Log file Viewer

[00:41:42] <borgoms> having trouble with a pip package. pip install xerox on a windows

[00:41:42] <borgoms> machine. error: `xerox.base.Pywin32NotFound`

[00:47:01] <tos9> borgoms: Seems pretty clear?

[00:50:44] <borgoms> i'm sleuthing rn

[01:56:51] <Ivo> borgoms, it needs the pywin32 package but it hasnt declared that proplerly in its code

[01:57:26] <Ivo> borgoms, try installing "pypiwin32" first

[15:49:44] <bhahn> hey guys, i’m trying to understand https://github.com/certifi/python-certifi/issues/26#issuecomment-146142947. this doesn’t mean that python itself is going to be built using openssl 1.0.2d right?

[16:01:35] <dstufft> bhahn: correct

[16:01:49] <dstufft> bhahn: the cryptography project will link against a different openssl

[16:04:49] <bhahn> ah

[16:13:39] <bhahn> so i’m trying to propagate this change back to requests and am having trouble

[16:14:03] <bhahn> do i need to rebuild requests into a wheel like cryptography?

[16:17:24] <dstufft> I'm not sure what you mean by propagate it back to requests, but you might be looking for #python-requests where lukasa is at to explain in more detail what he's doing :D

[16:21:16] <bhahn> oh, even better :)

[16:34:33] <nedbat> i am having difficulties where packages are installed twice, once as -e (in the venv/src directory, with a .egg-link file pointing to it), and once as a real installation (a .egg-info file). Any idea how that happens and what I can do to prevent it?

[16:36:21] <dstufft> nedbat: it's a none problem :(

[16:36:22] <dstufft> known

[16:36:26] <dstufft> not sure how I fucked that up

[16:36:49] <nedbat> dstufft: do you know how to make it happen? I thought a -e followed by a real install would do it, but it didn't.

[16:36:59] <nedbat> dstufft: or do you know what issue # it is?

[16:37:05] <dstufft> pretty sure it's the other way around

[16:37:14] <dstufft> real install followed by -e

[16:37:20] <nedbat> oh, i'll try that.

[16:40:03] <nedbat> dstufft: you are right: I get two when I try that. But in my real case, Python runs the older code, but pip thinks it has the newer code, so it won't upgrade. I'm not seeing that in my toy test here.

[16:46:55] <nedbat> dstufft: btw, I'm curious: what would help to get the issue and pr backlog reduced on pypa/pip ?

[16:48:47] <dstufft> nedbat: more people reviewing the PRs that are there and either finishing off PRs that aren't ready to be merged, or writing good PRs for issues. There's a handful of pip folks at all, and most don't have a ton of time. I'm the one with the most time but I'm split between pip and warehouse & all the other packaing stuff

[16:49:59] <dstufft> nedbat: basically manpower is the problem, and pip's code base sucks so it's hard to write/review PRs for it :/

[16:50:04] <nedbat> dstufft: it's hard to get people to do that style of work. I tweeted, "want to help with open source? coverage.py could use issue triagers!" It got 10 retweets, and no volunteers.

[16:50:38] <dstufft> xafer: has been going back through old issues and triaging them which has been awesome

[16:51:33] <nedbat> it's great to have other pairs of hands

[16:52:27] <willingc> dstufft, nedbat: pick me, pick me. On a serious note, keep asking :)

[16:53:35] <willingc> I think issue triaging needs a PR facelift as well as greater recognition.

[16:54:30] <nedbat> dstufft: I wrote a super-simple kit to experiment with: https://github.com/nedbat/version_dummy. When pip installs v2.0, all the messages say 2.0, except the very last one: Successfully installed version-dummy-1.0 ??? Where is that 1.0 coming from?

[16:55:11] <dstufft> nedbat: does ``pip list`` say 2.0 or 1.0 after the install is over

[16:56:41] <nedbat> 2.0

[16:57:02] <nedbat> (btw, i'm using pip v6.0.8 to match our real environment)

[16:58:05] <dstufft> nedbat: probably that's pkg_resources caching what's already installed and a bug wher we're not accurately clearing said cache

[16:58:14] <nedbat> ok

[20:37:29] <xafer> and there are still a lot to triage :)

[20:37:53] <nedbat> xafer: thanks for doing it.

[20:44:28] <pdobrogost_> xafer: I'd like to thank you, too.

[20:45:38] <xafer> :) well 500+ issues is indeed a lot :o

[20:53:10] <nedbat> xafer: i'm curious if you have ideas about how to interest others in doing that kind of work?

[20:58:20] <xafer> not really... it's already hard to convince myself it is a good use of my time :o

[20:59:19] <nedbat> xafer: :(

[21:00:45] <xafer> but it's a good way to learn about the project and its limitations