20:02:02 <jborean93> #startmeeting Ansible Windows Working Group
20:02:02 <zodbot> Meeting started Tue May 31 20:02:02 2022 UTC.
20:02:02 <zodbot> This meeting is logged and archived in a public location.
20:02:02 <zodbot> The chair is jborean93. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
20:02:02 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
20:02:02 <zodbot> The meeting name has been set to 'ansible_windows_working_group'
20:02:25 <jborean93> definitely not as punctual as nitz
20:02:25 <briantist> hello
20:02:29 <jborean93> hey
20:02:37 <jborean93> #topic Agenda https://github.com/ansible/community/issues/644
20:03:00 <jborean93> Nothing on the agenda so straight to the open floor
20:03:04 <jborean93> #topic Open Floor
20:03:11 <briantist> I updated the docs build PR; rather than remove the PR workflow, I removed the publish step (the one with the repo write access)
20:03:13 <jborean93> nitz is out sick today so it's just me
20:03:33 <briantist> so what remains would still build and validate the docs, and then it would post a comment on the PR with the details
20:03:55 <jborean93> i.e. failures in the build?
20:04:23 <briantist> the build step still has read-only, the PR step still only has PR write access (not repo contents write), which is the same permission that the current "comment when RST files are modified" workflow has
20:05:06 <briantist> the comment would link to the CI run which has the rendered docs attached as an artifact, so they can be downloaded and verified locally
20:05:21 <briantist> and it also shows diff style output of the docs changes, and a list of which files changes
20:05:21 <jborean93> cool, probably for the best for now. Will leave it to nitz to do the review though
20:05:47 <briantist> the comment looks like this: https://github.com/ansible-collections/community.hashi_vault/pull/220#issuecomment-1038393614
20:06:03 <briantist> except it won't have the links to the rendered stuff
20:06:15 <jborean93> does it edit the comment on new commits or add a new comment?
20:06:19 <briantist> edits
20:06:29 <jborean93> sweet
20:06:30 <briantist> don't want to be spamming the thread :)
20:07:01 <jborean93> definitely a good choice
20:07:38 <briantist> yeah that's fine we can wait for nitz, as I said permission is the same as the other workflow (that workflow can go away once this stuff is in place), but if he's still not liking that, then at the very least, just the build portion is sueful, since it will fail CI if the docs are not valid
20:08:17 <jborean93> yep agreed
20:08:36 <jborean93> I haven't had a chance to look into the hangs with pypsrp yet sorry
20:08:54 <briantist> no worries, I've got a million other things grabbing my attention
20:09:15 <briantist> yesterday I started writing an ansible galaxy proxy that uses Artifactory as its backend 😬
20:09:45 <briantist> the requests for them to add galaxy as a backend type have gone nowhere for ~3 years and that's where we store all our other artifacts
20:10:24 <briantist> and I looked into running galaxy_ng but... it's just more than I wanna deal with
20:11:41 <briantist> speaking of, thanks for this comment: https://github.com/ansible/ansible/issues/65933#issuecomment-567205147
20:12:03 <briantist> the format of the `galaxy_token` file doesn't appear to be documented anywhere else
20:12:22 <briantist> or I just couldn't find it
20:12:33 <jborean93> IIRC it's literally just a yaml file with `token: ...`
20:13:08 <briantist> it is! but I had no way to know that... b coca hinted it was probably yaml, but I didn't know what it was expecting
20:13:26 <jborean93> https://github.com/ansible/ansible/blob/devel/lib/ansible/galaxy/token.py#L101
20:13:46 <briantist> yeah trawling the source would've been my next stop
20:15:44 <jborean93> gl with getting galaxy support for Artifactory, the only real implementation I've seen is the pulp one which is what galaxy_ng uses on the backend
20:16:14 <jborean93> I know sivel has his own API to support a provides like API and I've written a mock one for testing in the past
20:16:20 <briantist> I have it working! install/download support working since yesterday, and as of today, upload it working
20:16:34 <jborean93> roles and collections?
20:16:35 <briantist> yes, actually sivel's project (amanda) was a huge inspiration for this
20:16:45 <briantist> oh not roles, collections only, that's all I care about
20:16:49 <jborean93> I suppose roles won't make sense here
20:16:50 <jborean93> yea
20:17:12 <jborean93> sweet though, always nice to see it being supported by more repository projects than just ours
20:17:31 <briantist> it wasn't intended but so much of the code resembles sivel's now that I am going to ask him to take a look and see if he wants to be credited lol
20:17:57 <briantist> the hard part will be making it work the way Artifactory's "virtual" repositories work
20:18:01 <jborean93> lol, he'll probably say you need to ensure his toiletwater collection is always available in the repo. Like Tom from Myspace :)
20:18:11 <briantist> lolllll
20:18:35 <briantist> I will gladly host it though it will be internal to my company
20:19:41 <jborean93> still sweet though
20:19:51 <briantist> the virtual repository thing is where the artifactory repository sort of shadows an upstream, and so if  you request something and it's not avialable local, then it downloads it and serves it and saves it, so next time it doesn't hit upstream
20:20:08 <briantist> that'll be really helpful for avoiding galaxy rate limiting which seems to be very strict :-/
20:20:30 <briantist> and it will avoid having to manually download/upload beforehand if I can make it transparent
20:21:13 <jborean93> yea my previous org had something like that for Maven artifacts. I wonder if they still support that or whether it needs to be explicitly allowed due to the supply chain concerns emerging today
20:21:16 <jborean93> these days*
20:22:03 <briantist> I think artifactory has a bunch of controls on those type of things (which I won't have at my disposal here)
20:24:44 <jborean93> cool, anything else you wanted to bring up?
20:24:59 <briantist> nope all good
20:25:14 <jborean93> alrighty, will catch you next time
20:25:18 <jborean93> #endmeeting