[15:42:38] <ionelmc> essentially that would equate to a distutils-only setup.py but I want some other stuff that are present in setuptools (like find_packages and include_package_data)
[15:43:06] <tos9> you don't need to use setuptools.setup to use find_packages I imagine
[15:43:34] <tos9> And IIRC include_package_data doesn't really "work" -- is there other stuff you want too?
[15:43:56] <ionelmc> tos9: no, i need to, i'm lazy and i don't want to reimplement it
[15:44:12] <tos9> ionelmc: Sorry, which was that a response to?
[15:44:43] <ionelmc> tos9: tos9> you don't need to use setuptools.setup to use find_packages I imagine
[15:44:57] <tos9> ionelmc: Yes, I wasn't saying to reimplement it
[15:45:11] <tos9> ionelmc: I'm saying I can't imagine distutils.core.setup(..., packages=setuptools.find_packages()) doesn't work
[15:45:20] <ionelmc> once you import setuptools your poor old distutils gets monkeypatched
[15:45:22] <tos9> find_packages is probably just a function that does the descending and returns the list.
[15:45:45] <tos9> ionelmc: oh. I don't know about the arcane details of that, but that's even if you import distutils.core first?
[15:46:03] <ionelmc> tos9: it monkeypatch all the distutils commands
[19:51:56] <ronny> the plan is to generate dists using setup_requires, wheels install in a separate virtualenv
[21:36:01] <DanielHolth> my setup-requires implementation wasn't the hit I hoped it would be. people seem to want everything to happen in setup.py. Even though setup-requires is something that pip should be doing in its role as a build system
[21:37:22] <jaraco> ronny, I added two comments in the setuptools_scm bug tracker about my response to reading the readme.txt.
[21:37:50] <dstufft> I think that, at least personally, people generally don't want to add more hacks ontop of an already fragile toolchain if they need to