pmxbot IRC Log Viewer

[01:06:34] <pmxbot> jaraco pushed 2 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[01:06:34] <pmxbot> Remove unused import

[01:06:34] <pmxbot> Use pytest to capture exception

[04:02:42] <pmxbot> jaraco pushed 3 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[04:02:42] <pmxbot> Backed out changeset: 40cc1fbecb1c

[04:02:42] <pmxbot> Move changelog entry

[04:02:42] <pmxbot> Catch, save, and restore any exceptions across the save_modules context. This corrects the latter of the two test failures. Ref #315.

[04:43:11] <pmxbot> jaraco pushed 2 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[04:43:11] <pmxbot> Make sure to monkey-patch the easy_install module in the setup context. Fixes the other former test failure. Ref #315.

[04:43:11] <pmxbot> Reindent

[14:53:46] <pmxbot> jaraco pushed 4 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[14:53:46] <pmxbot> Indent script for clarity

[14:53:46] <pmxbot> Extract variable for clarity of reading

[14:53:46] <pmxbot> Rewrite test to remove Windows-specific handling and instead capture the underlying expectation.

[14:53:46] <pmxbot> Rewrite file operations using context managers.

[15:24:04] <pmxbot> jaraco pushed 7 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[15:24:04] <pmxbot> Add compatibility shim for tarfile.open

[15:24:04] <pmxbot> Include __enter__ in shim

[15:24:04] <pmxbot> Try without self, as methods are set on the instance. Also include parameters to __exit__.

[15:24:04] <pmxbot> Remove spurious reference to self. Remove debugging code.

[15:24:04] <pmxbot> Extract common operation of dedent and left strip

[15:24:04] <pmxbot> Extract script variable for clarity

[15:24:04] <pmxbot> Reindent

[15:41:29] <pmxbot> jaraco pushed 6 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[15:41:29] <pmxbot> reset_setup_stop_context is apparently no longer needed.

[15:41:29] <pmxbot> Remove reset_setup_stop_context and reindent.

[15:41:29] <pmxbot> Remove unused import

[15:41:29] <pmxbot> Reindent for clarity

[15:41:29] <pmxbot> Slice the iterable rather than converting to a list and slicing that.

[15:41:29] <pmxbot> Please don't feign a monkey patch when one isn't used.

[16:02:25] <pmxbot> jaraco pushed 4 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[16:02:25] <pmxbot> Merge with master

[16:02:25] <pmxbot> Merge with remove-svn

[16:02:25] <pmxbot> Update changelog

[16:02:25] <pmxbot> Bumped to 10.0 in preparation for next release.

[16:07:33] <pmxbot> jaraco pushed 2 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[16:07:33] <pmxbot> Added tag 10.0 for changeset 0c4d18a747a6

[16:07:33] <pmxbot> Bumped to 10.1 in preparation for next release.

[16:23:19] <xafer> any reason why pip uses distlib.markers.interpret instead of markerlib like setuptools does ?

[17:03:09] <pmxbot> jaraco pushed 5 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:03:09] <pmxbot> Add change history to docs

[17:03:09] <pmxbot> Remove changes file from long_description.

[17:03:09] <pmxbot> Moved linkify logic to documentation builder as Sphinx extension.

[17:03:09] <pmxbot> Resave with excess whitespace removed

[17:03:09] <pmxbot> Add link to change history from readme.

[17:37:07] <pmxbot> jaraco pushed 1 commit to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:37:07] <pmxbot> Add test capturing failure. Ref #319.

[17:39:38] <pmxbot> jaraco pushed 2 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:39:38] <pmxbot> Ensure setuptools is present in the environment before invoking setup.py from easy_install. Fixes #319.

[17:39:38] <pmxbot> Update changelog

[17:42:28] <pmxbot> jaraco pushed 3 commits to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:42:28] <pmxbot> Bumped to 10.0.1 in preparation for next release.

[17:42:28] <pmxbot> Added tag 10.0.1 for changeset 4c41e2cdd70b

[17:42:28] <pmxbot> Bumped to 10.0.2 in preparation for next release.

[17:45:08] <pmxbot> jaraco pushed 1 commit to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:45:08] <pmxbot> Remove lingering reference to linked changelog.

[17:52:20] <pmxbot> jaraco pushed 1 commit to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:52:20] <pmxbot> Update copyright

[17:59:16] <pmxbot> jaraco pushed 1 commit to setuptools (https://bitbucket.org/pypa/setuptools/) :

[17:59:16] <pmxbot> Remove support for setup.py test. Use the recommended test runner (setup.py ptr).

[18:27:34] <ionelmc> jaraco: that's odd, setuptools' pypi page is missing the changelog now

[18:28:54] <sigmavirus24> ionelmc: the one that used to go on forever?

[18:29:33] <ionelmc> sigmavirus24: a testament to the effort poured in, yes

[18:29:50] <sigmavirus24> ionelmc: it's for the better

[18:31:25] <sigmavirus24> ionelmc: there were an unreasonable number of links on that page which was causing severe problems for installers that would spider links looking for packages to install

[19:06:52] <jaraco> ionelmc, I did that on purpose. Replaced it with a link at the top.

[19:07:09] <jaraco> To the same changelog, but in the docs.

[19:07:18] <jaraco> https://pythonhosted.org/setuptools/history.html

[19:10:16] <dstufft> jaraco: probably makes sense heh, the setuptools changelog was getting super long

[19:11:50] <dstufft> jaraco: is setuptools 10.0.dev0 supposed to be registered with PyPI?

[19:12:03] <dstufft> 10.1.dev0 I mean

[19:16:10] <jaraco> dstufft, hmm. Not really. That probably happened when I "registered" that version to refresh the readme.

[19:16:15] <jaraco> I'll remove it.

[19:29:53] <jaraco> dstufft: it's gone. Thanks for letting me know.

[19:34:11] <dstufft> jaraco: no problem! I only noticed cuz I saw it on https://warehouse.python.org/project/setuptools/ ;D

[19:34:23] <dstufft> probably warehouse shouldn't put the latest version up if the latest version is a pre-release

[19:35:37] <dstufft> jaraco: I find it funny that setuptools is now released more than any other project I know of compared to how in the past it was basically never released

[19:38:25] <dstufft> qwcode: ping

[19:42:23] <xafer> btw since you're discussing setuptools, are you aware that https://setuptools.pypa.io/en/latest/ points to version 5.4.3 ?

[19:53:23] <qwcode> dstufft, hey

[19:54:39] <qwcode> xafer, hah, the url for setuptools RTD project is https://bitbucket.org/ericholscher/setuptools

[19:55:02] <dstufft> ugh I forgot that I was going to try and get setuptools docs itself on RTD and not a fork

[19:55:28] <dstufft> qwcode: in pip 6 we added this https://github.com/pypa/pip/commit/6dc7f5ac3592d70c6d549c37e3f29bedc4e56798 which installs the dependencies in reverse order, which means that if foo depends on bar, bar is installed first instead of the other way around

[19:56:08] <dstufft> qwcode: I think that means https://github.com/pypa/pip/blob/develop/pip/req/req_set.py#L488-L505 can be deleted now?

[19:56:17] <dstufft> since distribute should already be after setuptools

[19:58:22] <qwcode> dstufft, well, it's time remove all of that isn't? Parts 1-3? (and document that distribute to setuptools upgrades won't work anymore)

[19:59:23] <qwcode> i.e. that was a helpful "hack" that we were only going to support for a period

[19:59:25] <dstufft> qwcode: well I was going to remove just that bit and release it as 6.0.4 to fix https://github.com/pypa/pip/issues/2272 and https://github.com/pypa/pip/issues/2289

[19:59:42] <qwcode> oh, that bit is erroring?

[19:59:50] <qwcode> looking...

[20:00:08] <dstufft> qwcode: well, what's erroring techincally is that InstallRequirement().installed_version was not actually returning the installed version

[20:00:22] <dstufft> it was returning either the installed version OR the version inside of an sdist

[20:01:07] <dstufft> so I changed installed_version to actually return the installed_version (or None if there is no installed_version)

[20:01:41] <dstufft> and if req.installed_version is None, then it blows up

[20:06:48] <qwcode> dstufft, I can imagine a pretty odd case where distribute still wouldn't be after setuptools, if e.g. they did: "pip install SomethingThatRequiresSetuptools Distribute"

[20:07:11] <qwcode> but it's odd

[20:08:12] <qwcode> very few people directly trying to install distribute at this point

[20:10:30] <dstufft> ah yea

[20:10:31] <dstufft> that's true

[20:10:49] <dstufft> though it shouldn't matter if we convince jaraco to release a .whl of the distribute wrapper too ;)

[20:11:14] <dstufft> since then we won't require distribute to be after setuptools I think

[20:11:26] <dstufft> maybe we would

[20:12:08] <dstufft> qwcode: maybe I should just add and req.installed_version is not None to that conditional

[20:13:04] <qwcode> not sure that a wheel would help. yea, maybe just fix the conditional. this stuff was a mind bender a year ago, even more so now...

[20:34:06] <qwcode> jaraco, dstufft fyi, I'm just noticing that https://setuptools.pypa.io/en/latest/ is serving from https://bitbucket.org/ericholscher/setuptools with hacks to make it work with RTD

[20:34:34] <dstufft> qwcode: yea, I set setuptools.pypa.io because I was going to fix the docs in setuptools... then I forgot

[20:34:44] <qwcode> ok, as long as its known : )

[20:35:55] <sigmavirus24> tsk tsk tsk dstufft =P

[21:05:01] <xafer> any feedback on https://github.com/pypa/pip/pull/2239 ?

[21:16:59] <sigmavirus24> I think moving the comments next to the if/elif/else clauses is unnecessary

[21:38:11] <xafer> indeed, just changed that

[21:45:02] <xafer> any reason why pip uses distlib.markers.interpret for environment markers instead of markerlib like setuptools does ?

[21:48:51] <sigmavirus24> no clue

[22:12:53] <tomprince> xafer: I think at one point, the hope was that distlib was going to become what packaging now is.

[22:32:12] <xafer> that's what I was wondering. distlib and packaging seem to have similar scopes

[22:36:57] <xafer> and if wheels didnt need distlib.scripts.ScriptMaker, distlib wouldnt be needed anymore in pip

[22:37:37] <tomprince> <dstufft yea, and i've given him feedback about the fact he uses distlib as his personal playground for non standardized standards and he just brushes it off as "well you can choose not to use them"

[22:51:02] <xafer> well that clears up things :o

[23:39:24] <qwcode> distlib was underlying his other project, "distil" http://pythonhosted.org//distil/overview.html, which theoretically could have replace pip

Log file Viewer

Help | Karma | Search:

#pypa-dev logs for Tuesday the 30th of December, 2014