18:00:24 <felixfontein> #startmeeting Ansible Community Meeting
18:00:24 <zodbot> Meeting started Wed Sep  8 18:00:24 2021 UTC.
18:00:24 <zodbot> This meeting is logged and archived in a public location.
18:00:24 <zodbot> The chair is felixfontein. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:24 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:00:24 <zodbot> The meeting name has been set to 'ansible_community_meeting'
18:00:24 <felixfontein> #topic Agenda https://github.com/ansible/community/issues/539
18:00:24 <felixfontein> abadger1999 acozine andersson007_ baptistemm bcoca briantist cyberpear cybette dericcrago dmsimard felixfontein geerlingguy gundalow gwmngilfen ikhan_ jillr jtanner lmodemal misc nitzmahone resmo samccann tadeboro cidrblock thaumos zbr: ping!
18:00:28 <felixfontein> #info Agenda: https://github.com/ansible/community/issues/539 / Topics: https://github.com/ansible-community/community-topics
18:00:33 <briantist> o/
18:00:38 <andersson007_> o/
18:00:39 * dericcrago waves
18:00:43 <felixfontein> #chair briantist andersson007_ dericcrago
18:00:43 <zodbot> Current chairs: andersson007_ briantist dericcrago felixfontein
18:00:45 <jillr> o/
18:00:51 <dmsimard> \o
18:00:57 <cybette> o/
18:01:01 <felixfontein> #chair jillr dmsimard cybette
18:01:01 <zodbot> Current chairs: andersson007_ briantist cybette dericcrago dmsimard felixfontein jillr
18:01:05 <abadger1999[m]> boa tarde :-)
18:01:16 <felixfontein> #chair abadger1999[m]
18:01:16 <zodbot> Current chairs: abadger1999[m] andersson007_ briantist cybette dericcrago dmsimard felixfontein jillr
18:01:40 <felixfontein> #topic Updates
18:01:40 <felixfontein> #info Tadej Borovšak (@tadeboro) has been elected as the chairperson for the Ansible Community Steering Committee
18:01:43 <felixfontein> congratulations, tadeboro! :)
18:01:54 <tadeboro> o/
18:02:01 <cybette> congrats tadeboro !
18:02:02 <gundalow> o/
18:02:04 <briantist> congrats! 🎉
18:02:07 <dericcrago> congrats tadeboro !
18:02:10 <jillr> congrats and well earned!
18:02:13 <dmsimard> yay
18:02:18 <andersson007_> 💪
18:02:22 <cyberpear> o/
18:02:25 <felixfontein> #chair tadeboro gundalow cyberpear
18:02:25 <zodbot> Current chairs: abadger1999[m] andersson007_ briantist cyberpear cybette dericcrago dmsimard felixfontein gundalow jillr tadeboro
18:02:31 <gundalow> This is great
18:02:35 <cybette> #info Reminder to register for Ansible Contributor Summit 2021.09, thanks! https://ansiblecs202109.eventbrite.com/?aff=irc
18:03:48 <gundalow> We'd really appreciate the help in promoting the event, especially Tuesday 28th Sept, which will be great for new contributors.
18:04:18 <gundalow> Should I paste Greg's updates?
18:04:38 <cybette> yes please
18:05:14 <gundalow> #info ansible-social will be bridged soon, EMS is working on it for us (it's a big room, so it can't be done selfservice). I had hoped it would be today but likely this week
18:05:46 <tadeboro> I am in talks with our marketing so hopefully we will have something ready early next week that (blog post promoting AnsibleFest and activities around it + all the regular social media things).
18:05:54 <gundalow> #info on the topic of improving this meeting format, Gwmngilfen is still researching, we can have a proper debate when he's actually here :P
18:05:56 <felixfontein> about today's meeting: all topics that are marked 'next_meeting' aren't really in a state for being discussed today IMO (the first two because Greg isn't around, the others because they need more input)
18:06:21 <cybette> tadeboro: that's great, thanks!
18:06:33 <gundalow> #info the admin bot is now operational in all our rooms (Let us know if we've missed any) and today we got the chance to actually deal with a spammer today, so we know it works :P
18:07:31 <gundalow> #info we have about a month left for new collections to be included in Ansible 5.0.0. See https://github.com/ansible-community/community-topics/issues/32 for details
18:08:47 <gundalow> #info If anyone would like to help review the new Instruqt Labs: https://github.com/ansible/instruqt/pull/46 Only need to review the tracks/*/track.yml files
18:08:47 <github-linkbot> https://github.com/ansible/instruqt/pull/46 | open, created 2021-09-07T09:32:55Z by Ompragash: Ansible Community Instruqt tracks  
18:08:52 <felixfontein> is there a topic people would like to discuss today? (from https://github.com/ansible-community/community-topics/issues)
18:09:51 <gundalow> No other updates from me
18:09:56 <tadeboro> I just wanted to give a quick update on inclusion process and how things are progressing, but I think this does not need to be a separate topic.
18:10:24 <tadeboro> I am almost done with the second round of reviews and things are progressing nicely in general.
18:10:24 <felixfontein> #topic Ansible collection inclusion process (https://github.com/ansible-community/community-topics/issues/32)
18:11:15 <tadeboro> But I did notice that some of the new collection maintainers would probably need some help/info with things like deprecations, backward-compatibiliy, removing functionality ...
18:12:10 <felixfontein> potentially also semver
18:12:16 <tadeboro> So maybe this is something we could talk about in Contributor Summit/AnsibleFest.
18:15:01 <remindbot[m]> <a href="https://matrix.to/#/@cybette:ansible.im">@cybette:ansible.im</a> cyb-clock chimes every 15 minutes during the community meeting
18:15:04 <tadeboro> But all in all, it fels like maintainers of inclusion candidates for Ansible 5.0.0 are quite responsive.
18:15:23 <tadeboro> So this is basically allI had to say about this.
18:15:48 <tadeboro> The current state of affairs is summarized in the description in https://github.com/ansible-community/community-topics/issues/32
18:15:50 <felixfontein> thanks a lot for all the reviewing work!
18:16:10 <felixfontein> #info current state can be found in https://github.com/ansible-community/community-topics/issues/32
18:16:26 <andersson007_> I've just set up a reminder to take a look too after the summit
18:17:14 <andersson007_> oh, 12 october..
18:17:22 <jillr> that meta issue is super helpful, thanks tadeboro
18:17:27 <andersson007_> no, will set up ton next week:)
18:17:41 <andersson007_> +1
18:17:59 <felixfontein> a slightly related topic: netbox.netbox (included in Ansible) is still not working with ansible-core 2.11, and doesn't seem to have active maintainers
18:18:30 * sc68cal perks up
18:18:39 <felixfontein> #chair sc68cal
18:18:39 <zodbot> Current chairs: abadger1999[m] andersson007_ briantist cyberpear cybette dericcrago dmsimard felixfontein gundalow jillr sc68cal tadeboro
18:18:42 <gundalow> I've ben chatting with Partner Engineering about this
18:19:04 <felixfontein> #topic Unmaintained / broken collections included in Ansible
18:19:12 <andersson007_> sc68cal: nice to see you here!:) welcome!
18:19:24 <tadeboro> And the previous maintainer now helps maintain networktocode.nautobot fork that is maintained and solved that particular issue, but in a backward-incompatible way.
18:19:29 <gundalow> We think we could add a CI test to detect if a plugin is removed in a minor/revision release
18:20:04 <gundalow> deric.crago: is `networktocode.nautobot` something you've looked at, as part of c.networking?
18:20:29 <felixfontein> networktocode.nautobot is one of the Ansible 5 inclusion candidates
18:20:32 <dericcrago> no
18:20:36 <tadeboro> gundalow: Is netbox maintained by a Red Hat partner? I got the feeling that this collection is maintained by community.
18:20:50 <andersson007_> this is my main concern - now they responsive, in a year it can get totally dead
18:21:08 <tadeboro> Not sure what plans networktocode has with the nautobot.
18:21:23 <felixfontein> also the netbox.netbox repo is not in gh.com/ansible-collections
18:21:57 <andersson007_> moreover, we control things in ansible-collections org but we can't assign new maintainers in outer repos..
18:22:30 <andersson007_> third party ones i mean
18:22:42 <sc68cal> it sounds to me like it would be useful for the new maintainers of netbox.netbox to move it under gh.com/ansible-collections
18:22:54 <sc68cal> when those new maintainers are appointed
18:23:13 <andersson007_> we can kick out such collections from the package of course but it doesn't feel good for the project reputation
18:23:14 <felixfontein> can networktocode.nautobot be used as a replacement for netbox.netbox? or is it incompatible (on the service/... side, I have no idea what it actually works on :) )
18:23:18 <gundalow> I know some people at Networktocode, I can reach out to them
18:23:28 <gundalow> If there are a list of specific questions
18:23:40 <sc68cal> felixfontein: I think it is incompatible? I have been looking at nautobot as a consumer of netbox API
18:23:45 <tadeboro> netbox is a bit special since the main maintainer is now working on nautobot collection.
18:23:56 <felixfontein> andersson007_: well, if the collection doesn't get fixed, removing it might be the best choice, since right now it's broken and cannot be used in Ansible anyway
18:24:04 * samccann shows up 30 min late
18:24:12 <felixfontein> #chair samccann
18:24:12 <zodbot> Current chairs: abadger1999[m] andersson007_ briantist cyberpear cybette dericcrago dmsimard felixfontein gundalow jillr samccann sc68cal tadeboro
18:24:15 <andersson007_> felixfontein: sure
18:24:50 <sc68cal> I think kicking it out is the right call since it is broken
18:24:52 <abadger1999> I do think that it's inevitable that collections will need to exit the package.
18:25:03 <tadeboro> Do we have a means of getting the collection fixed and fixes merged?
18:25:33 <sc68cal> a new maintainer was appointed in may, but appears to have gone dark
18:25:49 <andersson007_> felixfontein: i'm trying to say that there are big downsides of inclusion third party things outside of ansible-collections
18:26:10 <felixfontein> andersson007_: there definitely are! and we still have to finish https://github.com/ansible-community/community-topics/issues/34 :)
18:26:31 <andersson007_> felixfontein: yeah
18:26:38 <felixfontein> tadeboro: there has already been a fix around for quite some time now, so that part is solved
18:26:53 <felixfontein> no idea who can actually merge it though (nobody did so far)
18:27:01 <abadger1999> We probably want a procedure right away and criteria down the road for removing a collection from the package.
18:27:26 <sc68cal> I don't think anyone is active in netbox.netbox who can merge it
18:27:29 <tadeboro> Ugh, maybe we should prioitize 34.
18:27:45 <andersson007_> felixfontein: maybe we should require to move repos to ansible-collections if maintainers want to see their stuff included? Just brainstorming
18:28:05 <felixfontein> andersson007_: you mean just for community collections, or for all collections?
18:28:12 <andersson007_> for all
18:28:17 <felixfontein> andersson007_: (whatever makes a collection "community" :) )
18:28:23 <dmsimard> andersson007_: I'm not convinced everyone would be OK with that
18:28:45 <felixfontein> I think if we require that, we're mainly reducing the number of collections included in Ansible ;)
18:28:53 <tadeboro> andersson007_: Not sure how this would work with collections such as Sensu Go, for example.
18:29:19 <samccann> yeah not sure how certified collections would react to that? I don't think all of them are in ansible-collections repo today?
18:29:25 <andersson007_> and another question is why do we include new collections? people can easily install them from galaxy:)
18:29:37 <andersson007_> just brainstorming, folks:)
18:29:37 <Zhenech> foreman/satellite is not in that org
18:29:43 <Zhenech> (and no intention to move :P)
18:29:55 <felixfontein> #chair Zhenech
18:29:55 <zodbot> Current chairs: Zhenech abadger1999[m] andersson007_ briantist cyberpear cybette dericcrago dmsimard felixfontein gundalow jillr samccann sc68cal tadeboro
18:30:00 <remindbot[m]> <a href="https://matrix.to/#/@cybette:ansible.im">@cybette:ansible.im</a> cyb-clock chimes every 15 minutes during the community meeting
18:30:24 <felixfontein> IMO collections can be hosted outside gh.com/ansible-collections, or even outside GH
18:30:27 <dmsimard> andersson007_: I would say that collections were included post-split in the best interests of backwards compatibility and to keep the "batteries-included" approach that the ansible package always had
18:30:34 <tadeboro> What bothers me with netbox is that by removing it from the Ansible package is that ansible's runtime will point to a tombstone.
18:30:39 <felixfontein> (just look at openstack, I don't think they want to migrate to GH :) )
18:30:57 <andersson007_> dmsimard: yeah, it's fair for stuff that was in ansible/ansible before the split
18:31:01 <felixfontein> tadeboro: it won't point to a tombstone, it will just point to something not included
18:31:22 <dmsimard> andersson007_: that said, I think "why?" is a good question and we should have consensus on an answer
18:31:29 <sc68cal> anyone running it on ansible > 3.4 already gets a tombstone where it breaks and their playbooks die with an error
18:31:56 <tadeboro> felixfontein: Right, but the fact is that things will still break.
18:32:10 <felixfontein> sc68cal: that's not a tombstone, just a fatal error :)
18:32:17 <sc68cal> :)
18:32:20 <felixfontein> tadeboro: it currently doesn't work anyway (basically since Ansible 4.0.0)
18:32:32 <felixfontein> tadeboro: so it will just be the error message that changes
18:32:46 <tadeboro> I wonder what would happen in the pre-collections world. Would the ansible core devs fix the modules that went bad?
18:33:05 <dmsimard> tadeboro: probably not if they were "community supported"
18:33:13 <sc68cal> I think they would have had to refactor it, since it's an argspec thing that changed
18:33:14 <Zhenech> tadeboro, the once that were "support: community"? no.
18:33:19 <gundalow> Ok, update from networktocode
18:33:19 <gundalow> 1. https://github.com/netbox-community/ansible_modules/discussions/526 talks about new maintainer
18:33:19 <gundalow> 2. networktocode.nautobot and netbox.netbox will become incompatible over time as the two upstream projects move away from each other
18:33:20 <felixfontein> potentially they would, or deprecate them.
18:33:30 <felixfontein> maybe the originally bad code wouldn't have been merged in the first case (though maybe it already was)
18:35:24 <felixfontein> gundalow: 1) unfortunately that new maintainer hasn't been active so far, at least not w.r.t. this problem
18:35:54 <gundalow> Seems like there are various people wanting to help (hi sc68cal :))
18:35:59 <felixfontein> (not sure whether he actually has permissions on that repo by now)
18:36:34 <felixfontein> there's actually a PR that fixes the current main problem: https://github.com/netbox-community/ansible_modules/pull/558
18:36:34 <github-linkbot> https://github.com/netbox-community/ansible_modules/pull/558 | open, created 2021-07-14T01:58:02Z by nahun: Quick fix to support ansible-core 2.11  
18:36:49 <gundalow> Seems like it needs a bit more work
18:36:49 <tadeboro> Maybe we should move the repo to ansible-collections so that people who actually want to help can do so because we can appoint new maintainers?
18:37:27 <sc68cal> I think moving the repo would be wise, because of that issue of getting new maintainers added
18:37:39 <samccann> do we want to own a collection that 'competes' with a different one that may (or may not) start being maintained?
18:38:29 <sc68cal> does the pedigree of the collection count? Was it spun out of ansible at some point into where it exists now in gh.com?
18:38:44 <sc68cal> or is it the reverse
18:38:57 <tadeboro> samccann: Competes? As far as I can tell, the old maintainers just want to get the netbox collection off their bcks because at least some of them are now working on nautobot.
18:38:59 <dmsimard> samccann: I recall briefly discussing that a few months back (what if we had two nginx collections?) but since nautobot is a fork of netbox, they will diverge so I don't feel there is a notion of competition strictly speaking (outside of the fork itself, which is beyond ansible)
18:39:00 <andersson007_> i think we won't have enough meeting in the future to think of every unmaintained collection
18:39:23 <andersson007_> meetingS
18:39:28 <samccann> tadeboro dmsimard: okay thanks for the explanation
18:39:29 <felixfontein> sc68cal: collections that took modules when ansible/ansible was split were automatically included in Ansible 2.10 ("grandfathered")
18:40:06 <tadeboro> sc68cal: netbox content was part ansible in 2.9 and then split into separate collection we are talking about today.
18:40:28 <sc68cal> ack. thanks felixfontein and tadeboro
18:40:55 <sc68cal> so I would say that historical fact at least is a point in favor of moving the repo into ansible-collections
18:41:39 <sc68cal> the netbox community, at least from my vantage point as a user, it appears they only focus on netbox itself and pynetbox
18:42:47 <felixfontein> I guess the main question is who can actually move the repo? because we can't (since we don't have access to https://github.com/netbox-community/)
18:43:20 <tadeboro> felixfontein: I think the old maintainer could. So maybe we should start asking there?
18:43:28 <gundalow> I've just got the details
18:43:46 <gundalow> Signing up for the new netbox Slack and I'll ask them
18:43:57 <abadger1999> do we have someone who wants to maintain the collection?
18:44:30 <tadeboro> abadger1999: https://github.com/netbox-community/ansible_modules/discussions/526#discussioncomment-1293334 seems to indicate that we do.
18:44:30 <abadger1999> If we don't have someone who wants to maintain it, I think it might be better to drop it rather than work to move it and then have it be unmaintained under our namespace.
18:44:50 <sc68cal> abadger1999: myself
18:44:57 <gundalow> ah, they are discussing maintainers in netbox's Slack instance yesterday
18:45:00 <remindbot[m]> <a href="https://matrix.to/#/@cybette:ansible.im">@cybette:ansible.im</a> cyb-clock chimes every 15 minutes during the community meeting
18:45:08 <tadeboro> sc68cal is one of those nice people ;)
18:45:10 <sc68cal> I am working off a verbal agreement with my employer for some time to devote to this
18:45:13 <abadger1999> <nod> Okay.
18:45:27 <sc68cal> once I get the formal go ahead I will post a concrete commitment
18:45:36 <sc68cal> think 20% time
18:46:04 <abadger1999> We can fork the github repo instead of moving it if need be, I think.  The more important thing might be getting access to galaxy for the collection.
18:46:21 <briantist> to abadger1999 's point though, we should be prepared to drop collections with no maintainers, and redirect folks who use the Ansible package to keep installing it manually if they want it (if not this collection, it's going to happen at some point)
18:46:45 <sc68cal> briantist: +1
18:47:00 <tadeboro> briantist: Yep. We should probably do this as part of the https://github.com/ansible-community/community-topics/issues/34
18:47:40 <briantist> yeah that's probably the right place for discussion on that
18:47:42 <felixfontein> maybe we (as Ansible maintainers) should say that either the collection gets fixed by $DATE_IN_THE_FUTURE, or it will get removed from Ansible 5; where $DATE_IN_THE_FUTURE is somewhere close to the inclusion deadline for new collections?
18:49:13 <tadeboro> Since it is getting late today, I say we mark the 34 as next_meeting and gather some feedback in that issue for the next meeting. Things like felixfontein suggested can go in that issue so that others can comment on.
18:49:14 <jillr> I concur with felixfontein's idea, in general
18:49:19 <andersson007_> as the meeting is almost done, would be nice if people put their thoughts in the issue
18:49:43 <felixfontein> sounds good to me
18:49:46 <andersson007_> tadeboro: was a bit faster:)
18:50:01 <felixfontein> I'll switch to open floor then
18:50:19 <tadeboro> andersson007_: But it feels nice to have people who are aligned with my mind ;)
18:50:23 <felixfontein> #topic open floor
18:50:29 <andersson007_> tadeboro: :)
18:52:24 <felixfontein> so, does anyone have questions, topics, ideas ... for open floor?
18:53:07 <tadeboro> I must admit that last few weeks were quite harsh and I would not mind ending a few minutes early ;)
18:53:18 <jillr> same  :)
18:53:38 <felixfontein> same for me actually :)
18:53:57 <felixfontein> tadeboro: I'm glad you still managed to review the inclusion candidates!
18:54:21 <gundalow> FYI I'm chatting with the netbox folks. I'll report back what I hear
18:54:43 <felixfontein> oh, btw, there's a PR for split-controller testing (https://github.com/ansible/ansible/pull/75605), and depending on the collection split-controller testing will require fixing quite some integration tests
18:54:43 <github-linkbot> https://github.com/ansible/ansible/pull/75605), | open, created 2021-08-31T07:44:18Z by mattclay: ansible-test - split controller/target testing [needs_rebase,WIP,feature,affects_2.12] 
18:55:17 <felixfontein> the PR isn't complete yet, so some things can still change, but some other things will not
18:55:22 <gundalow> sc68cal: Are you on netdev-community.slack.com ?
18:55:30 <briantist> any tips on testing split controller testing felixfontein ?
18:55:44 <felixfontein> if you maintain a collection which has integration tests, you might want to try running your tests with that branch :)
18:55:45 <gundalow> mattclay: ^
18:55:45 <sc68cal> gundalow: no I am not
18:56:10 <sc68cal> gundalow: what's that slack instance for?
18:56:17 <tadeboro> felixfontein: I did review them because I felt bad for people who were actually fixing things and waiting for my feedback.
18:56:29 <samccann> felixfontein - that is a ...big PR.  Do you know if there are docs needs associated with it? (looks like it's targeted for 2.12?)
18:56:37 <jillr> I'm going to run away early and try to sneak some lunch in before more meetings, thanks y'all!
18:56:43 <felixfontein> briantist: the main one is that you should look for usages of output_dir or role_path or lookup('file', ...) - if you are testing modules (and not plugins, which run on the controller), these are points which could break
18:56:54 <gundalow> sc68cal: I only just found out about it, though seems to be fairly active. It's the NetBox Community. Ther eis a `#netbox-ansible` channel which had an existing discussion about people becoming maintainers
18:56:58 <briantist> felixfontein: are there any immediate changes that need to be done in the invocation, or does invocation not usually have to change?
18:57:18 <sc68cal> gundalow: ahhhh! okay awesome. I was struggling to find where netbox was doing chat
18:57:26 <sc68cal> gundalow: i'll join it
18:57:40 <briantist> right, I'm only testing plugins right now, but there will be modules (I know I've been saying that for almost a year... but it's getting close!)
18:58:08 <tadeboro> Enjoy your meal jillr!
18:58:26 <sc68cal> gundalow: do you happen to have the invite URL to join that slack instance?
18:58:41 <gundalow> oh, stupid slack, sure
18:58:59 <tadeboro> I am also leaving for today. Thank you all!
18:58:59 <felixfontein> briantist: i.e. for the hashi_vault lookup it won't really affect you ;)
18:58:59 <felixfontein> samccann: it is indeed! I'm not aware of documentation so far
18:58:59 <felixfontein> and yes, I think it's targeted for 2.12, it's probably the main feature for 2.12 that isn't done yet
18:58:59 <felixfontein> the main difference is that so far, controller == target in ansible-test's integration testing for modules. which isn't the case anymore with split-controller testing.
18:59:02 <sc68cal> ah actually I found it via google, and found the mailing list post
18:59:22 <gwmngilfen> gundalow another Slack bridge candidate? ;)
18:59:49 <gundalow> #info If you are interested in netbox.netbox join `#netbox-ansible` on https://netdev-community.slack.com/join/shared_invite/zt-mtts8g0n-Sm6Wutn62q_M4OdsaIycrQ
19:00:00 <gundalow> Gwm (non-work): hum, not sure
19:00:34 <gwmngilfen> Well, the option is there, should we need it
19:01:02 <gwmngilfen> tadeboro congrats BTW
19:01:11 <felixfontein> briantist: invocation should be backwards compatible
19:01:11 <felixfontein> there is a new syntax to specfiy exactly what should run on controller and target, but the old options are translated automatically to the new ones
19:01:11 <felixfontein> (how exactly they do might still change a bit though)
19:01:11 <felixfontein> #endmeeting