line in python-dev that goes something like "if > you don't want to port to Python 3, you don't have to" is to be > credible, then ongoing maintainers who choose not to port to Python 3 > should not be subject to pressure exhibited by sites like the "wall of > shame". I am a sometimes developer who has no idea what site you are talking about. But given that it exists and gets way too much attention, I can the usefulness of the proposal, which seemed senseless until I read this post. There is a fairly short list of the best and most widely used library packages that I *hope* will either get ported or replaced by a new Py 3 package. Basically, enough so that 'don't use Python 3 because it lacks 3rd-party libraries' can be put to rest. But I expect that substantial portions of the stuff on PyPI never will be. -- Terry Jan Reedy From ubershmekel at gmail.com Tue Dec 13 11:29:13 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Tue, 13 Dec 2011 12:29:13 +0200 Subject: [Catalog-sig] "python 2 only" classifier Message-ID: Hi catalog-sig, I'm the wall of shame guy, This was just brought to my attention by Chris via reddit so I'm sorry for chiming in late. Toshio Kuratomi is correct: > Additionally, I'm not certain of the use case cited. Isn't the utility of > a site like http://python3wos.appspot.com/ in seeing which popular or > widely depended upon packages have no python3 version? With that in mind, > the listing on that page wouldn't seem to depend on whether a package's > author intends to port to python3. 1. If this forum likes it I'll add a lock icon next to packages with the "Programming Language :: Python :: 2 :: Only" classifier with the following mouse-over: "The maintainers of this code have declared it will run on Python 2 only for the foreseeable future". This might lessen the social pressure on the maintainers to port. I don't think it'll help the PR issue. 2. A more useful classifier (in terms of PR) could be "Programming Language :: Python :: 2 :: Py3k equivalent exists". Which would denote that I can entirely remove the red package from the wall. This is currently done manually for the following: 'multiprocessing', 'simplejson', 'argparse', 'uuid', 'setuptools', 'Jinja'. So feel free to tell me if there are more I should add to this list. A classifier would mean more work for everyone, though it is the pure and correct way to go about this. And I hope I'm not considered the bad guy here. Please do contact me with any questions or suggestions. I only want to promote python 3 though I know the site has a dual edge to its sword. --Yuval From fuzzyman at gmail.com Tue Dec 13 13:17:54 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Tue, 13 Dec 2011 12:17:54 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: On 13 December 2011 10:29, Yuval Greenfield wrote: > Hi catalog-sig, I'm the wall of shame guy, > > This was just brought to my attention by Chris via reddit so I'm sorry > for chiming in late. > > Toshio Kuratomi is correct: > > Additionally, I'm not certain of the use case cited. Isn't the utility > of > > a site like http://python3wos.appspot.com/ in seeing which popular or > > widely depended upon packages have no python3 version? With that in > mind, > > the listing on that page wouldn't seem to depend on whether a package's > > author intends to port to python3. > > 1. If this forum likes it I'll add a lock icon next to packages with > the "Programming Language :: Python :: 2 :: Only" classifier with the > following mouse-over: "The maintainers of this code have declared it > will run on Python 2 only for the foreseeable future". This might > lessen the social pressure on the maintainers to port. I don't think > it'll help the PR issue. > > 2. A more useful classifier (in terms of PR) could be "Programming > Language :: Python :: 2 :: Py3k equivalent exists". Which would denote > that I can entirely remove the red package from the wall. This is > currently done manually for the following: 'multiprocessing', > 'simplejson', 'argparse', 'uuid', 'setuptools', 'Jinja'. So feel free > to tell me if there are more I should add to this list. unittest2 has a Python 3 port: unittest2py3k docutils is also Python 3 compatible (since version 0.6), even if they're not using the trove classifier. http://docutils.sourceforge.net/ All the best, Michael Foord > A classifier > would mean more work for everyone, though it is the pure and correct > way to go about this. > > And I hope I'm not considered the bad guy here. Please do contact me > with any questions or suggestions. I only want to promote python 3 > though I know the site has a dual edge to its sword. > > > --Yuval > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From ubershmekel at gmail.com Tue Dec 13 14:24:02 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Tue, 13 Dec 2011 15:24:02 +0200 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord wrote: > On 13 December 2011 10:29, Yuval Greenfield wrote: >> Toshio Kuratomi is correct: >> > Additionally, I'm not certain of the use case cited. ?Isn't the utility >> > of >> > a site like http://python3wos.appspot.com/ in seeing which popular or >> > widely depended upon packages have no python3 version? ?With that in >> > mind, >> > the listing on that page wouldn't seem to depend on whether a package's >> > author intends to port to python3. >> > > unittest2 has a Python 3 port: unittest2py3k > > docutils is also Python 3 compatible (since version 0.6), even if they're > not using the trove classifier. > My initial response was to remove modules that had python 3 equivalents on the wall. This makes sense for eg setuptools and jinja as it wouldn't be accurate to give them a double green listing and skew the percentage. Packages that have a single listing on the top 200 and have a python 3 equivalent I should mark as green. I guess the best solution would be to consolidate these packages. I'll work on it. Concerning docutils, I sent an email to Lea Wiemann (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove classifier though I don't think I can be held responsible for this one. Yuval Greenfield From fuzzyman at gmail.com Tue Dec 13 14:39:58 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Tue, 13 Dec 2011 13:39:58 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: On 13 December 2011 13:24, Yuval Greenfield wrote: > On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord wrote: > > On 13 December 2011 10:29, Yuval Greenfield > wrote: > >> Toshio Kuratomi is correct: > >> > Additionally, I'm not certain of the use case cited. Isn't the > utility > >> > of > >> > a site like http://python3wos.appspot.com/ in seeing which popular or > >> > widely depended upon packages have no python3 version? With that in > >> > mind, > >> > the listing on that page wouldn't seem to depend on whether a > package's > >> > author intends to port to python3. > >> > > > > unittest2 has a Python 3 port: unittest2py3k > > > > docutils is also Python 3 compatible (since version 0.6), even if they're > > not using the trove classifier. > > > > My initial response was to remove modules that had python 3 > equivalents on the wall. This makes sense for eg setuptools and jinja > as it wouldn't be accurate to give them a double green listing and > skew the percentage. > > Packages that have a single listing on the top 200 and have a python 3 > equivalent I should mark as green. I guess the best solution would be > to consolidate these packages. I'll work on it. > > Concerning docutils, I sent an email to Lea Wiemann > (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove > classifier though I don't think I can be held responsible for this > one. > What do you mean by held responsible? If a package is Python 3 compatible, or has a Python 3 version/port/alternative distribution, shouldn't "wall of shame" (I *really* dislike the name for what it's worth and cringe every time I write it) reflect the truth? All the best, Michael > > > Yuval Greenfield > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From ubershmekel at gmail.com Tue Dec 13 14:50:50 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Tue, 13 Dec 2011 15:50:50 +0200 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: I'm not sure this is the case, but easy_install/pip should check if the package is python 3 compatible or not before installing. To quote you: If a package is Python 3 compatible, or has a Python 3 version/port/alternative distribution, shouldn't "The Python Package Index" reflect the truth? Yuval Greenfield On Tue, Dec 13, 2011 at 3:39 PM, Michael Foord wrote: > > > On 13 December 2011 13:24, Yuval Greenfield wrote: >> >> On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord wrote: >> > On 13 December 2011 10:29, Yuval Greenfield >> > wrote: >> >> Toshio Kuratomi is correct: >> >> > Additionally, I'm not certain of the use case cited. ?Isn't the >> >> > utility >> >> > of >> >> > a site like http://python3wos.appspot.com/ in seeing which popular or >> >> > widely depended upon packages have no python3 version? ?With that in >> >> > mind, >> >> > the listing on that page wouldn't seem to depend on whether a >> >> > package's >> >> > author intends to port to python3. >> >> >> > >> > unittest2 has a Python 3 port: unittest2py3k >> > >> > docutils is also Python 3 compatible (since version 0.6), even if >> > they're >> > not using the trove classifier. >> > >> >> My initial response was to remove modules that had python 3 >> equivalents on the wall. This makes sense for eg setuptools and jinja >> as it wouldn't be accurate to give them a double green listing and >> skew the percentage. >> >> Packages that have a single listing on the top 200 and have a python 3 >> equivalent I should mark as green. I guess the best solution would be >> to consolidate these packages. I'll work on it. >> >> Concerning docutils, I sent an email to Lea Wiemann >> (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove >> classifier though I don't think I can be held responsible for this >> one. > > > > What do you mean by held responsible? If a package is Python 3 compatible, > or has a Python 3 version/port/alternative distribution, shouldn't "wall of > shame" (I *really* dislike the name for what it's worth and cringe every > time I write it) reflect the truth? > > All the best, > > Michael > >> >> >> >> Yuval Greenfield > > > > > -- > > http://www.voidspace.org.uk/ > > May you do good and not evil > May you find forgiveness for yourself and forgive others > > May you share freely, never taking more than you give. > -- the sqlite blessing http://www.sqlite.org/different.html > > From fuzzyman at gmail.com Tue Dec 13 15:22:38 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Tue, 13 Dec 2011 14:22:38 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: On 13 December 2011 13:50, Yuval Greenfield wrote: > I'm not sure this is the case, but easy_install/pip should check if > the package is python 3 compatible or not before installing. To quote > you: > > If a package is Python 3 compatible, or has a Python 3 > version/port/alternative distribution, shouldn't "The Python Package > Index" reflect the truth? > My initial reply went privately to Yuval (goddamn lists that don't have reply-to set correctly ;-). Anyway, even though the wall-of-shame states that it is based on the PyPI classifier, I think it would be less of a publicity own goal for the Python community, and would be much more *useful*, if packages that did support Python 3 showed up as green... All the best, Michael Foord > > Yuval Greenfield > > On Tue, Dec 13, 2011 at 3:39 PM, Michael Foord wrote: > > > > > > On 13 December 2011 13:24, Yuval Greenfield > wrote: > >> > >> On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord > wrote: > >> > On 13 December 2011 10:29, Yuval Greenfield > >> > wrote: > >> >> Toshio Kuratomi is correct: > >> >> > Additionally, I'm not certain of the use case cited. Isn't the > >> >> > utility > >> >> > of > >> >> > a site like http://python3wos.appspot.com/ in seeing which > popular or > >> >> > widely depended upon packages have no python3 version? With that > in > >> >> > mind, > >> >> > the listing on that page wouldn't seem to depend on whether a > >> >> > package's > >> >> > author intends to port to python3. > >> >> > >> > > >> > unittest2 has a Python 3 port: unittest2py3k > >> > > >> > docutils is also Python 3 compatible (since version 0.6), even if > >> > they're > >> > not using the trove classifier. > >> > > >> > >> My initial response was to remove modules that had python 3 > >> equivalents on the wall. This makes sense for eg setuptools and jinja > >> as it wouldn't be accurate to give them a double green listing and > >> skew the percentage. > >> > >> Packages that have a single listing on the top 200 and have a python 3 > >> equivalent I should mark as green. I guess the best solution would be > >> to consolidate these packages. I'll work on it. > >> > >> Concerning docutils, I sent an email to Lea Wiemann > >> (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove > >> classifier though I don't think I can be held responsible for this > >> one. > > > > > > > > What do you mean by held responsible? If a package is Python 3 > compatible, > > or has a Python 3 version/port/alternative distribution, shouldn't "wall > of > > shame" (I *really* dislike the name for what it's worth and cringe every > > time I write it) reflect the truth? > > > > All the best, > > > > Michael > > > >> > >> > >> > >> Yuval Greenfield > > > > > > > > > > -- > > > > http://www.voidspace.org.uk/ > > > > May you do good and not evil > > May you find forgiveness for yourself and forgive others > > > > May you share freely, never taking more than you give. > > -- the sqlite blessing http://www.sqlite.org/different.html > > > > > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From fuzzyman at gmail.com Tue Dec 13 15:45:52 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Tue, 13 Dec 2011 14:45:52 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: On 13 December 2011 14:36, David Goodger wrote: > Sorry, I have no idea what this message is about. > Some context/explanation please? > What action are you asking for? > docutils supports Python 3 (according to the project website), but does not use the Python 3 trove classifier on PyPI. That means it incorrectly shows up as not supporting Python 3 on the "wall of shame" website. http://python3wos.appspot.com/ All the best, Michael Foord > > Note that Lea Wiemann is no longer active in the Docutils project. > > -- David Goodger > > On Tue, Dec 13, 2011 at 08:24, Yuval Greenfield > wrote: > > On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord > wrote: > >> On 13 December 2011 10:29, Yuval Greenfield > wrote: > >>> Toshio Kuratomi is correct: > >>> > Additionally, I'm not certain of the use case cited. Isn't the > utility > >>> > of > >>> > a site like http://python3wos.appspot.com/ in seeing which popular > or > >>> > widely depended upon packages have no python3 version? With that in > >>> > mind, > >>> > the listing on that page wouldn't seem to depend on whether a > package's > >>> > author intends to port to python3. > >>> > >> > >> unittest2 has a Python 3 port: unittest2py3k > >> > >> docutils is also Python 3 compatible (since version 0.6), even if > they're > >> not using the trove classifier. > >> > > > > My initial response was to remove modules that had python 3 > > equivalents on the wall. This makes sense for eg setuptools and jinja > > as it wouldn't be accurate to give them a double green listing and > > skew the percentage. > > > > Packages that have a single listing on the top 200 and have a python 3 > > equivalent I should mark as green. I guess the best solution would be > > to consolidate these packages. I'll work on it. > > > > Concerning docutils, I sent an email to Lea Wiemann > > (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove > > classifier though I don't think I can be held responsible for this > > one. > > > > > > Yuval Greenfield > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From goodger at python.org Tue Dec 13 15:36:29 2011 From: goodger at python.org (David Goodger) Date: Tue, 13 Dec 2011 09:36:29 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: Sorry, I have no idea what this message is about. Some context/explanation please? What action are you asking for? Note that Lea Wiemann is no longer active in the Docutils project. -- David Goodger On Tue, Dec 13, 2011 at 08:24, Yuval Greenfield wrote: > On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord wrote: >> On 13 December 2011 10:29, Yuval Greenfield wrote: >>> Toshio Kuratomi is correct: >>> > Additionally, I'm not certain of the use case cited. ?Isn't the utility >>> > of >>> > a site like http://python3wos.appspot.com/ in seeing which popular or >>> > widely depended upon packages have no python3 version? ?With that in >>> > mind, >>> > the listing on that page wouldn't seem to depend on whether a package's >>> > author intends to port to python3. >>> >> >> unittest2 has a Python 3 port: unittest2py3k >> >> docutils is also Python 3 compatible (since version 0.6), even if they're >> not using the trove classifier. >> > > My initial response was to remove modules that had python 3 > equivalents on the wall. This makes sense for eg setuptools and jinja > as it wouldn't be accurate to give them a double green listing and > skew the percentage. > > Packages that have a single listing on the top 200 and have a python 3 > equivalent I should mark as green. I guess the best solution would be > to consolidate these packages. I'll work on it. > > Concerning docutils, I sent an email to Lea Wiemann > (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove > classifier though I don't think I can be held responsible for this > one. > > > Yuval Greenfield From fuzzyman at gmail.com Tue Dec 13 17:21:24 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Tue, 13 Dec 2011 16:21:24 +0000 Subject: [Catalog-sig] Hosting documentation on PyPI Message-ID: Hey all, PyPI does documentation hosting, which is awesome. Are there docs or a tutorial on using this feature? PyPI itself links to this page, which is a dead link: http://wiki.python.org/moin/PyPiHostingDocumentation Linked to from: http://pypi.python.org/pypi All the best, Michael Foord -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From a.badger at gmail.com Tue Dec 13 17:57:52 2011 From: a.badger at gmail.com (Toshio Kuratomi) Date: Tue, 13 Dec 2011 08:57:52 -0800 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: <20111213165752.GJ8294@unaka.lan> On Tue, Dec 13, 2011 at 09:36:29AM -0500, David Goodger wrote: > Sorry, I have no idea what this message is about. > Some context/explanation please? > What action are you asking for? > > Note that Lea Wiemann is no longer active in the Docutils project. > I think Yuval is asking that the pypi page for docutils be updated to have:: Programming Language :: Python :: 3 as well as:: Programming Language :: Python :: 2 I'm not sure if that's right, though -- I just checked the docutils page and it has :: Programming Language :: Python which seems to be a valid method of marking a module as belonging to multiple sub-categories. (This thread has some other ideas as well, but those don't directly impact docutils) -Toshio > -- David Goodger > > On Tue, Dec 13, 2011 at 08:24, Yuval Greenfield wrote: > > On Tue, Dec 13, 2011 at 2:17 PM, Michael Foord wrote: > >> On 13 December 2011 10:29, Yuval Greenfield wrote: > >>> Toshio Kuratomi is correct: > >>> > Additionally, I'm not certain of the use case cited. ?Isn't the utility > >>> > of > >>> > a site like http://python3wos.appspot.com/ in seeing which popular or > >>> > widely depended upon packages have no python3 version? ?With that in > >>> > mind, > >>> > the listing on that page wouldn't seem to depend on whether a package's > >>> > author intends to port to python3. > >>> > >> > >> unittest2 has a Python 3 port: unittest2py3k > >> > >> docutils is also Python 3 compatible (since version 0.6), even if they're > >> not using the trove classifier. > >> > > > > My initial response was to remove modules that had python 3 > > equivalents on the wall. This makes sense for eg setuptools and jinja > > as it wouldn't be accurate to give them a double green listing and > > skew the percentage. > > > > Packages that have a single listing on the top 200 and have a python 3 > > equivalent I should mark as green. I guess the best solution would be > > to consolidate these packages. I'll work on it. > > > > Concerning docutils, I sent an email to Lea Wiemann > > (LeWiemann at gmail.com) and am cc-ing David Goodger for the trove > > classifier though I don't think I can be held responsible for this > > one. > > > > > > Yuval Greenfield > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 198 bytes Desc: not available URL: From carl at oddbird.net Tue Dec 13 18:51:03 2011 From: carl at oddbird.net (Carl Meyer) Date: Tue, 13 Dec 2011 10:51:03 -0700 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: <4EE79087.4010101@oddbird.net> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 12/13/2011 06:50 AM, Yuval Greenfield wrote: > I'm not sure this is the case, but easy_install/pip should check if > the package is python 3 compatible or not before installing. Pip does not check the trove classifiers - I'm not aware of any other way to check for Python version support. I'm not sure that it would be a good idea for pip to start checking trove classifiers - too many packages just don't use them, or haven't updated them correctly. I guess it could be an overridable warning, perhaps. And I suppose that would encourage people to start setting them correctly... Carl -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.10 (GNU/Linux) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk7nkIcACgkQ8W4rlRKtE2f7EwCglViVM1gDFHqnKMHg+/3n8dYI rvIAn0fP22eEC7vGsbGvlmoqppBrlFOI =HrZI -----END PGP SIGNATURE----- From martin at v.loewis.de Wed Dec 14 01:11:58 2011 From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Wed, 14 Dec 2011 01:11:58 +0100 Subject: [Catalog-sig] Hosting documentation on PyPI In-Reply-To: References: Message-ID: <4EE7E9CE.9040803@v.loewis.de> > PyPI does documentation hosting, which is awesome. Are there docs or a > tutorial on using this feature? As a package author, if you go to your package's page, you find documentation for that. Regards, Martin From fuzzyman at gmail.com Wed Dec 14 01:26:59 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Wed, 14 Dec 2011 00:26:59 +0000 Subject: [Catalog-sig] Hosting documentation on PyPI In-Reply-To: <4EE7E9CE.9040803@v.loewis.de> References: <4EE7E9CE.9040803@v.loewis.de> Message-ID: On 14 December 2011 00:11, "Martin v. L?wis" wrote: > > PyPI does documentation hosting, which is awesome. Are there docs or a > > tutorial on using this feature? > > As a package author, if you go to your package's page, you find > documentation for that. > > Thanks I found it. I was trying to find a link I could give to a friend who wanted instructions. All the best, Michael Foord > Regards, > Martin > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrism at plope.com Wed Dec 14 04:36:20 2011 From: chrism at plope.com (Chris McDonough) Date: Tue, 13 Dec 2011 22:36:20 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: <1323833780.3502.42.camel@thinko> On Tue, 2011-12-13 at 12:29 +0200, Yuval Greenfield wrote: > 1. If this forum likes it I'll add a lock icon next to packages with > the "Programming Language :: Python :: 2 :: Only" classifier with the > following mouse-over: "The maintainers of this code have declared it > will run on Python 2 only for the foreseeable future". This might > lessen the social pressure on the maintainers to port. I don't think > it'll help the PR issue. For the record, at least one package is already using the new "2-only" classifier now: http://pypi.python.org/pypi/Pylons/1.0.1rc1 From tjreedy at udel.edu Wed Dec 14 06:55:04 2011 From: tjreedy at udel.edu (Terry Reedy) Date: Wed, 14 Dec 2011 00:55:04 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: On 12/13/2011 5:29 AM, Yuval Greenfield wrote: > Hi catalog-sig, I'm the wall of shame guy, > And I hope I'm not considered the bad guy here. Please do contact me > with any questions or suggestions. I only want to promote python 3 > though I know the site has a dual edge to its sword. I promote Python 3 also, but I would have called such a site something like python3progress. I would make the second sentence something "The packages listed in [green] are those known to work with Python 3 or have an equivalent that does." Also, list the equivalents somehow. Please change the page, if not the url. I am really sure that doing so will promote Python 3 better than the current page. -- Terry Jan Reedy From chrism at plope.com Thu Dec 15 06:01:35 2011 From: chrism at plope.com (Chris McDonough) Date: Thu, 15 Dec 2011 00:01:35 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: Message-ID: <1323925295.2556.48.camel@thinko> On Tue, 2011-12-13 at 12:29 +0200, Yuval Greenfield wrote: > 2. A more useful classifier (in terms of PR) could be "Programming > Language :: Python :: 2 :: Py3k equivalent exists". Which would denote > that I can entirely remove the red package from the wall. This is > currently done manually for the following: 'multiprocessing', > 'simplejson', 'argparse', 'uuid', 'setuptools', 'Jinja'. So feel free > to tell me if there are more I should add to this list. "zc.buildout" is actually Python 3.2 compatible now, FWIW. Its 2.X release line is 3.X compatible but the author doesn't want that to be the default version pulled down yet, so he has "egg-only" releases of it on PyPI which prevents easy_install/pip from treating it as "most recent" for users of Python 3. However, if you use Python 3, and ask for it, you indeed get a Python 3 compat version, - C From martin at v.loewis.de Fri Dec 16 00:40:12 2011 From: martin at v.loewis.de (=?ISO-8859-15?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Fri, 16 Dec 2011 00:40:12 +0100 Subject: [Catalog-sig] PyPI outage Message-ID: <4EEA855C.90309@v.loewis.de> I just rebooted PyPI for maintenance, and that reboot failed (or the file system check took some long time without reporting back). After waiting for 20 minutes, I power-cycled the machine, which brought it back. Regards, Martin From ubershmekel at gmail.com Fri Dec 16 02:53:54 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Fri, 16 Dec 2011 03:53:54 +0200 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1323833780.3502.42.camel@thinko> References: <1323833780.3502.42.camel@thinko> Message-ID: On Wed, Dec 14, 2011 at 5:36 AM, Chris McDonough wrote: > On Tue, 2011-12-13 at 12:29 +0200, Yuval Greenfield wrote: > > > 1. If this forum likes it I'll add a lock icon next to packages with > > the "Programming Language :: Python :: 2 :: Only" classifier with the > > following mouse-over: "The maintainers of this code have declared it > > will run on Python 2 only for the foreseeable future". This might > > lessen the social pressure on the maintainers to port. I don't think > > it'll help the PR issue. > > For the record, at least one package is already using the new "2-only" > classifier now: > > http://pypi.python.org/pypi/Pylons/1.0.1rc1 > > > > I added both features (python2only and py3equivalents). It took a lot longer than I expected and cost $0.10 of GAE quotas... Please do critique: http://python3wos.appspot.com/ Yuval Greenfield -------------- next part -------------- An HTML attachment was scrubbed... URL: From lists at zopyx.com Fri Dec 16 09:28:48 2011 From: lists at zopyx.com (Andreas Jung) Date: Fri, 16 Dec 2011 09:28:48 +0100 Subject: [Catalog-sig] Massive problems with PyPI since days Message-ID: <4EEB0140.3020205@zopyx.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi there, we encounter lots of buildout problems (triggered by various Jenkins instances running at different hosters) similar to: Restored 'README.txt' Restored 'dev.cfg' A products A products/README.txt Checked out revision 6884. + bin/python bootstrap.py Downloading http://pypi.python.org/packages/source/d/distribute/distribute-0.6.24.tar.gz Traceback (most recent call last): File "bootstrap.py", line 67, in ez['use_setuptools'](to_dir=tmpeggs, download_delay=0, no_fake=True) File "", line 145, in use_setuptools File "", line 124, in _do_download File "", line 193, in download_setuptools File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 124, in urlopen return _opener.open(url, data, timeout) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 389, in open response = self._open(req, data) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 407, in _open '_open', req) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 367, in _call_chain result = func(*args) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1146, in http_open return self.do_open(httplib.HTTPConnection, req) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1121, in do_open raise URLError(err) urllib2.URLError: Recording test results Is there any severe problem with PyPI? These errors come and come. Triggering a Jenkins manually after such a failure often results in a successful build. Andreas - -- ZOPYX Limited | zopyx group Charlottenstr. 37/1 | The full-service network for Zope & Plone D-72070 T?bingen | Produce & Publish www.zopyx.com | www.produce-and-publish.com - ------------------------------------------------------------------------ E-Publishing, Python, Zope & Plone development, Consulting -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQGUBAEBAgAGBQJO6wE/AAoJEADcfz7u4AZjvdkLwIqA3dtNFig20L4sQPikEiXq c1YNeZ8qaZTcGN3urvnVP9XXG8wGd8eebGP/Y/hwSzyvtLFxDgrx0XEmbIFgAjCL 7t1tjiHScG/rN/F8ABg0CXBHqI9sjbzJ12ZmrivmzJLhlqE/BKsEQOuAQ6ThgwmZ 2MqfxNkjWqndN8MIgvy7E1tQBH5McuEEDOdwe/uMXeXxy/jpJwiFEKqqBhMjEULN PTxxF6RJGDJ3TiXK51dZkSJse8IcwRnm4A2mb828Ig3sfX4RTrvJ3Hl0kzyXqVl0 eH28/10Dbi67GH3uWtZ21+8kQ1WD3hGvsqTxNAoBJoPF+/2tLSmdC6WsLpa6+q/v vdjlNOf0mvzSM5g8EiJohBsII6DDRHyEIQw9Ukf94482BNdmsiMtAf+v4za/mbw2 uZkbScBGXqeJ8fyQq//cNRz+mMaOZsXwd2pNXkXCgIlUhZAWL8E2wRBjwMCH3DFc fozfJCJ9EINKKHHMhfWricbOkOdSOtY= =jxhs -----END PGP SIGNATURE----- -------------- next part -------------- A non-text attachment was scrubbed... Name: lists.vcf Type: text/x-vcard Size: 310 bytes Desc: not available URL: From ubershmekel at gmail.com Fri Dec 16 10:04:05 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Fri, 16 Dec 2011 11:04:05 +0200 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: <4EEB0140.3020205@zopyx.com> References: <4EEB0140.3020205@zopyx.com> Message-ID: I get a massive ton of xmlrpclib timeouts recently as well. Even a few regular webpages are extremely slow eg: http://pypi.python.org/pypi?:action=browse&c=214 Yuval Greenfield On Fri, Dec 16, 2011 at 10:28 AM, Andreas Jung wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi there, > > we encounter lots of buildout problems (triggered by various > Jenkins instances running at different hosters) similar to: > > Restored 'README.txt' > Restored 'dev.cfg' > A products > A products/README.txt > Checked out revision 6884. > + bin/python bootstrap.py > Downloading > > http://pypi.python.org/packages/source/d/distribute/distribute-0.6.24.tar.gz > Traceback (most recent call last): > File "bootstrap.py", line 67, in > ez['use_setuptools'](to_dir=tmpeggs, download_delay=0, no_fake=True) > File "", line 145, in use_setuptools > File "", line 124, in _do_download > File "", line 193, in download_setuptools > File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 124, in urlopen > return _opener.open(url, data, timeout) > File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 389, in open > response = self._open(req, data) > File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 407, in _open > '_open', req) > File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 367, in > _call_chain > result = func(*args) > File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1146, in http_open > return self.do_open(httplib.HTTPConnection, req) > File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1121, in do_open > raise URLError(err) > urllib2.URLError: by protocol> > Recording test results > > Is there any severe problem with PyPI? These errors come and come. > Triggering a Jenkins manually after such a failure often results in a > successful build. > > Andreas > > - -- > ZOPYX Limited | zopyx group > Charlottenstr. 37/1 | The full-service network for Zope & Plone > D-72070 T?bingen | Produce & Publish > www.zopyx.com | www.produce-and-publish.com > - ------------------------------------------------------------------------ > E-Publishing, Python, Zope & Plone development, Consulting > > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.11 (Darwin) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iQGUBAEBAgAGBQJO6wE/AAoJEADcfz7u4AZjvdkLwIqA3dtNFig20L4sQPikEiXq > c1YNeZ8qaZTcGN3urvnVP9XXG8wGd8eebGP/Y/hwSzyvtLFxDgrx0XEmbIFgAjCL > 7t1tjiHScG/rN/F8ABg0CXBHqI9sjbzJ12ZmrivmzJLhlqE/BKsEQOuAQ6ThgwmZ > 2MqfxNkjWqndN8MIgvy7E1tQBH5McuEEDOdwe/uMXeXxy/jpJwiFEKqqBhMjEULN > PTxxF6RJGDJ3TiXK51dZkSJse8IcwRnm4A2mb828Ig3sfX4RTrvJ3Hl0kzyXqVl0 > eH28/10Dbi67GH3uWtZ21+8kQ1WD3hGvsqTxNAoBJoPF+/2tLSmdC6WsLpa6+q/v > vdjlNOf0mvzSM5g8EiJohBsII6DDRHyEIQw9Ukf94482BNdmsiMtAf+v4za/mbw2 > uZkbScBGXqeJ8fyQq//cNRz+mMaOZsXwd2pNXkXCgIlUhZAWL8E2wRBjwMCH3DFc > fozfJCJ9EINKKHHMhfWricbOkOdSOtY= > =jxhs > -----END PGP SIGNATURE----- > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From jim at zope.com Fri Dec 16 15:21:03 2011 From: jim at zope.com (Jim Fulton) Date: Fri, 16 Dec 2011 09:21:03 -0500 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: <4EEB0140.3020205@zopyx.com> References: <4EEB0140.3020205@zopyx.com> Message-ID: Are you using a mirror? Jim On Fri, Dec 16, 2011 at 3:28 AM, Andreas Jung wrote: > -----BEGIN PGP SIGNED MESSAGE----- > Hash: SHA1 > > Hi there, > > we encounter lots of buildout problems (triggered by various > Jenkins instances running at different hosters) similar to: > > Restored 'README.txt' > Restored 'dev.cfg' > A ? ?products > A ? ?products/README.txt > Checked out revision 6884. > + bin/python bootstrap.py > Downloading > http://pypi.python.org/packages/source/d/distribute/distribute-0.6.24.tar.gz > Traceback (most recent call last): > ?File "bootstrap.py", line 67, in > ? ez['use_setuptools'](to_dir=tmpeggs, download_delay=0, no_fake=True) > ?File "", line 145, in use_setuptools > ?File "", line 124, in _do_download > ?File "", line 193, in download_setuptools > ?File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 124, in urlopen > ? return _opener.open(url, data, timeout) > ?File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 389, in open > ? response = self._open(req, data) > ?File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 407, in _open > ? '_open', req) > ?File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 367, in _call_chain > ? result = func(*args) > ?File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1146, in http_open > ? return self.do_open(httplib.HTTPConnection, req) > ?File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1121, in do_open > ? raise URLError(err) > urllib2.URLError: by protocol> > Recording test results > > Is there any severe problem with PyPI? These errors come and come. > Triggering a Jenkins manually after such a failure often results in a > successful build. > > Andreas > > - -- > ZOPYX Limited ? ? ? ? ? | zopyx group > Charlottenstr. 37/1 ? ? | The full-service network for Zope & Plone > D-72070 T?bingen ? ? ? ?| Produce & Publish > www.zopyx.com ? ? ? ? ? | www.produce-and-publish.com > - ------------------------------------------------------------------------ > E-Publishing, Python, Zope & Plone development, Consulting > > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.11 (Darwin) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iQGUBAEBAgAGBQJO6wE/AAoJEADcfz7u4AZjvdkLwIqA3dtNFig20L4sQPikEiXq > c1YNeZ8qaZTcGN3urvnVP9XXG8wGd8eebGP/Y/hwSzyvtLFxDgrx0XEmbIFgAjCL > 7t1tjiHScG/rN/F8ABg0CXBHqI9sjbzJ12ZmrivmzJLhlqE/BKsEQOuAQ6ThgwmZ > 2MqfxNkjWqndN8MIgvy7E1tQBH5McuEEDOdwe/uMXeXxy/jpJwiFEKqqBhMjEULN > PTxxF6RJGDJ3TiXK51dZkSJse8IcwRnm4A2mb828Ig3sfX4RTrvJ3Hl0kzyXqVl0 > eH28/10Dbi67GH3uWtZ21+8kQ1WD3hGvsqTxNAoBJoPF+/2tLSmdC6WsLpa6+q/v > vdjlNOf0mvzSM5g8EiJohBsII6DDRHyEIQw9Ukf94482BNdmsiMtAf+v4za/mbw2 > uZkbScBGXqeJ8fyQq//cNRz+mMaOZsXwd2pNXkXCgIlUhZAWL8E2wRBjwMCH3DFc > fozfJCJ9EINKKHHMhfWricbOkOdSOtY= > =jxhs > -----END PGP SIGNATURE----- > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > -- Jim Fulton http://www.linkedin.com/in/jimfulton From lists at zopyx.com Fri Dec 16 16:06:06 2011 From: lists at zopyx.com (Andreas Jung) Date: Fri, 16 Dec 2011 16:06:06 +0100 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: References: <4EEB0140.3020205@zopyx.com> Message-ID: <4EEB5E5E.4090209@zopyx.com> -----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 No. Jim Fulton wrote: > Are you using a mirror? > > Jim > > On Fri, Dec 16, 2011 at 3:28 AM, Andreas Jung > wrote: Hi there, > > we encounter lots of buildout problems (triggered by various Jenkins > instances running at different hosters) similar to: > > Restored 'README.txt' Restored 'dev.cfg' A products A > products/README.txt Checked out revision 6884. + bin/python > bootstrap.py Downloading > http://pypi.python.org/packages/source/d/distribute/distribute-0.6.24.tar.gz > > Traceback (most recent call last): > File "bootstrap.py", line 67, in > ez['use_setuptools'](to_dir=tmpeggs, download_delay=0, no_fake=True) > File "", line 145, in use_setuptools File "", line > 124, in _do_download File "", line 193, in > download_setuptools File > "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 124, in urlopen > return _opener.open(url, data, timeout) File > "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 389, in open > response = self._open(req, data) File > "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 407, in _open > '_open', req) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line > 367, in _call_chain result = func(*args) File > "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1146, in > http_open return self.do_open(httplib.HTTPConnection, req) File > "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1121, in do_open > raise URLError(err) urllib2.URLError: Address family not supported by protocol> Recording test results > > Is there any severe problem with PyPI? These errors come and come. > Triggering a Jenkins manually after such a failure often results in > a successful build. > > Andreas > >> >> _______________________________________________ Catalog-SIG mailing >> list Catalog-SIG at python.org >> http://mail.python.org/mailman/listinfo/catalog-sig >> > > > - -- ZOPYX Limited | zopyx group Charlottenstr. 37/1 | The full-service network for Zope & Plone D-72070 T?bingen | Produce & Publish www.zopyx.com | www.produce-and-publish.com - ------------------------------------------------------------------------ E-Publishing, Python, Zope & Plone development, Consulting -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iQGUBAEBAgAGBQJO615eAAoJEADcfz7u4AZjwN8LwMoQjJE0vOSU8Fa2Bz+EgAp0 cE3HOY/yiesXkM9IycmFrxJYd96AHv7f4PE5vvABALmXAXHkATwVnF6PC38Kjycs 72XocNKyNK4hDxS0P+/LLW4FODzoeFdBZIVSlviJujw+eOX422cEyVsgPqRJoIVF eLIlSWHTR7z28NaWW3dwXjoJaPzgEfDGCEVKfaehKB3x09DZxrMH478ivEd6001R l0ib1Xm+fF9NwCSsyGQThTzrcWvIp0WIUZUhXIzhoqqe2txvxTbHpT26TXdp3N4x 880MJ3Ko2xiIXEFFkgNaP3ajYZhF0K4D5urIjx861NfGps49UGYEX35y30NCEFas rkmmDv4y4sZB12GtSnURHzDhzkqOGfqXAFV9bATZ1AWcG7gcoN4VjnwzFmWbHiDt ohiw/831PBMUbBet5kheCHm8oD0jLnvWOy0mrcslikvUX+3sIJhyu1YxRRcsF9jQ iUe8kUwzU16YDy2DoDXk3q36uhWOyPA= =Z/0v -----END PGP SIGNATURE----- -------------- next part -------------- A non-text attachment was scrubbed... Name: lists.vcf Type: text/x-vcard Size: 310 bytes Desc: not available URL: From chrism at plope.com Fri Dec 16 18:53:12 2011 From: chrism at plope.com (Chris McDonough) Date: Fri, 16 Dec 2011 12:53:12 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> Message-ID: <1324057992.2556.61.camel@thinko> On Fri, 2011-12-16 at 03:53 +0200, Yuval Greenfield wrote: > On Wed, Dec 14, 2011 at 5:36 AM, Chris McDonough > wrote: > On Tue, 2011-12-13 at 12:29 +0200, Yuval Greenfield wrote: > > > 1. If this forum likes it I'll add a lock icon next to > packages with > > the "Programming Language :: Python :: 2 :: Only" classifier > with the > > following mouse-over: "The maintainers of this code have > declared it > > will run on Python 2 only for the foreseeable future". This > might > > lessen the social pressure on the maintainers to port. I > don't think > > it'll help the PR issue. > > > For the record, at least one package is already using the new > "2-only" > classifier now: > > http://pypi.python.org/pypi/Pylons/1.0.1rc1 > > > > > > > I added both features (python2only and py3equivalents). It took a lot > longer than I expected and cost $0.10 of GAE quotas... > > > Please do critique: > > > http://python3wos.appspot.com/ Thanks for adding the feature. The way you've decided to do it, I guess folks can't really opt out of the "shame of being red" by using a python 2 only tag. Also, given the composition of packages on there, that chart will be all-red more or less forever; there's unlikely to be many newly charting python 3-compat packages any time soon, at least none that overtake the "will always be red" ones which have been around for 8 + years. Lame. - C From stefan-usenet at bytereef.org Fri Dec 16 20:13:30 2011 From: stefan-usenet at bytereef.org (Stefan Krah) Date: Fri, 16 Dec 2011 20:13:30 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <1324057992.2556.61.camel@thinko> References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> Message-ID: <20111216191330.GA32647@sleipnir.bytereef.org> [Not replying to Chris, I don't have the previous mail] Chris McDonough wrote: > > I added both features (python2only and py3equivalents). It took a lot > > longer than I expected and cost $0.10 of GAE quotas... > > > > > > Please do critique: > > > > > > http://python3wos.appspot.com/ I don't think this site is very useful in its current form. I immediately found two false positives: 1) ordereddict is part of: http://docs.python.org/dev/library/collections.html#collections.OrderedDict 2) pylint supports Python3 (from their README): "2to3 is integrated into the distutils installation process and will be run as a build step when invoked by the python3 interpreter: NO_SETUPTOOLS=1 python3 setup.py install --no-compile" Also, the site would look a lot greener if there were just two rows for all zope.* and plone.* packages. Stefan Krah From richard at python.org Fri Dec 16 23:20:02 2011 From: richard at python.org (Richard Jones) Date: Sat, 17 Dec 2011 09:20:02 +1100 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: <4EEB5E5E.4090209@zopyx.com> References: <4EEB0140.3020205@zopyx.com> <4EEB5E5E.4090209@zopyx.com> Message-ID: I'm curious. Why were the mirrors not your first step in attempting to address your automated buildout issues? Richard On 17 December 2011 02:06, Andreas Jung wrote: > No. > > Jim Fulton wrote: >> Are you using a mirror? >> >> Jim >> >> On Fri, Dec 16, 2011 at 3:28 AM, Andreas Jung >> wrote: Hi there, >> >> we encounter lots of buildout problems (triggered by various Jenkins >> instances running at different hosters) similar to: >> >> Restored 'README.txt' Restored 'dev.cfg' A ? ?products A >> products/README.txt Checked out revision 6884. + bin/python >> bootstrap.py Downloading >> http://pypi.python.org/packages/source/d/distribute/distribute-0.6.24.tar.gz >> >> > Traceback (most recent call last): >> File "bootstrap.py", line 67, in >> ez['use_setuptools'](to_dir=tmpeggs, download_delay=0, no_fake=True) >> File "", line 145, in use_setuptools File "", line >> 124, in _do_download File "", line 193, in >> download_setuptools File >> "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 124, in urlopen >> return _opener.open(url, data, timeout) File >> "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 389, in open >> response = self._open(req, data) File >> "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 407, in _open >> '_open', req) File "/opt/python/2.6.4/lib/python2.6/urllib2.py", line >> 367, in _call_chain result = func(*args) File >> "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1146, in >> http_open return self.do_open(httplib.HTTPConnection, req) File >> "/opt/python/2.6.4/lib/python2.6/urllib2.py", line 1121, in do_open >> raise URLError(err) urllib2.URLError: > Address family not supported by protocol> Recording test results >> >> Is there any severe problem with PyPI? These errors come and come. >> Triggering a Jenkins manually after such a failure often results in >> a successful build. >> >> Andreas >> >>> >>> _______________________________________________ Catalog-SIG mailing >>> list Catalog-SIG at python.org >>> http://mail.python.org/mailman/listinfo/catalog-sig >>> >> >> >> > > - -- > ZOPYX Limited ? ? ? ? ? | zopyx group > Charlottenstr. 37/1 ? ? | The full-service network for Zope & Plone > D-72070 T?bingen ? ? ? ?| Produce & Publish > www.zopyx.com ? ? ? ? ? | www.produce-and-publish.com > - ------------------------------------------------------------------------ > E-Publishing, Python, Zope & Plone development, Consulting > > > -----BEGIN PGP SIGNATURE----- > Version: GnuPG v1.4.11 (Darwin) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iQGUBAEBAgAGBQJO615eAAoJEADcfz7u4AZjwN8LwMoQjJE0vOSU8Fa2Bz+EgAp0 > cE3HOY/yiesXkM9IycmFrxJYd96AHv7f4PE5vvABALmXAXHkATwVnF6PC38Kjycs > 72XocNKyNK4hDxS0P+/LLW4FODzoeFdBZIVSlviJujw+eOX422cEyVsgPqRJoIVF > eLIlSWHTR7z28NaWW3dwXjoJaPzgEfDGCEVKfaehKB3x09DZxrMH478ivEd6001R > l0ib1Xm+fF9NwCSsyGQThTzrcWvIp0WIUZUhXIzhoqqe2txvxTbHpT26TXdp3N4x > 880MJ3Ko2xiIXEFFkgNaP3ajYZhF0K4D5urIjx861NfGps49UGYEX35y30NCEFas > rkmmDv4y4sZB12GtSnURHzDhzkqOGfqXAFV9bATZ1AWcG7gcoN4VjnwzFmWbHiDt > ohiw/831PBMUbBet5kheCHm8oD0jLnvWOy0mrcslikvUX+3sIJhyu1YxRRcsF9jQ > iUe8kUwzU16YDy2DoDXk3q36uhWOyPA= > =Z/0v > -----END PGP SIGNATURE----- > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > From tjreedy at udel.edu Fri Dec 16 23:53:49 2011 From: tjreedy at udel.edu (Terry Reedy) Date: Fri, 16 Dec 2011 17:53:49 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <20111216191330.GA32647@sleipnir.bytereef.org> References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> Message-ID: On 12/16/2011 2:13 PM, Stefan Krah wrote: >>> Please do critique: >>> http://python3wos.appspot.com/ > > I don't think this site is very useful in its current form. I think in its current form it is a net negative. From time to time, Python 2 package developers have shown up on pydev list and expressed their resentment at being 'pressured' to convert yesterday. It makes it hard to have a conversation. I always wondered what they were talking about. Core developers have mostly tried to avoid resentment-inducing pressure. At least now I know one source of the blowback, even if it is misdirected. -- Terry Jan Reedy From ubershmekel at gmail.com Sat Dec 17 00:01:14 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Sat, 17 Dec 2011 01:01:14 +0200 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <20111216191330.GA32647@sleipnir.bytereef.org> References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> Message-ID: On Fri, Dec 16, 2011 at 9:13 PM, Stefan Krah wrote: > [Not replying to Chris, I don't have the previous mail] > > Chris McDonough wrote: > > > I added both features (python2only and py3equivalents). It took a lot > > > longer than I expected and cost $0.10 of GAE quotas... > > > > > > > > > Please do critique: > > > > > > > > > http://python3wos.appspot.com/ > > I don't think this site is very useful in its current form. I immediately > found two false positives: > > 1) ordereddict is part of: > > http://docs.python.org/dev/library/collections.html#collections.OrderedDict > > > 2) pylint supports Python3 (from their README): > > "2to3 is integrated into the distutils installation process and will be > run as a > build step when invoked by the python3 interpreter: > > NO_SETUPTOOLS=1 python3 setup.py install --no-compile" > > > Also, the site would look a lot greener if there were just two rows for > all zope.* and plone.* packages. > > > Stefan Krah > > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > 1) Fixed. 2) This is the same issue as with docutils. I believe that either a maintainer of PyPI needs to add the trove classifier for them or someone should contact Logilab and ask them to add the trove classifier. 3) Why is it that plone and zope have 81 packages on PyPI? In the next top 100 there are 7 green packages (cssutils, argparse, pytest, pyramid, versiontools, PyYAML, scipy). So if I consolidate plone and zope we get to about 59/200 vs 52/200. Not a huge jump but a worthwhile one. Yuval Greenfield -------------- next part -------------- An HTML attachment was scrubbed... URL: From stefan-usenet at bytereef.org Sun Dec 18 00:37:35 2011 From: stefan-usenet at bytereef.org (Stefan Krah) Date: Sun, 18 Dec 2011 00:37:35 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> Message-ID: <20111217233735.GA11590@sleipnir.bytereef.org> Yuval Greenfield wrote: > On Fri, Dec 16, 2011 at 9:13 PM, Stefan Krah > 1) ordereddict is part of: > 2) pylint supports Python3 (from their README): > > 1) Fixed. > 2) This is the same issue as with docutils. I believe that either a maintainer > of PyPI needs to add the trove classifier for them or someone should contact > Logilab and ask them to add the trove classifier. The persons who are running PyPI should only edit individual pages in exceptional situations (illegal content etc.). I was under the impression that you want to to promote Python3 adoption. Why don't you contact Logilab and check the rest of the packages manually? > 3) Why is it that plone and zope have 81 packages on PyPI? I don't think it is the top priority of PyPI to facilitate generating automated statistics. Also, anyone can upload a package without any editorial process. > In the next top 100 > there are 7 green packages > (cssutils, argparse, pytest, pyramid, versiontools, PyYAML, scipy). So if I > consolidate plone and zope we get to about 59/200 vs 52/200. Not a huge jump > but a worthwhile one. I don't know enough about the organization of the zope.* hierarchy. Perhaps some people want to see that e.g. zope.interface is green. On the whole, I believe that users of zope and plone are already acutely aware if/when Python3 compatibility will happen. For the rest of Python users, it would be far more interesting to see a wider variety of packages. Regarding the lock icon on a red background: Authors who use the Python::2::Only tag want to be left alone. It is completely beyond me why someone who once made a package freely available should have a moral obligation to follow language changes forever. What are you going to do in 10 years? Leave the Pylons entry with a lock icon? How about maintaining a "wall of shame" for free packages written in K&R C or free packages written in FORTRAN 77? If an author has decided to use the classifier, he already accepts that his package might get less attention in the future. This isn't an easy decision, so please respect the intent of the classifier and remove such packages from your site. Stefan Krah From martin at v.loewis.de Sun Dec 18 20:37:44 2011 From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Sun, 18 Dec 2011 20:37:44 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <20111217233735.GA11590@sleipnir.bytereef.org> References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: <4EEE4108.7020808@v.loewis.de> > The persons who are running PyPI should only edit individual pages in > exceptional situations (illegal content etc.). Indeed. People frequently request that we change meta-data about packages (in particular broken URLs); we always point them to the package authors. Regards, Martin From fuzzyman at gmail.com Sun Dec 18 20:46:57 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Sun, 18 Dec 2011 19:46:57 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> Message-ID: On 16 December 2011 23:01, Yuval Greenfield wrote: > On Fri, Dec 16, 2011 at 9:13 PM, Stefan Krah wrote: > >> [Not replying to Chris, I don't have the previous mail] >> >> Chris McDonough wrote: >> > > I added both features (python2only and py3equivalents). It took a lot >> > > longer than I expected and cost $0.10 of GAE quotas... >> > > >> > > >> > > Please do critique: >> > > >> > > >> > > http://python3wos.appspot.com/ >> >> I don't think this site is very useful in its current form. I immediately >> found two false positives: >> >> 1) ordereddict is part of: >> >> >> http://docs.python.org/dev/library/collections.html#collections.OrderedDict >> >> >> 2) pylint supports Python3 (from their README): >> >> "2to3 is integrated into the distutils installation process and will be >> run as a >> build step when invoked by the python3 interpreter: >> >> NO_SETUPTOOLS=1 python3 setup.py install --no-compile" >> >> >> Also, the site would look a lot greener if there were just two rows for >> all zope.* and plone.* packages. >> >> >> Stefan Krah >> >> >> _______________________________________________ >> Catalog-SIG mailing list >> Catalog-SIG at python.org >> http://mail.python.org/mailman/listinfo/catalog-sig >> > > > > 1) Fixed. > 2) This is the same issue as with docutils. I believe that either a > maintainer of PyPI needs to add the trove classifier for them or someone > should contact Logilab and ask them to add the trove classifier. > I strongly believe that your website would be of much greater use to the Python community (instead of actively harmful) if packages that have Python 3 support show up in green. I'm very pleased (thank you) you've done this with packages like ordereddict and unittest2, but some (additional) manual curation for packages that do support Python 3 will complete the job. All the best, Michael Foord > 3) Why is it that plone and zope have 81 packages on PyPI? In the next top > 100 there are 7 green packages > (cssutils, argparse, pytest, pyramid, versiontools, PyYAML, scipy). So if I > consolidate plone and zope we get to about 59/200 vs 52/200. Not a huge > jump but a worthwhile one. > > > Yuval Greenfield > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From martin at v.loewis.de Sun Dec 18 23:06:00 2011 From: martin at v.loewis.de (=?ISO-8859-1?Q?=22Martin_v=2E_L=F6wis=22?=) Date: Sun, 18 Dec 2011 23:06:00 +0100 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: <4EEB0140.3020205@zopyx.com> References: <4EEB0140.3020205@zopyx.com> Message-ID: <4EEE63C8.8050007@v.loewis.de> > Is there any severe problem with PyPI? There were. The e1000 card kept complaining e1000: eth0: e1000_clean_tx_irq: Detected Tx Unit Hang under heavy load. Not sure whether that's a new problem, or just now occurring due to increased load; notably, the problems were much reduced over this weekend. People reported that this may be due to bugs in the Linux e1000 driver, but none of the suggested changes (most recent driver, various module parameters) had any positive effect. I have now moved PyPI to a different machine (the same as www.python.org), and see whether I can get the network card (or perhaps the cabling) replaced. www.python.org is a quite larger machine, so it should be able to cope with the extra load. Regards, Martin From ubershmekel at gmail.com Mon Dec 19 00:09:32 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Mon, 19 Dec 2011 01:09:32 +0200 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <20111217233735.GA11590@sleipnir.bytereef.org> References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: On Sun, Dec 18, 2011 at 1:37 AM, Stefan Krah wrote: > Yuval Greenfield wrote: > > On Fri, Dec 16, 2011 at 9:13 PM, Stefan Krah > > > 1) ordereddict is part of: > > 2) pylint supports Python3 (from their README): > > > > 1) Fixed. > > 2) This is the same issue as with docutils. I believe that either a > maintainer > > of PyPI needs to add the trove classifier for them or someone should > contact > > Logilab and ask them to add the trove classifier. > > The persons who are running PyPI should only edit individual pages in > exceptional situations (illegal content etc.). > > > I was under the impression that you want to to promote Python3 adoption. > Why don't you contact Logilab and check the rest of the packages manually? > > > I'd already contacted several maintainers of packages with such a request. Starting this project I assumed PyPI would be a reliable source for this information, so I hope you can understand my mindset. I guess the WOS is also good for motivating package owners to fix their classifiers. On that note, I think PyPI should require python modules to declare which version they're compatible with or tested on. > If an author has decided to use the classifier, he already accepts > that his package might get less attention in the future. This isn't > an easy decision, so please respect the intent of the classifier and > remove such packages from your site. > > > The WOS's main goal is to measure Python3 adoption in some non-trivial fashion. Total downloads isn't a perfect comparison as many downloads happened in the 8 years Python2 had a head start. Still, I don't know of a better metric. Removing packages that are a gigantic part of the python community would completely defeat the purpose of the site. If/when Richard implements the statistics call to PyPI, I might be able to implement more metrics. On Sun, Dec 18, 2011 at 9:46 PM, Michael Foord wrote: ... > some (additional) manual curation for packages that do support Python 3 > will complete the job. > > Fine, fine, fine.... I added these cute little red triangles to the newly green docutils and pylint. Are there others I should know of? Yuval Greenfield ps, some of you might prefer this limited edition better: http://python3wos.appspot.com/?pygmalion=1 -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrism at plope.com Mon Dec 19 05:58:56 2011 From: chrism at plope.com (Chris McDonough) Date: Sun, 18 Dec 2011 23:58:56 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: <1324270736.2556.157.camel@thinko> On Mon, 2011-12-19 at 01:09 +0200, Yuval Greenfield wrote: > > Fine, fine, fine.... I added these cute little red triangles to the > newly green docutils and pylint. Are there others I should know of? zc.buildout and zc.recipe.egg are others, FWIW. Thanks, - C From g.brandl at gmx.net Mon Dec 19 09:14:53 2011 From: g.brandl at gmx.net (Georg Brandl) Date: Mon, 19 Dec 2011 09:14:53 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: Am 19.12.2011 00:09, schrieb Yuval Greenfield: > If an author has decided to use the classifier, he already accepts > that his package might get less attention in the future. This isn't > an easy decision, so please respect the intent of the classifier and > remove such packages from your site. > > > > The WOS's main goal is to measure Python3 adoption in some non-trivial fashion. Then perhaps you could give it a non-trivial name? Georg From jannis at leidel.info Mon Dec 19 11:26:46 2011 From: jannis at leidel.info (Jannis Leidel) Date: Mon, 19 Dec 2011 11:26:46 +0100 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: <4EEE63C8.8050007@v.loewis.de> References: <4EEB0140.3020205@zopyx.com> <4EEE63C8.8050007@v.loewis.de> Message-ID: On 18.12.2011, at 23:06, Martin v. L?wis wrote: >> Is there any severe problem with PyPI? > > There were. The e1000 card kept complaining > > e1000: eth0: e1000_clean_tx_irq: Detected Tx Unit Hang > > under heavy load. Not sure whether that's a new problem, > or just now occurring due to increased load; notably, > the problems were much reduced over this weekend. > > People reported that this may be due to bugs in the Linux > e1000 driver, but none of the suggested changes (most recent > driver, various module parameters) had any positive effect. > > I have now moved PyPI to a different machine (the same > as www.python.org), and see whether I can get the network > card (or perhaps the cabling) replaced. www.python.org is > a quite larger machine, so it should be able to cope with > the extra load. Ah, I wasn't able to SSH into this new box, can you transfer the accounts from ximinez to www? Jannis From fuzzyman at gmail.com Mon Dec 19 12:05:10 2011 From: fuzzyman at gmail.com (Michael Foord) Date: Mon, 19 Dec 2011 11:05:10 +0000 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: On 18 December 2011 23:09, Yuval Greenfield wrote: > On Sun, Dec 18, 2011 at 1:37 AM, Stefan Krah wrote: > >> Yuval Greenfield wrote: >> > On Fri, Dec 16, 2011 at 9:13 PM, Stefan Krah < >> stefan-usenet at bytereef.org> >> > 1) ordereddict is part of: >> > 2) pylint supports Python3 (from their README): >> > >> > 1) Fixed. >> > 2) This is the same issue as with docutils. I believe that either a >> maintainer >> > of PyPI needs to add the trove classifier for them or someone should >> contact >> > Logilab and ask them to add the trove classifier. >> >> The persons who are running PyPI should only edit individual pages in >> exceptional situations (illegal content etc.). >> >> >> I was under the impression that you want to to promote Python3 adoption. >> Why don't you contact Logilab and check the rest of the packages manually? >> >> >> > > I'd already contacted several maintainers of packages with such a request. > Starting this project I assumed PyPI would be a reliable source for this > information, so I hope you can understand my mindset. I guess the WOS is > also good for motivating package owners to fix their classifiers. > > On that note, I think PyPI should require python modules to declare which > version they're compatible with or tested on. > > The use of trove classifiers at all isn't compulsory on PyPI, so although I agree it would be *preferable* for all authors to use them correctly I don't think PyPI is going to require it any time soon. > >> If an author has decided to use the classifier, he already accepts >> that his package might get less attention in the future. This isn't >> an easy decision, so please respect the intent of the classifier and >> remove such packages from your site. >> >> >> > The WOS's main goal is to measure Python3 adoption in some non-trivial > fashion. Total downloads isn't a perfect comparison as many downloads > happened in the 8 years Python2 had a head start. Still, I don't know of a > better metric. Removing packages that are a gigantic part of the python > community would completely defeat the purpose of the site. > > If/when Richard implements the statistics call to PyPI, I might be able to > implement more metrics. > > > On Sun, Dec 18, 2011 at 9:46 PM, Michael Foord wrote: > ... > > some (additional) manual curation for packages that do support Python 3 >> will complete the job. >> >> > Fine, fine, fine.... I added these cute little red triangles to the newly > green docutils and pylint. Are there others I should know of? > Thank you, a great improvement. All the best, Michael Foord > > > Yuval Greenfield > > > ps, some of you might prefer this limited edition better: > http://python3wos.appspot.com/?pygmalion=1 > > _______________________________________________ > Catalog-SIG mailing list > Catalog-SIG at python.org > http://mail.python.org/mailman/listinfo/catalog-sig > > -- http://www.voidspace.org.uk/ May you do good and not evil May you find forgiveness for yourself and forgive others May you share freely, never taking more than you give. -- the sqlite blessing http://www.sqlite.org/different.html -------------- next part -------------- An HTML attachment was scrubbed... URL: From ubershmekel at gmail.com Mon Dec 19 12:49:07 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Mon, 19 Dec 2011 13:49:07 +0200 Subject: [Catalog-sig] Massive problems with PyPI since days In-Reply-To: <4EEE63C8.8050007@v.loewis.de> References: <4EEB0140.3020205@zopyx.com> <4EEE63C8.8050007@v.loewis.de> Message-ID: On Mon, Dec 19, 2011 at 12:06 AM, "Martin v. L?wis" wrote: > > Is there any severe problem with PyPI? > > There were. The e1000 card kept complaining > > e1000: eth0: e1000_clean_tx_irq: Detected Tx Unit Hang > > under heavy load. Not sure whether that's a new problem, > or just now occurring due to increased load; notably, > the problems were much reduced over this weekend. > > People reported that this may be due to bugs in the Linux > e1000 driver, but none of the suggested changes (most recent > driver, various module parameters) had any positive effect. > > I have now moved PyPI to a different machine (the same > as www.python.org), and see whether I can get the network > card (or perhaps the cabling) replaced. www.python.org is > a quite larger machine, so it should be able to cope with > the extra load. > > My logs show that between 2011-12-19 10:49:15 and 2011-12-19 00:58:17 (UTC) were the last PyPI errors. So it seems fixed. Thank you, Yuval Greenfield -------------- next part -------------- An HTML attachment was scrubbed... URL: From ubershmekel at gmail.com Mon Dec 19 13:35:03 2011 From: ubershmekel at gmail.com (Yuval Greenfield) Date: Mon, 19 Dec 2011 14:35:03 +0200 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: On Mon, Dec 19, 2011 at 1:05 PM, Michael Foord wrote: > > The use of trove classifiers at all isn't compulsory on PyPI, so although > I agree it would be *preferable* for all authors to use them correctly I > don't think PyPI is going to require it any time soon. > > > Are we happy that it isn't compulsory? I mean, there are a lot of compulsory lines in setup.py/cfg, why not require a compatibility tag? >From a backwards compatibility standpoint, I guess a warning could be emitted for a year or 2 until the compatibility classifiers are actually mandatory. I'm sure I'm not the only one who's struggled with eg Scapy working only on 2.6 and not 2.7, which is only a minor occurrence compared to the great py3k divide. Yuval Greenfield -------------- next part -------------- An HTML attachment was scrubbed... URL: From martin at v.loewis.de Mon Dec 19 17:41:54 2011 From: martin at v.loewis.de (martin at v.loewis.de) Date: Mon, 19 Dec 2011 17:41:54 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: <20111219174154.Horde.o2PrDlNNcXdO72lSVdamEMA@webmail.df.eu> >> The use of trove classifiers at all isn't compulsory on PyPI, so although >> I agree it would be *preferable* for all authors to use them correctly I >> don't think PyPI is going to require it any time soon. >> >> >> > Are we happy that it isn't compulsory? I mean, there are a lot of > compulsory lines in setup.py/cfg, why not require a compatibility tag? Packages on PyPI don't even need to use setup.py at all, and may use autoconf or plain make instead. There is always a conflict of interest wrt. PyPI between package authors and package users. Package users often demand that package authors do something, and then ask that PyPI enforces this. Package authors then explain reasons why they don't want to do that, and threaten to host their packages elsewhere. In the specific case, I don't think that requiring version classifiers would be useful. First, tools (like setuptools) might start filling out "default" values for these, so that users don't have to modify their setuo.pys. Furthermore, proper declaration wouldn't have helped you either: if the 2.7 classifier is missing, is it because the software doesn't work on 2.7, or is it that the author hasn't added 2.7 to the list? Regards, Martin From stefan-usenet at bytereef.org Mon Dec 19 18:56:36 2011 From: stefan-usenet at bytereef.org (Stefan Krah) Date: Mon, 19 Dec 2011 18:56:36 +0100 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> Message-ID: <20111219175636.GA23419@sleipnir.bytereef.org> Yuval Greenfield wrote: > I'd already contacted several maintainers of packages with such a request. > Starting this project I assumed PyPI would be a reliable source for this > information, so I hope you can understand my mindset. No, unfortunately PyPI isn't so reliable in that respect. Short of downloading and checking each package manually, the next best option is to check: http://code.activestate.com/pypm/ Here we go: http://code.activestate.com/pypm/pylint/ -> green http://code.activestate.com/pypm/logilab-common/ -> green http://code.activestate.com/pypm/logilab-astng/ -> green http://code.activestate.com/pypm/paste/ -> green http://code.activestate.com/pypm/pycrypto/ -> green http://code.activestate.com/pypm/python-dateutil/ -> green http://code.activestate.com/pypm/zconfig/ -> green http://code.activestate.com/pypm/zdaemon/ -> green http://code.activestate.com/pypm/pyopenssl/ -> green http://code.activestate.com/pypm/ordereddict/ -> green [ordereddict is still red on your site] http://code.activestate.com/pypm/yolk/ -> green That was tough. Stefan Krah From chrism at plope.com Mon Dec 19 20:14:11 2011 From: chrism at plope.com (Chris McDonough) Date: Mon, 19 Dec 2011 14:14:11 -0500 Subject: [Catalog-sig] "python 2 only" classifier In-Reply-To: <20111219175636.GA23419@sleipnir.bytereef.org> References: <1323833780.3502.42.camel@thinko> <1324057992.2556.61.camel@thinko> <20111216191330.GA32647@sleipnir.bytereef.org> <20111217233735.GA11590@sleipnir.bytereef.org> <20111219175636.GA23419@sleipnir.bytereef.org> Message-ID: <1324322051.2556.161.camel@thinko> On Mon, 2011-12-19 at 18:56 +0100, Stefan Krah wrote: > Yuval Greenfield wrote: > > I'd already contacted several maintainers of packages with such a request. > > Starting this project I assumed PyPI would be a reliable source for this > > information, so I hope you can understand my mindset. > > No, unfortunately PyPI isn't so reliable in that respect. Short of > downloading and checking each package manually, the next best option > is to check: > > http://code.activestate.com/pypm/ > > Here we go: > > http://code.activestate.com/pypm/pylint/ -> green > http://code.activestate.com/pypm/logilab-common/ -> green > http://code.activestate.com/pypm/logilab-astng/ -> green > > http://code.activestate.com/pypm/paste/ -> green > http://code.activestate.com/pypm/pycrypto/ -> green > http://code.activestate.com/pypm/python-dateutil/ -> green > http://code.activestate.com/pypm/zconfig/ -> green > http://code.activestate.com/pypm/zdaemon/ -> green > http://code.activestate.com/pypm/pyopenssl/ -> green > > http://code.activestate.com/pypm/ordereddict/ -> green > [ordereddict is still red on your site] > > http://code.activestate.com/pypm/yolk/ -> green I can speak from experience that paste is not 3.X compatible. I'm not sure why PyPM would list it compatible. Maybe it installs, but it doesn't work. - C From ash at newthink.net Fri Dec 23 18:13:16 2011 From: ash at newthink.net (Ash Christopher) Date: Fri, 23 Dec 2011 10:13:16 -0700 Subject: [Catalog-sig] Python Packages on PyPi no longer attributed to my account. Message-ID: <9A32E7C6-782C-4656-A1A0-6583C8FCFD45@gmail.com> Hi catalog-sig, Noticed an issue where one of my packages isn't attributed to my PyPi account. http://pypi.python.org/pypi/django-clean/ <--- this is attached to my account http://pypi.python.org/pypi/django-concurrent-server/ <--- this is not attached I am trying to submit an update to django-concurrent-server but obviously since it isn't on my account it is rejected. Is there any way to rectify this? Ash From martin at v.loewis.de Mon Dec 26 14:30:19 2011 From: martin at v.loewis.de (martin at v.loewis.de) Date: Mon, 26 Dec 2011 14:30:19 +0100 Subject: [Catalog-sig] Python Packages on PyPi no longer attributed to my account. In-Reply-To: <9A32E7C6-782C-4656-A1A0-6583C8FCFD45@gmail.com> References: <9A32E7C6-782C-4656-A1A0-6583C8FCFD45@gmail.com> Message-ID: <20111226143019.Horde.LgO0ftjz9kRO_Hbrwvd0HCA@webmail.df.eu> Zitat von Ash Christopher : > Hi catalog-sig, > > Noticed an issue where one of my packages isn't attributed to my > PyPi account. > > http://pypi.python.org/pypi/django-clean/ <--- this is attached > to my account > http://pypi.python.org/pypi/django-concurrent-server/ <--- this > is not attached > > I am trying to submit an update to django-concurrent-server but > obviously since it isn't on my account it is rejected. > > Is there any way to rectify this? I'm not sure why you say "no longer". AFAICT, django-concurrent-server was created and uploaded by user 'bancek'; you never had a role in this. To fix this, you need to talk to 'bancek' and ask him to adjust the package ownership. Regards, Martin From songofacandy at gmail.com Tue Dec 27 13:05:34 2011 From: songofacandy at gmail.com (INADA Naoki) Date: Tue, 27 Dec 2011 21:05:34 +0900 Subject: [Catalog-sig] Can't upload MSI file via web form. Message-ID: I want to upload binary distribution for my package in MSI format. But PyPI shows:: """ Error processing form invalid distribution file """ What's wrong? -- INADA Naoki? From martin at v.loewis.de Wed Dec 28 10:41:03 2011 From: martin at v.loewis.de (martin at v.loewis.de) Date: Wed, 28 Dec 2011 10:41:03 +0100 Subject: [Catalog-sig] Can't upload MSI file via web form. In-Reply-To: References: Message-ID: <20111228104103.Horde.zD2-C1NNcXdO_uQvyQbDhwA@webmail.df.eu> Zitat von INADA Naoki : > I want to upload binary distribution for my package in MSI format. > But PyPI shows:: > > """ > Error processing form > > invalid distribution file > """ > > What's wrong? You didn't say what exactly you did, so it's difficult to say what's wrong. Most likely, you didn't specify "MS Windows MSI Installer" for "File Type". Alternatively, the file may not be an MSI file, or there may be a bug in PyPI. Regards, Martin From walterwefft at gmail.com Fri Dec 30 13:55:01 2011 From: walterwefft at gmail.com (gflanagan) Date: Fri, 30 Dec 2011 12:55:01 +0000 Subject: [Catalog-sig] pypi not accessible 30th Dec 2011 12:51 GMT Message-ID: http://www.downforeveryoneorjustme.com/http://pypi.python.org/ From tjreedy at udel.edu Fri Dec 30 23:42:39 2011 From: tjreedy at udel.edu (Terry Reedy) Date: Fri, 30 Dec 2011 17:42:39 -0500 Subject: [Catalog-sig] pypi not accessible 30th Dec 2011 12:51 GMT In-Reply-To: References: Message-ID: On 12/30/2011 7:55 AM, gflanagan wrote: > > > http://www.downforeveryoneorjustme.com/http://pypi.python.org/ Back up now. Thanks whoever worked on today's problems. -- Terry Jan Reedy From chris at simplistix.co.uk Sat Dec 31 12:29:44 2011 From: chris at simplistix.co.uk (Chris Withers) Date: Sat, 31 Dec 2011 11:29:44 +0000 Subject: [Catalog-sig] bunch of spam packages Message-ID: <4EFEF228.8040907@simplistix.co.uk> http://pypi.python.org/pypi/girlfriends/1.0 http://pypi.python.org/pypi/house/0.9 http://pypi.python.org/pypi/hardwork/0.9 http://pypi.python.org/pypi/car/0.9 ...all spamvertising the same website. While we're at it, can we get rid of the stupid nested list printers again: http://pypi.python.org/pypi?%3Aaction=search&term=nested+lists&submit=search cheers, Chris -- Simplistix - Content Management, Batch Processing & Python Consulting - http://www.simplistix.co.uk From tjreedy at udel.edu Sat Dec 31 23:14:28 2011 From: tjreedy at udel.edu (Terry Reedy) Date: Sat, 31 Dec 2011 17:14:28 -0500 Subject: [Catalog-sig] bunch of spam packages In-Reply-To: <4EFEF228.8040907@simplistix.co.uk> References: <4EFEF228.8040907@simplistix.co.uk> Message-ID: On 12/31/2011 6:29 AM, Chris Withers wrote: > http://pypi.python.org/pypi/girlfriends/1.0 > http://pypi.python.org/pypi/house/0.9 > http://pypi.python.org/pypi/hardwork/0.9 > http://pypi.python.org/pypi/car/0.9 > > ...all spamvertising the same website. In chinese, making them really stupid. The account should be deleted. > While we're at it, can we get rid of the stupid nested list printers again: > > http://pypi.python.org/pypi?%3Aaction=search&term=nested+lists&submit=search There are about 150 duplicate 'a simple printer of nested lists' listings. Perhaps pypi needs to be first-package-registration moderated. -- Terry Jan Reedy