Log file Viewer

[05:41:19] <Nakato> lifeless: https://review.openstack.org/#/c/278210

[10:01:20] <lifeless> Nakato: ack

[11:05:33] <dstufft> woo threads on distutils-sig

[11:05:37] <dstufft> time to respond

[12:05:10] <dstufft> sigmavirus24_awa: what do you think about ``twine wheel`` and ``twine sdist``? ``twine wheel`` could create a wheel from a given sdist, and ``twine sdist`` could create a sdist from a given directory on disk.

[12:25:03] <ronny> dstufft: i suppose its intended that twone uses the build tool pep?

[12:25:12] <dstufft> yea.

[12:25:45] <ronny> sounds good

[12:25:59] <ronny> please ensure that the sdist archive format is consistent over platforms

[12:26:17] <ronny> and also that it is usable as a library for devpi ^^

[15:38:55] <sigmavirus24> dstufft: how would that relate to https://github.com/pypa/twine/issues/160 ?

[15:39:34] <sigmavirus24> dstufft: or rather, should that replace ^

[15:39:42] <sigmavirus24> I think I like the UX better in 160 as it is proposed

[15:41:06] <dstufft> sigmavirus24: not sure yet. well the UX specifically isn't something I'm super concerned about rather than the idea of twine replacing ``setup.py sdist`` and ``setup.py bdist``

[15:41:30] <sigmavirus24> dstufft: right

[15:41:37] <sigmavirus24> That's part of the idea in 160

[15:41:54] <sigmavirus24> I (personally) prefer having a "build" command instead of mimicking/proliferating the sdist/wheel/etc. as separate commands

[15:42:21] <sigmavirus24> That said, I'm all on board with this

[15:42:27] <sigmavirus24> I haven't followed that thread very closely though

[15:42:37] <sigmavirus24> because I need more hours in a day

[15:42:43] <dstufft> sigmavirus24: I think maybe hold off on #160? till we shake out some stuff in distutils-sig?

[15:43:36] <sigmavirus24> dstufft: yeah, it's not top of my priorities at the moment to be honest

[18:37:57] <lifeless> dstufft: ping

[18:38:22] <dstufft> lifeless: pong

[18:38:34] <lifeless> dstufft: re: teh abstraction pep

[18:38:53] <lifeless> dstufft: I've rewritten it entirely once already; I'm happy to do another entire iteration

[18:39:11] <lifeless> dstufft: Guido never had a concern about asking devs to redo things to match what he wanted :)

[18:39:28] <lifeless> dstufft: but - I'd need at least a hint about what you want different :>

[18:45:11] <dstufft> lifeless: One of the main reason I'm doing a competing PEP as opposed to that, is writing the PEP out is the best way I can think of to try and figure out what it is that's rattling around in my head as a sort of rubber duck. I realized I started essentially writing a new PEP in a response to the email thread kicking back up that sort of made a few connections in my head I think. The other thing Is that I'm not sure it's actually better

[18:45:11] <dstufft> that the current approach, so I don't want to just throw away what you have now, this will provide two sort of different kinds of ways to approach this, and then the process can decide which one is the path we want to go down (assuming I don't end up throwing it out because it ends up being untennable to do it, or you don't throw yours out because you love mine so much)

[18:45:29] <dstufft> I've been trying to write an email reply since last night heh

[18:46:45] <dstufft> So I've been writing all day to try and get something formulated ASAP

[18:47:14] <lifeless> okay

[18:47:26] <lifeless> just - feel free to lean on my time, so that you're not the bottleneck

[18:50:16] <dstufft> lifeless: will do :) thanks!

[23:50:01] <LiquidLight> 'allo everybody :) is https://virtualenv.pypa.io/en/latest actually broken or am I having a special moment?