ansible_core_public_irc_meeting
LOGS
15:03:50 <sdoran> #startmeeting Ansible Core Public IRC Meeting
15:03:50 <zodbot> Meeting started Thu Feb 25 15:03:50 2021 UTC.
15:03:50 <zodbot> This meeting is logged and archived in a public location.
15:03:50 <zodbot> The chair is sdoran. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:03:50 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:03:50 <zodbot> The meeting name has been set to 'ansible_core_public_irc_meeting'
15:03:57 <sdoran> Hello hello.
15:04:02 <sdoran> Nothing on the agenda for today.
15:06:25 <sdoran> #info At the last meeting, changes to the schedule were proposed. Based on votes, we will be changing the schedule to once per week, alternating Tuesday and Thursday. The meeting time will remain the same.
15:06:37 <felixfontein> hi!
15:06:43 <sdoran> 👋
15:07:56 <felixfontein> hmm, looks like nobody but meeting regulars actually voted
15:08:10 <sdoran> Darn confirmation bias.
15:08:29 <bcoca> felixfontein: kind of exemplifies why we think the meetings might need to be cut down if those are only people showing up
15:10:02 <felixfontein> indeed.... other people don't seem interested until they actually are for a specific topic, and then they're not anymore...
15:10:33 <bcoca> or until after the fact and then complain they were not able to give feedback at the time they were ignore it
15:11:09 <felixfontein> btw, bcoca, I created a PR for some more ansible-doc (and some code it indirectly calls): https://github.com/ansible/ansible/pull/73601 - are you the right person to review this, or whom should I bug?
15:12:33 <felixfontein> and I have a PR for improving the error messages when redirects are broken: https://github.com/ansible/ansible/pull/73603 - though I think a better global solution is needed, than to fix the error messages for every plugin type individually
15:13:53 <bcoca> both are on my list, i'm still probably best for reviews on ansible-doc
15:14:16 <bcoca> for the redirect issue, ping @nitzmahone
15:14:27 <felixfontein> ok, thanks, will do!
15:14:35 <bcoca> i can also take a look .. if you have the patience, he is also probalby very busy
15:15:26 <bcoca> just a gues (have not read it) you want to show the 'attempted names' .. not just one used?
15:15:41 <bcoca> ^ also on my list to add that functionality
15:15:57 <felixfontein> sure, feel free :) that PR only improves the messages for some specific situations
15:16:40 <felixfontein> you mean for modules/actions in playbooks?
15:19:06 <bcoca> plugins, they all use same redirected loader logic
15:20:29 <felixfontein> well, not all actually :)
15:20:39 <felixfontein> test and filter plugins do something different apparently
15:20:58 <briantist> I am lurking (usually), admit I only started joining core meetings due to having something on the agenda once. But I abstained from voting because even though I usually lurk, the meetings are mostly for others and I don't add much,  doesn't make sense for me to influence the schedule :)
15:24:56 <bcoca> felixfontein: not really, see the specific loader class, it ends up using return super
15:25:10 <bcoca> felixfontein: i just 'fixed' some corner cases on that code in PR
15:25:21 <bcoca> https://github.com/ansible/ansible/pull/73714
15:26:20 <bcoca> briantist: mostly schedule is for people like you that want to discuss an issue in rare moments .. the thing is lately we have had none to discuss in many meetings
15:28:01 <bcoca> this also consumes a lot of core team effort to clear time for these meetings .. and we all have long lists to do , so adding back 1h/week for soemthing else  we can always use, but we dont want to sacrafice our availability to the community for important discussions for that 1h/week
15:28:01 <sdoran> Since we have no official topics for today, I'll go ahead and wrap up the meeting in five minutes.
15:28:01 <sivel> anyone want to rehash anything about hash_behaviour?
15:28:01 <sivel> :troll:
15:28:01 <bcoca> just trying to gauge how much does the community still need this 1h/week
15:28:01 * sdoran bans sivel
15:28:10 * bcoca smacks sivel with tuna fish
15:28:26 * bcoca hides '-' in group names
15:28:27 <sdoran> Discussion about PRs can continue in #ansible-devel (and is welcome there any time).
15:28:35 <briantist> makes sense.. if I had a strong opinion I'd share it, but the burden on the team members is exactly why I prefer to let you all decide, and I'm good to go with what works best
15:28:52 <bcoca> ^ mostly PRs shoudl be discussed there, this meeting should be for bigger design concerns and project decisions
15:29:06 <sdoran> briantist: The main thing is we want to provide times that give everyone an opportunity to attend at least one meeting.
15:29:12 <sdoran> So all input is helpful.
15:29:58 <bcoca> sadly i suspect that most peop;le that will want to have input on this decision wont until they try to get to a meeting and only the previous schedule worked for them (also have biweeklies like new proposed)
15:30:45 <briantist> sdoran understood thank you! I think I've seen relevant discussion on (common) timezones and all that, so I have nothing else to add; and for me personally, I tend to be pretty flexible on times.
15:30:56 <sdoran> 👍
15:31:08 <felixfontein> if you want to discuss PRs, I always have some ;)
15:31:33 <felixfontein> how about: https://github.com/ansible/ansible/pull/73508
15:31:41 <felixfontein> validate-modules: make sure that options that potentially contain secret data have no_log set
15:32:10 <bcoca> felixfontein: i have 60
15:32:23 <felixfontein> bcoca: you have commit rights ;)
15:32:35 <bcoca> not for my own
15:32:39 <sdoran> Right.
15:32:41 <bcoca> we require review
15:32:53 <sdoran> We tend to smack teammates with tuna for merging their own PRs. ;)
15:32:55 <felixfontein> this one actually might need some discussion, since it will produce a lot of false positives
15:33:20 * felixfontein doesn't like fish, so this sounds pretty scary ;)
15:33:27 <bcoca> felixfontein: same reason 'key' was not added to original  (we did but then realized 100% of existing modules was false positive)
15:33:47 <bcoca> felixfontein: check the size of tuna, its pretty big impact
15:34:04 <felixfontein> bcoca: this being a sanity check has the advantage that false positives won't flood end-users, only maintainers / module authors
15:34:35 <bcoca> why i also want to remove runtime at one point .. but still 99% of time its false positive
15:34:52 <bcoca> now that we have at least 2 modules that do use '*key*' for secrets
15:35:08 <bcoca> but as you said, its less pitchfork inducing on author side
15:37:04 <sdoran> Alright, going to wrap up now. Thank you everyone for attending.
15:37:06 <felixfontein> btw, feature freeze for devel branch was until ~end of march, or something like that?
15:37:06 <sdoran> #endmeeting