Log file Viewer

[00:29:33] <StevenK> lifeless: Prod to chime in on https://bitbucket.org/pypa/setuptools/pull-requests/164/first-shot-at-removing-usage-of-_markerlib re platform_python_implementation

[01:46:31] <lifeless> StevenK: done

[01:47:12] <StevenK> lifeless: Thanks. Did you want to update PEP 508, or shall I do so after I go and buy lunch things?

[01:47:35] <lifeless> I think you should, no need for a handoff

[02:10:08] <pmxbot> jaraco pushed 4 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[02:10:08] <pmxbot> Move six to pkg_resources for use there.

[02:10:08] <pmxbot> Use six in pkg_resources.

[02:10:08] <pmxbot> Merge feature/issue-229

[02:10:08] <pmxbot> Added tag 19.3b1 for changeset 66fa131a0d77

[04:36:23] <StevenK> lifeless: I'm adding platform_python_implementation, or using it to replace python_implementation?

[05:05:19] <lifeless> StevenK: just s/python_implementation/platform_\1/

[05:05:54] <StevenK> Right, it's a replacement

[05:08:10] <StevenK> lifeless: That's a little greedy, it will replace platform.python_implementation calls :-P

[05:08:22] <lifeless> :)

[05:20:08] <StevenK> lifeless: https://github.com/pypa/interoperability-peps/pull/62 should be what we're after.

[13:37:30] <JayFK> is it normal that the warehouse container running gulp watch is on 100% CPU all the time?

[13:38:08] <JayFK> i suspect that's a side effect of vboxfs, but you never know :)

[13:42:18] <dstufft> JayFK: um, never noticed

[14:55:22] <ronny> xafer: thanks for the investigation :)

[14:58:23] <xafer> np :) I'd rather have pip stick with latest pytest

[14:58:43] <xafer> so long it keeps 2.6 compatibility :-P

[15:08:20] <ronny> xafer: when is python2.6 support drop sheduled for pip?

[15:09:24] <ronny> eh dstufft ?

[15:12:39] <xafer> no idea, but pip 8 will start complaining: autoslave/core/damages/scripts/bluecrash_exporter

[15:12:46] <xafer> erf

[15:13:06] <ronny> that was not the link you are looking for

[15:13:08] <xafer> I meant https://github.com/pypa/pip/commit/41cf790d7a8db6a0778799e93378c8ea1390e788#diff-b670e3b192038c9ffe810c1a12c0c51fR184

[15:13:17] <ronny> *jedi sheningans*

[15:14:44] <xafer> so maybe pip 9 or 10 will drop python 2.6 :)

[15:22:26] <ronny> xafer: do you know where th eol#s for older pythons are listed?

[15:22:39] <xafer> in PEPs

[15:22:41] <ronny> (i want to make pytest warn on eol#d and known to be broken versions

[15:23:14] <xafer> Search "PEP pythonX.X release schedule"

[15:23:42] <xafer> Don't know if they are centralized somewhere

[15:25:44] <ronny> found it, we should also warn for 3.0, 3.1, and starting february about 3.2

[15:40:47] <dstufft> ronny: xafer FTR there's no set date or version # for dropping 2.6

[15:41:18] <dstufft> we added a pending deprecation warning (but a special one that doesn't follow the version based deprecation schedule) to try and push people off 2.6 if possible

[15:41:29] <dstufft> but actually dropping it will be based on usage numbers

[15:45:52] <ronny> dstufft: py.test will follow on that - and rasie pytest-warnings for eol python version

[15:46:16] <ronny> dstufft: it might be nivce to get the same for all eol#d python version

[15:46:22] <xafer> we could also say pip8 will be the last version for 2.6 and keep maintaining it without major upgrade

[15:46:47] <dstufft> well, one thing I'd like to do next time I run the numbers, is do a pip version breakdown by Python version

[15:47:10] <dstufft> it's possible that a signifcant number of 2.6 users are also not upgrading their pip

[15:47:21] <xafer> but I'm not sure maintaining two pip branches would lower the work

[15:47:53] <dstufft> if the bulk of the 2.6 users are using, say pip < 6, then there probably isn't a big gain to continue supporting 2.6 in pip 8+

[17:04:28] <xafer> looking forward for those numbers :)

[17:08:34] <ronny> dstufft: for when is pipe 8 sheduled

[18:13:48] <umeshksingla_> Hello guys, do we only these commands for pip:

[18:14:32] <umeshksingla_> InstallCommand, DownloadCommand, UninstallCommand, FreezeCommand, ListCommand, ShowCommand, SearchCommand, WheelCommand, HashCommand, HelpCommand ?

[18:26:31] <dstufft> umeshksingla_: https://github.com/pypa/pip/blob/develop/pip/commands/__init__.py#L6-L16

[18:26:46] <umeshksingla_> thanks

[18:27:08] <umeshksingla_> i had doubt that pip only has 10-11 commands

[18:27:24] <umeshksingla_> anyways, i found out an issue on github

[18:27:34] <umeshksingla_> in pypa/pip repo #1737

[18:27:45] <umeshksingla_> https://github.com/pypa/pip/issues/1737

[18:30:47] <dstufft> nlh: ping

[19:13:04] <tchaypo> umeshksingla_: if I'm reading it right, https://pypi.python.org/pypi/editdistance "implements Levenshtein distance with C++ and Cython." which probably also makes it unsuitable for pip

[19:14:39] <umeshksingla_> okay.. sorry, my bad. i didn’t notice.

[19:14:59] <umeshksingla_> so do i need to implement Levenshtein myself?

[19:22:58] <tchaypo> I don't think that's necessarily the case; if you can find a pure-python implementation it could possibly be vendored in (see https://github.com/pypa/pip/tree/develop/pip/_vendor for some of the stuff that's already being pulled in this way)

[19:24:24] <tchaypo> but the big caveat to that is that the developers generally want to avoid pip getting too big and having to spend time updating the vendored packages, so things don't get added there unless they're really high value. So probably the first thing to do is look for what pure-python implementations are available and find out if they'd be included

[19:25:19] <tchaypo> I'm not a pip developer just a very occasional contributor, so take my advice with a grain of salt

[19:27:48] <umeshksingla_> Okayy..I’ll have a look what all is available already in pure python.

[19:27:56] <umeshksingla_> thanks for the advice though

[19:36:17] <umeshksingla_> @tchaypo, i think this is the one: https://pypi.python.org/pypi/pylev

[20:16:36] <nlh> dstufft: pong (if you're still around!)

[20:19:25] <nlh> dstufft: about to watch a film - email me and I will take a look tomorrow AM :D