15:01:35 <Shrews> #startmeeting Ansible Core Public IRC Meeting
15:01:35 <zodbot> Meeting started Thu Nov 12 15:01:35 2020 UTC.
15:01:35 <zodbot> This meeting is logged and archived in a public location.
15:01:35 <zodbot> The chair is Shrews. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:35 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:01:35 <zodbot> The meeting name has been set to 'ansible_core_public_irc_meeting'
15:01:40 <sdoran> \o
15:01:44 <Shrews> #chair mattclay bcoca
15:01:44 <zodbot> Current chairs: Shrews bcoca mattclay
15:01:53 <bcoca> darn .. it was a trap!
15:01:54 <Shrews> #info https://github.com/ansible/community/issues/570
15:01:57 * mattclay waves
15:01:58 <Shrews> lol
15:02:03 <Shrews> looks like an empty agenda today
15:02:17 <Shrews> we'll just open the floor for a bit
15:02:19 <Shrews> #topic Open floor
15:02:21 <bcoca> open floor then, close after 5 mins if nothing/noone
15:02:27 <Shrews> ++
15:02:46 <Shrews> much easier than 2 days ago
15:02:52 <sdoran> Quite the contrast.
15:03:19 <Shrews> #chair sdoran
15:03:19 <zodbot> Current chairs: Shrews bcoca mattclay sdoran
15:03:23 <felixfontein> hi
15:03:25 <Shrews> just so you don't feel left out :-P
15:03:37 <Shrews> hi felix
15:03:50 <felixfontein> if you are bored I guess I can come up with some questions :P
15:05:07 <Shrews> open floor, so fire away
15:05:26 <felixfontein> I was originally joking
15:05:36 <felixfontein> but let me take a look at some open PRs...
15:06:13 <felixfontein> mattclay: do you think the import tests you mention in https://github.com/ansible/ansible/pull/71734#issuecomment-717483287 need to be there before that PR can continue?
15:06:58 <sivel> he likely won't be online for at least another hour I think
15:07:06 <mattclay> I'm here.
15:07:11 <sivel> oh hey :)
15:07:24 <felixfontein> sivel: he wrote something a few minutes ago in here :)
15:07:36 <sivel> oh, a "/me
15:07:44 <felixfontein> assuming it wasn't sleep-typing
15:07:48 <felixfontein> (if that is a thing)
15:08:10 <Shrews> #link https://github.com/ansible/ansible/pull/71734#issuecomment-717483287
15:10:03 <mattclay> Since we'll be running validate-modules on non-module plugins, most issues with imports will still be exposed without the additional sanity test. They're just not likely to be reported as nicely, and we'll only cover the version of python used for validate-modules instead of all supported versions.
15:11:01 <mattclay> So we could get by without it, but I think we should plan on extending the test sometime during the 2.11 dev cycle.
15:11:33 <mattclay> My biggest concern is that we won't find time for it and it won't get done.
15:11:59 <felixfontein> I agree. I already started playing around with the current import test (https://github.com/ansible/ansible/pull/72497), it needs a less-minimal venv though to not choke on jinja2 for every plugin
15:12:16 <mattclay> Yeah, it needs a venv with the minimal requirements for running ansible itself.
15:12:43 <felixfontein> I guess creating an empty one and installing the requirements is not a good idea, since that would run a lot, right?
15:12:44 <mattclay> Since those are the only requirements that can be guaranteed to be available to all plugins on the controller.
15:13:34 <mattclay> We already do that, so we can continue the pattern. At some point I'd like to optimize that be pre-generating those environments in the test container, but that should come later.
15:13:52 <felixfontein> btw, are the minimal venvs for the module import test always re-created during test execution? I guess they are since they don't have to install something, but I'm not 100% sure (didn't spend enough time digging through it)
15:13:59 <mattclay> Yes, they are.
15:14:11 <felixfontein> ok. I think I'll try that then
15:14:14 <mattclay> Well, effectively -- at least in CI, since we're always running with `--docker`.
15:14:24 <mattclay> If you're running without `--docker` they'll be persisted.
15:15:05 <felixfontein> for non-minimal venvs that might be a problem, if no longer used requirements are not removed. not sure how often that actually happens though.
15:16:26 <mattclay> I'm not too worried about us removing ansible-base requirements at this point, but yeah, if we did, it would be a problem.
15:17:18 <felixfontein> I guess it would suffice to add explicit `pip uninstall` calls for such requirements at that point
15:17:51 <mattclay> I'd probably just hash the requirements file at that point and use it as part of the temp path to the venv.
15:18:19 <mattclay> But don't worry about that now, we can save that for when we actually do remove requirements.
15:18:54 <felixfontein> sounds good.
15:19:11 <felixfontein> I'll try to work on that the next days
15:19:33 <felixfontein> who needs to/should review #71734?
15:21:41 <Shrews> i think mattclay is the primary on that. i see sivel was requested but not sure if he'll poke at it
15:22:02 <mattclay> I'd volunteer, but I'm going to be on PTO starting tomorrow, and I won't have time to review it before then.
15:22:24 <mattclay> I can review when I get back though.
15:23:22 <felixfontein> if you don't mind asking, for how long will you be gone?
15:23:34 <felixfontein> and in any case, enjoy your PTO!
15:23:55 * bcoca repowers clonning vat
15:24:30 <mattclay> I'll be back November 30th.
15:24:45 <gundalow> mattclay: enjoy!
15:26:02 <Shrews> ok, any other topics?
15:27:16 <felixfontein> bcoca: why did you turn it off in the first place?
15:27:41 <felixfontein> I think nothing from my side. at least nothing I remember now ;)
15:28:38 <Shrews> cool. i'll give everyone back 31 minutes to their day then
15:28:42 <Shrews> thanks everyone
15:28:56 <Shrews> #endmeeting