18:00:04 <felixfontein> #startmeeting Ansible Community Meeting
18:00:04 <zodbot> Meeting started Wed Jun 23 18:00:04 2021 UTC.
18:00:04 <zodbot> This meeting is logged and archived in a public location.
18:00:04 <zodbot> The chair is felixfontein. Information about MeetBot at http://wiki.debian.org/MeetBot.
18:00:04 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
18:00:04 <zodbot> The meeting name has been set to 'ansible_community_meeting'
18:00:04 <felixfontein> #topic Agenda https://github.com/ansible/community/issues/539
18:00:04 <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!
18:00:08 <felixfontein> #info Agenda: https://github.com/ansible/community/issues/539 / Topics: https://github.com/ansible-community/community-topics
18:00:14 * gundalow waves
18:00:15 <briantist> o/
18:00:20 <abadger1999> Hi
18:00:23 <andersson007_> o/
18:00:24 <cybette> o/
18:00:27 <gwmngilfen> o/
18:00:28 <jillr> o/  I'm half here (sorry, lot going on) but I'll chime in as I can
18:00:33 <felixfontein> #chair gundalow briantist abadger1999 andersson007_ cybette gwmngilfen jillr
18:00:33 <zodbot> Current chairs: abadger1999 andersson007_ briantist cybette felixfontein gundalow gwmngilfen jillr
18:00:51 <felixfontein> #topic Updates
18:00:51 <felixfontein> #info ansible-core 2.11.2, ansible-base 2.10.11, and Ansible 2.9.23 have been released
18:01:13 <briantist> (same as jill, in the middle of another meeting)
18:01:16 <dmsimard> o/
18:01:22 <tadeboro> o/
18:01:24 <felixfontein> #chair briantist dmsimard tadeboro
18:01:24 <zodbot> Current chairs: abadger1999 andersson007_ briantist cybette dmsimard felixfontein gundalow gwmngilfen jillr tadeboro
18:01:37 <cidrblock> hey all
18:01:45 <felixfontein> #chair cidrblock
18:01:45 <zodbot> Current chairs: abadger1999 andersson007_ briantist cidrblock cybette dmsimard felixfontein gundalow gwmngilfen jillr tadeboro
18:01:48 <felixfontein> hi cidrblock!
18:01:50 <cyberpear> o/
18:01:56 <felixfontein> #chair cyberpear
18:01:56 <zodbot> Current chairs: abadger1999 andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gundalow gwmngilfen jillr tadeboro
18:01:58 <samccann> \o
18:01:59 <abadger1999> Welcome!
18:02:14 <felixfontein> #chair samccann
18:02:14 <zodbot> Current chairs: abadger1999 andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gundalow gwmngilfen jillr samccann tadeboro
18:02:25 <cybette> #info The Bullhorn issue 29 will be pushed back by a week (new release date: July 1, 2021)
18:04:32 <felixfontein> any other updates?
18:04:37 <acozine> o/
18:04:46 <felixfontein> #chair acozine
18:04:46 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gundalow gwmngilfen jillr samccann tadeboro
18:04:54 <lmodemal> o/
18:05:01 <felixfontein> #chair lmodemal
18:05:01 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gundalow gwmngilfen jillr lmodemal samccann tadeboro
18:05:19 <felixfontein> #topic Standardising IRC modes for spam prevention + Matrix
18:05:19 <felixfontein> #info Discussion: https://github.com/ansible-community/community-topics/issues/28
18:05:22 <felixfontein> gwmngilfen is busy with family stuff, but will try to do his best to respond if there are questions about this
18:06:20 <gwmngilfen> o/
18:06:36 <felixfontein> #chair gwmngilfen
18:06:36 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gundalow gwmngilfen jillr lmodemal samccann tadeboro
18:06:56 <gwmngilfen> i just want to make sure all the prtal rooms work and that no spam is allowed in as a result. i think i have op in all the relevant rooms, so its just about permission
18:07:29 <cyberpear> prtal?
18:07:36 <abadger1999> portal
18:07:42 <gwmngilfen> ta
18:07:58 <gwmngilfen> i.e that users coming from atrix can speak in protected rooms
18:08:11 <abadger1999> what makes a room a portal room vs a non-portal room?
18:08:11 <gwmngilfen> *matrix, stupid fingers
18:08:44 <cyberpear> +q is the only one I'd avoid unless actively getting spammed... it's harder for folks to jump on webchat and ask a quick question if they don't have a registered nick
18:08:53 <gwmngilfen> its an implemetation detail
18:09:21 <gwmngilfen> cyberpear: we're already using +r in some rooms, so I'd argue +q is more user friendly than +e (because then at least they can join
18:09:52 <cyberpear> yeah, +r is even worse, IMO
18:09:53 <gundalow> We don't currently require registered nickname in #ansible on Libera.chat
18:09:56 <abadger1999> Does +r redirect them somewhere that they are told why they can't join?
18:10:05 <gundalow> abadger1999: not currently
18:10:15 <abadger1999> +q seems like it might be the equivalent of an error that occurs silently.
18:10:33 <gwmngilfen> abadger1999: portal rooms are when you join an irc room from matrix (so #ansible-community:libera.chat is created), plumbed rooms are where you add a bridge to an existing room. the net result is the same and we already have our rooms
18:11:29 <cyberpear> gwmngilfen++ thanks for the explanation!
18:11:29 <zodbot> cyberpear: Karma for gsutcliffe changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
18:11:50 <gwmngilfen> i'm happy to drop +qe entirely if we're not spammed, but this room definitely had some
18:12:05 <gwmngilfen> i'm unclear if +e can be used with +r
18:12:06 <felixfontein> yes, we definitely had spam here recently. though it's already a few days ago at least
18:12:49 <cyberpear> possible to have a bot to toggle +q when a channel is extra busy?
18:13:02 <cybette> maybe we keep +qe but drop it on #ansible-community for contributor summits and other one-time needs?
18:13:08 <abadger1999> Maybe we shouldn't change permissions until we're telling people to join via matrix?
18:13:27 <jtanner> would matrix be required in that case?
18:13:45 <gwmngilfen> that wouuld be my thought, then users on the irc side are likely authed already
18:13:45 <gwmngilfen> no. irc is not going away
18:13:53 <gwmngilfen> jtanner ^
18:13:56 <jtanner> k
18:14:21 <gwmngilfen> i have a long writeup on how it all looks from my view coming, i didnt quite finish it for today
18:14:22 <abadger1999> Alright, it's not a silent error (webchat as an anonymous user tells me i cannot send to the #ansible-jboss channel).  But it doesn't tell me what to do
18:14:47 <felixfontein> so how many people actually use irc via webchat (without being registered)?
18:14:57 <cyberpear> maybe something for the libera folks to improve?
18:15:15 <cyberpear> I know I've pointed folks there in the past for quick on-off discussions
18:15:43 <abadger1999> You can be anonymous in a regular client too
18:15:48 <cybette> I think we had a few in past contributor summits (using webchat) but I think the plan going forward is also to have matrix links instead of irc webchat links
18:16:16 <jillr> I'd venture a guess that the people most likely to use webchat are the folks least knowelagble about troubleshooting irc who arrive there from the docs
18:16:25 <briantist> I imagine irc via webchat (especially without registering) is the primary way that new users pop in, especially if they don't otherwise use irc
18:16:36 <briantist> yeah what jillr said
18:16:41 <cybette> yeah docs, ansible.com pages etc. contain most of the webchat links
18:16:57 <felixfontein> ok, so basically that means we shouldn't use +r or +q, except when strictly necessary?
18:17:22 <cyberpear> felixfontein: that's my feeling
18:17:24 <gwmngilfen> i'm good with that, i can document the needed commands for an op
18:17:40 <briantist> I think that's correct felixfontein
18:17:40 <gwmngilfen> its already in the linked hackmd buyt i'll make it clearer
18:17:46 <gwmngilfen> *but
18:18:03 <abadger1999> I've seen +r used with a redirect to a channel where the topic and people/a bit tell you to register
18:18:07 <felixfontein> is it possible to set +e already, so ops just need to add +q to make it work for matrix users?
18:18:39 <gwmngilfen> i dont think so, +e is an exempion and there's nothing to be exempt to :)
18:19:20 <gwmngilfen> but its just one /mode command for an op
18:20:30 <gwmngilfen> I think this is fine then. I had formed the impression that we were using +r more widely, and that's quite hostile on the matrix side. it seems we're not, so I'm good ;)
18:20:57 <felixfontein> should we vote on something, like that we do not want +qe or +re set by default?
18:21:00 <gwmngilfen> thanks for the feedback all
18:21:10 <cybette> cyb-clock chimes: 20 mintes into meeting, 15 min on IRC modes
18:21:35 <gwmngilfen> seems we're largely in agreement, but im good with a vote if desired
18:22:05 <felixfontein> if nobody explicitly wants a vote, I'll switch to the next topic in 1-2 minutes :)
18:22:55 <gundalow> We can always change the mode flags in the future if needed. IMHO no need to vote
18:23:09 <felixfontein> sounds good to me.
18:23:12 <felixfontein> #topic Improve ansible-collections issue templates
18:23:12 <felixfontein> #info Discussion: https://github.com/ansible-community/community-topics/issues/23
18:23:15 <felixfontein> #info PR: https://github.com/ansible-collections/.github/pull/4
18:23:15 <github-linkbot> https://github.com/ansible-collections/.github/pull/4 | open, created 2021-06-17T16:18:02Z by mariolenz: Add collection version to bug report template  
18:23:17 <felixfontein> Basically the question is: are we happy with the PR and want to merge it now, or does anyone want it changed?
18:23:42 <gundalow> LGTM
18:23:43 <felixfontein> (we already discussed this last week, today we can basically approve the PR or ask for changes)
18:23:45 <acozine> me too
18:23:49 <acozine> the example really helps
18:24:02 <gundalow> Will need the same in https://github.com/ansible-collections/community.general/tree/main/.github/ISSUE_TEMPLATE
18:24:05 <felixfontein> I also like the current version
18:24:06 <dmsimard> lgtm
18:24:15 <jillr> +1
18:24:17 <gundalow> (as community.general is using the new Issue Forms)
18:24:22 <felixfontein> VOTE: should we merge https://github.com/ansible-collections/.github/pull/4 as is?
18:24:25 <tadeboro> +1 for merge
18:24:27 <abadger1999> +1
18:24:29 <samccann> +1
18:24:32 <jillr> +1
18:24:32 <felixfontein> +1
18:24:37 <lmodemal> +1
18:24:40 <acozine> +1
18:24:41 <gundalow> +1
18:24:53 <andersson007_> +1
18:24:55 <cybette> +1
18:25:09 <felixfontein> #agreed merge https://github.com/ansible-collections/.github/pull/4
18:25:12 <felixfontein> good :)
18:25:22 <felixfontein> #topic Move shared ansible glue python core to a standalone python package
18:25:25 <felixfontein> #info Discussion: https://github.com/ansible-community/community-topics/issues/26
18:25:28 <felixfontein> zbr: ^
18:26:18 <felixfontein> the basic idea is to have a place for utility code which works with ansible
18:26:44 <felixfontein> I guess antsibull-changelog and antsibull could also offload some shared code there
18:27:00 <gundalow> cidrblock: ^ Are you looking at ansible-lint and runner things?
18:27:25 <gundalow> need to be careful we don't end up with a dumping ground library that's difficult to upgrade
18:27:34 <felixfontein> (can anyone with sufficient rights merge https://github.com/ansible-collections/.github/pull/4 ?)
18:28:04 <gundalow> Merged #4
18:28:06 * samccann afk for a few min
18:28:14 <felixfontein> gundalow: thanks!
18:28:27 <cyberpear> the proposal seems fine to me
18:29:35 <samccann> #unchair samccann
18:29:35 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gundalow gwmngilfen jillr lmodemal tadeboro
18:29:42 <samccann> (gonna be afk for longer than I thought)
18:30:03 <felixfontein> see you later samccann!
18:30:24 <andersson007_> the proposal seems fine to me to as well as the name of the repo
18:30:49 <felixfontein> I also like the proposal. I'm not 100% happy with the name, but I cannot come up with a better one :)
18:32:00 <abadger1999> <nod> i wish the content about ansible-runner wasn't true but her we can merge them later if those issues go away
18:32:12 <abadger1999> S/content/comment
18:33:48 <felixfontein> since ansible-runner does more than just being a library, I guess it's ok for it to have these dependencies. but I would avoid them for a library (especially python-daemon; pexpect might be ok depending for what it is needed)
18:34:40 * acozine thinks of names
18:34:58 <tadeboro> Last time I checked, ansible-runner was the thing that knows how to use EEs, and since those are what paying customers are interested in, I assumed that Red Hat will control it more tightly compared to molecule or linter.
18:35:23 <abadger1999> <nod>
18:35:44 <acozine> ansible-dynasty (because it works across various "reigns")? ansible-bridge (because it links different versions)?
18:36:30 <andersson007_> i like ansible-bridge but i'd keep this for a better case:)
18:36:50 <cybette> haha i like the sound of ansible-dynasty
18:36:53 <andersson007_> sounds like a cool project:)
18:36:54 <felixfontein> ansible-glue? :)
18:37:12 <andersson007_> felixfontein: this also came to my mind:)
18:37:23 <andersson007_> first
18:37:30 <cybette> naming is hard (but fun)
18:37:32 <felixfontein> (ansiglue? ;) )
18:37:42 <acozine> ansible-keyring (hm, probably too ssh-sounding)
18:37:54 <andersson007_> haha
18:38:07 <felixfontein> acozine: that sounds too much like key management
18:38:42 <acozine> ansible-fob? probably same objection
18:39:25 <andersson007_> (zbr: won't be happy)
18:39:58 <abadger1999> Ansiglue is cool but i don't want to cast a bike shed vote
18:40:07 <acozine> ansible-match?
18:40:10 <acozine> I can stop any time
18:40:13 <bcoca> aaa abstracting ansible api
18:40:31 <felixfontein> bcoca: already taken: https://pypi.org/project/aaa/
18:40:32 <acozine> ansible-link?
18:40:50 <acozine> too close to lint, probably
18:41:11 <bcoca> versionless-ansible
18:41:27 <bcoca> ansiwrap
18:41:27 <cybette> how about like zbr suggested, put the suggestions in the issue as comments and we can vote on them?
18:41:37 <bcoca> wrappable
18:42:45 <bcoca> agen, ansible generations
18:42:54 <cybette> *vote on them in the issue, not limited to this meeting
18:42:56 <felixfontein> do we want to use the `ansible-` prefix, or try to avoid it? (mainly to avoid confusion with potential future 'real' projects)
18:43:36 <felixfontein> (we also have the `antsibull-` prefix :) )
18:44:03 <acozine> heh, wrapsible
18:44:05 <felixfontein> acozine: it's better to have one name per comment, so it's possible to vote on them individually :)
18:44:10 <acozine> oops
18:44:14 <acozine> I can fix that
18:44:30 <cyberpear> do we have a good description of everything the project will contain? -- I like "glue" but will it be more than that?
18:45:01 * gundalow has to drop
18:45:03 <gundalow> #unchair gundalow
18:45:03 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gwmngilfen jillr lmodemal tadeboro
18:45:06 <gundalow> Thanks all
18:45:08 <acozine> bye gundalow!
18:45:10 <felixfontein> bye gundalow!
18:45:16 <cidrblock> sorry, stepped away..... If I understand, there's about 500 lines of code here common to molecule and ansible-lint that is being suggested moved to a new python package?
18:45:53 <felixfontein> cyberpear: I think it's mainly code that interfaces with some aspects of ansible that tries to be agnostic of the ansible version, and that's potentially useful to more projects
18:46:14 <cybette> cyb-clock chimes: 45 minutes into meeting, 20 min on ansible glue python code
18:46:30 <felixfontein> cidrblock: I don't know how much it is exactly, but I think that was the motivation. I think there's some more code in antsibull-changelog and antsibull which could also go there
18:47:33 <bcoca> gluable!
18:47:57 <cidrblock> without looking at the code and it's complexity, I cannot say one way or another if a new package makes sense here....  new deps come at a cost.. release cadence, versioning, testing etc.
18:48:51 <felixfontein> it's probably a good idea to have a `ansi*` or `antsi*` name, so it's clearer that the package belongs there (and is not a random other dependency that's easier to miss if you don't notice the ansible puns)
18:49:41 <felixfontein> andersson007_: already taken: https://pypi.org/project/agnostic/
18:50:00 <bcoca> ansible-atom
18:50:06 <andersson007_> felixfontein: ok, ansignostic
18:50:14 <bcoca> andersson007_:++
18:50:34 <felixfontein> feel free to add more ideas to the issue :)
18:50:39 <andersson007_> fixed in the issue:)
18:50:52 <felixfontein> should we vote on whether we think such a package is useful?
18:51:05 <bcoca> andersson007_: though .. ansi agnostic .. isnt that unicode?
18:51:08 <abadger1999> Sounds good.
18:51:25 <andersson007_> bcoca: I've no idea
18:51:29 <abadger1999> And then the rest can be done in the ticket
18:51:47 * lmodemal hard stop...see you all next time :)
18:51:58 <acozine> bye lmodemal!
18:51:59 <felixfontein> does someone thinks this needs to be discussed further before we vote?
18:52:02 <felixfontein> bye lmodemal!
18:52:10 <tadeboro> Well, at least molecule and ansible-lint are already sharing some code already, so the need is there.
18:52:12 <acozine> #unchair lmodemal
18:52:12 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gwmngilfen jillr tadeboro
18:52:51 <felixfontein> tadeboro: true, and making one project depending on the other for that shared code is something we should avoid IMO
18:53:21 <cidrblock> without reviewing the complexity of the code, understanding who and how it will be maintained, I am not in  a place to suggest this be done.
18:53:33 <felixfontein> VOTE: do we want a new Python library for ansible 'glue' code with as few dependencies as possible, and hopefully a stable interface?
18:54:01 <tadeboro> Yep, currently, molecule depends on ansible-lint, which is fun when we try to fix bugs in molecule that actually come from ansible-lint.
18:54:01 <cidrblock> --- (no vote)
18:54:02 <andersson007_> +1
18:54:22 <felixfontein> +1
18:54:22 <acozine> 0
18:54:25 <tadeboro> +1
18:54:28 <jillr> -- (abstain)
18:54:32 <abadger1999> +1
18:54:38 <cybette> 0
18:54:54 <felixfontein> cidrblock: I think that basically the maintainers of the projects which depend on the shared code will maintain it
18:55:08 <cidrblock> I missed that commitment
18:55:40 <felixfontein> i.e. the community ;)
18:55:41 <cidrblock> if that is true, then I will defer to the owners/maintainers of the projects that need it
18:56:37 <felixfontein> #chair
18:56:37 <zodbot> Current chairs: abadger1999 acozine andersson007_ briantist cidrblock cyberpear cybette dmsimard felixfontein gwmngilfen jillr tadeboro
18:57:24 <briantist> I will also defer to the (potential) maintainers. It sounds like a good idea to me, but I have little familiarity with the details.
18:59:46 <felixfontein> so we have 4 x +1, 2 x 0, 1 x abstain, 2 x defer to (potenial) maintainers
19:00:17 <cybette> cyb-clock chimes: 1 HOUR into meeting, 35 min on ansible glue python code
19:01:10 <felixfontein> #info we have 4 x +1, 2 x 0, 1 x abstain, 2 x defer to (potenial) maintainers
19:01:22 <acozine> nobody opposed it, so I think the motion passes
19:01:36 <felixfontein> ok, so please vote on the suggestions in the issue, so we have something to pick a name from :)
19:01:39 <felixfontein> #topic open floor
19:02:23 <cidrblock> hard stop here.  see you all later!
19:02:35 <abadger1999> Since cidrblock is here, I'd like to encourage all the steering committee members  who haven't voted yet (acozine cidrblock, cyberpear, zbr, thaumos), to vote on https://github.com/ansible-community/community-topics/issues/11
19:02:38 <abadger1999> cidrblock: ^
19:02:39 <felixfontein> do you have something spontaneous, or something that you just came up with?
19:03:24 <felixfontein> abadger1999: thanks for bringing that up!
19:04:23 <abadger1999> cidrblock (et al): also, if you aren't going to be at meetings, it would be great to let us know and also if we can treat your vote as a +0 unless we hear otherwise.  Lacking your vote holds up making decisions on close votes.
19:05:02 <abadger1999> felixfontein: I suppose at some point we'll have to discuss removing people from the steering committee as well, if they aren't showing up.
19:05:39 <jillr> should we add an todo to have another convo about how the steering committee operates?
19:06:01 <felixfontein> sounds like a good idea
19:06:03 * jillr acknowledges they've been checked out of most of this meeting as well  :(
19:06:25 <andersson007_> it should be detailed policy imo described in the doc
19:06:49 <andersson007_> we have a pretty abstract thing now, imo
19:06:51 <jillr> +1
19:07:14 <acozine> I'm +1 on adding the `disk_info` module, and we've already merged the `groupby_as_dict` filter (community-topics/issues/11)
19:07:30 <felixfontein> (and we have some docs on filters by now :) )
19:08:33 <abadger1999> acozine: Thanks :-)
19:09:07 <felixfontein> acozine: can you add that to the issue? so we don't forget about your vote :)
19:09:18 <acozine> yep
19:09:22 <felixfontein> thanks :)
19:09:38 <abadger1999> #info current vote tally on accepting disk_info ( https://github.com/ansible-community/community-topics/issues/11 ) is +1:4 (@abadger, @Andersson007, @tadeboro, @acozine), +0:2 (@jillr @felixfontein) -1:1 (@gundalow)
19:09:57 <cybette> "current responsibilities of Ansible Community Steering Committee members" in the blog post https://www.ansible.com/blog/ansible-community-steering-committee
19:11:34 <abadger1999> #info Steering committee announcement has committee member responsibilities listed: https://www.ansible.com/blog/ansible-community-steering-committee
19:11:51 <abadger1999> Thanks cybette !
19:11:53 <andersson007_> a more detailed version should be in the doc
19:12:02 <cybette> andersson007_: +1
19:12:30 <andersson007_> 1) how members are elected, how often, when 2) ...
19:12:47 <andersson007_> s/when/where/
19:13:25 * cybette and cyb-clock have to drop soon...
19:13:49 <andersson007_> 3) quorum
19:14:04 <andersson007_> that's it for now:)
19:14:42 <felixfontein> ompragash might also be interested in the above discussion
19:14:51 <felixfontein> #endmeeting