15:02:00 <zbyszek> #startmeeting FESCO (2019-06-07)
15:02:01 <zodbot> Meeting started Fri Jun  7 15:02:00 2019 UTC.
15:02:01 <zodbot> This meeting is logged and archived in a public location.
15:02:01 <zodbot> The chair is zbyszek. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:02:01 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:02:01 <zodbot> The meeting name has been set to 'fesco_(2019-06-07)'
15:02:01 <zbyszek> #meetingname fesco
15:02:01 <zbyszek> #chair nirik, bowlofeggs, jforbes, zbyszek, bookwar, sgallagh, contyk, mhroncok, otaylor
15:02:01 <zodbot> The meeting name has been set to 'fesco'
15:02:01 <zodbot> Current chairs: bookwar bowlofeggs contyk jforbes mhroncok nirik otaylor sgallagh zbyszek
15:02:03 <zbyszek> #topic init process
15:02:09 <mhroncok> hi
15:02:11 <jforbes> .hello2
15:02:12 <zodbot> jforbes: jforbes 'Justin M. Forbes' <jforbes@redhat.com>
15:02:12 <contyk> .hello psabata
15:02:15 <bcotton> .hello2
15:02:15 <zodbot> contyk: psabata 'Petr Šabata' <psabata@redhat.com>
15:02:15 <zbyszek> .hello2
15:02:18 <zodbot> bcotton: bcotton 'Ben Cotton' <bcotton@redhat.com>
15:02:21 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl>
15:02:24 <otaylor> .hello2
15:02:25 <zodbot> otaylor: otaylor 'Owen Taylor' <otaylor@redhat.com>
15:02:33 <mhroncok> .hello churchyard
15:02:34 <zodbot> mhroncok: churchyard 'Miro Hrončok' <mhroncok@redhat.com>
15:02:43 * zbyszek 's coffee machine broke. They are sad.
15:02:43 <bookwar> .hello2
15:02:44 <zodbot> bookwar: bookwar 'Aleksandra Fedorova' <alpha@bookwar.info>
15:03:06 <bcotton> RIP zbyszek's coffee machine
15:03:20 <bowlofeggs> mhroncok: heh, yeah
15:03:22 <bowlofeggs> .hello2
15:03:23 <zodbot> bowlofeggs: bowlofeggs 'Randy Barlow' <rbarlow@redhat.com>
15:03:43 <zbyszek> So we have quorum. Let's start.
15:03:54 <zbyszek> = Followups =
15:03:54 <zbyszek> #topic #2063 F31 Change: Ibus-typing-booster default for Indian languages
15:03:57 <zbyszek> .fesco 2063
15:03:59 <zodbot> zbyszek: Issue #2063: F31 Change: Ibus-typing-booster default for Indian languages - fesco - Pagure.io - https://pagure.io/fesco/issue/2063
15:03:59 <zbyszek> https://pagure.io/fesco/issue/2063
15:04:33 <contyk> +1 to reject
15:04:40 <jforbes> +1 to reject as well
15:04:45 <mhroncok> +1 to reject
15:04:53 <bowlofeggs> +1 reject
15:04:53 <zbyszek> +1 too (as in the ticket)
15:05:48 <zbyszek> #agree This change is rejected due to lack of response. Change owner is free to resubmit it again. (+5, 0, 0)
15:05:58 <zbyszek> #topic #2140 F31 System-Wide Change: RPM 4.15
15:05:58 <zbyszek> .fesco 2140
15:05:58 <zbyszek> https://pagure.io/fesco/issue/2140
15:06:00 <zodbot> zbyszek: Issue #2140: F31 System-Wide Change: RPM 4.15 - fesco - Pagure.io - https://pagure.io/fesco/issue/2140
15:06:14 <mhroncok> ah, an easy one
15:06:22 <mhroncok> python3-brainwash
15:06:28 <zbyszek> It involves Python and Unicode, so it's fun :)
15:06:39 <mhroncok> the change in rpm is good
15:06:57 <mhroncok> I'm not sure if the "depracation" steps were good and I said everyhting on the ticket
15:07:16 <ffesti> yeah, sorry for the mess
15:07:52 <zbyszek> I think the only way out is forward at this point...
15:08:09 <contyk> agreed
15:08:10 <mhroncok> I'd be much happier if we had a backward compatible way of using the python bindings
15:10:09 <zbyszek> mhroncok: I think it's too late for that, because some of the projects have been update to expect the new incompatible API
15:10:27 <mhroncok> zbyszek: the new api should remain as is
15:10:38 <mhroncok> I just want the old API to have a flag to opt for this
15:11:24 <zbyszek> But isn't there just one API (no "old" and "new"), that changed the return value? A I missing something?
15:11:33 <mhroncok> that's a different API
15:11:58 <mhroncok> what is returned is an essential part of an API
15:12:44 <mhroncok> brbr, sorry, I got a package delivery
15:12:44 <ffesti> the issue is that the old API doesn't make any sense
15:12:52 <smooge> I am a bit confused.. the topic says rpm-4.15 but we seem to be talking about python3-brainwash
15:13:50 <ffesti> yes, technically these are separate issues
15:14:05 <ffesti> as we changed the API in the 4.14 package already
15:14:11 * mhroncok is back
15:14:52 <zbyszek> ... and 4.14 is in F30 already ...
15:15:05 <mhroncok> this was rawhide only AFAIK
15:15:11 <jforbes> I voted in ticket, and my +1 doesn't change
15:15:26 <bowlofeggs> mhroncok: was the package another cup of coffee to replace the one bcotton took?
15:16:21 <mhroncok> I've said everything I needed to say in the ticket. if FESCo is OK with this, we should vote and proceed
15:16:52 <contyk> I'm +1
15:17:11 <zbyszek> +1 too, as in the ticket
15:17:20 <mhroncok> I'm 0
15:18:14 <zbyszek> nirik was +1 in the ticket
15:18:16 <otaylor> +1
15:19:14 <otaylor> (I don't think we should be nit-picking the details of the Python bindings at the fesco level - cleaning up from earlier api messes is always hard and there is seldom a perfect way forwrad)
15:19:56 <bowlofeggs> i guess since the API break is already in rawhide, i'll be +1 too
15:20:00 <mhroncok> what I'm saying is that this will break the packager stack
15:20:18 <mhroncok> rawhide has hacks that make rpmlint work
15:20:48 <bcotton> "i guess since the API break is already in rawhide, i'll be +1 too" these votes always make me the saddest
15:21:04 <zbyszek> Yeah, but the change is about rawhide, so the breakage is in rawhide too?
15:21:10 <zbyszek> bookwar: vote?
15:21:46 <mhroncok> zbyszek: rawhide is half broken, this makes it broken
15:22:08 <bowlofeggs> bcotton: heh, yeah… ☹
15:22:22 <bowlofeggs> i also wish they were following semver
15:22:36 <bowlofeggs> 4.14 → 4.15 should not have an API break
15:23:18 <bookwar> zbyszek: +1 on change
15:23:30 <bookwar> -5 on how we handled it :)
15:23:51 <zbyszek> #agreed Change is accepted (+6, 1, -0)
15:24:09 <zbyszek> mhroncok: let's work on unbreaking rawhide
15:24:21 <mhroncok> #action mhroncok to put a I've told you so note in the top drawer
15:24:44 <zbyszek> #topic #2139 F30: change default stream for avocado module
15:24:44 <zbyszek> .fesco 2139
15:24:44 <zbyszek> https://pagure.io/fesco/issue/2139
15:24:45 <zodbot> zbyszek: Issue #2139: F30: change default stream for avocado module - fesco - Pagure.io - https://pagure.io/fesco/issue/2139
15:24:52 <mhroncok> zbyszek: yes, I'm trying, but there is no sane way of doing it without breaking the code on older fedoras
15:25:28 <mhroncok> ad avocado: I voted +1 because I didn't know this cannot be done
15:25:39 <mhroncok> is there actually anyhing to vote on, since it is impossible?
15:25:51 <bowlofeggs> i was -1 in the ticket due to the stable users not being able to follow it
15:25:53 <jforbes> I was under the impression that this request was withdrawn because of this
15:26:23 <contyk> me too
15:26:26 <contyk> I think we can close this
15:26:49 <zbyszek> It was withdrawn for F30 certainly.
15:26:57 <contyk> but -1 if we need a vote, for the same technical and procedural reasons
15:27:06 <contyk> well, it was about f30
15:27:15 <zbyszek> Right, I misunderstood.
15:27:29 <zbyszek> Nothing to see here, let's move along.
15:28:17 <zbyszek> #info the request has already been withdrawn
15:28:22 <zbyszek> #topic #2135 set default module stream for cri-o
15:28:23 <zbyszek> .fesco 2135
15:28:23 <zbyszek> https://pagure.io/fesco/issue/2135
15:28:24 <zodbot> zbyszek: Issue #2135: set default module stream for cri-o - fesco - Pagure.io - https://pagure.io/fesco/issue/2135
15:29:23 <jforbes> Sorry, I thought I voted in the ticket on this one, but seemingly not
15:29:26 <mhroncok> "best to hold off on this for now"
15:29:28 <zbyszek> mhroncok and bowlofeggs voted +1 in the ticket
15:29:36 <otaylor> this seems to be on hold waiting for some idea of how to coordinate this with k8s
15:30:03 <jforbes> And that was why.
15:31:24 <contyk> I don't think that's a blocker for this
15:31:56 <contyk> it's a potential follow-up work but I think I can give +1 to this
15:32:10 <jforbes> contyk: I think when the requester says "best to hold off on this for now" it is specifically a blocker for us moving forward
15:32:21 <contyk> having a default stream means you get *something* when you try dnf install cri-o, rather than an error
15:32:36 <contyk> but okay, up to them
15:32:49 <mhroncok> jforbes: well I think we can still give the maintainer the right to do so, even if they decide not to
15:33:48 <jforbes> mhroncok: just as likely they might decide to do something else though.
15:34:11 <mhroncok> ok. let's move on?
15:34:23 <jforbes> seems the right course
15:35:02 <zbyszek> OK, let's move on.
15:35:46 <zbyszek> #topic #2136 F32 Self-Contained Change: Track Changes in Taiga
15:35:46 <zbyszek> .fesco 2136
15:35:47 <zbyszek> https://pagure.io/fesco/issue/2136
15:35:48 <zodbot> zbyszek: Issue #2136: F32 Self-Contained Change: Track Changes in Taiga - fesco - Pagure.io - https://pagure.io/fesco/issue/2136
15:35:55 <zbyszek> Another controversial one.
15:36:34 <bcotton> everyone loves it except mhroncok ;-)
15:37:01 <mhroncok> bcotton: oh I love it personally, except I'm afraid that it's another bump on the packaging road
15:37:17 <bcotton> mhroncok: i understand the concern
15:37:36 <mhroncok> everybody seems to only see the program manager's pov
15:37:41 <bcotton> what i haven't seen in the community feedback is specific problems other than "i like the wiki and it works"
15:38:37 <zbyszek> Hmm, thinking about it some more, I think that some of the concerns might be valid.
15:39:03 <bowlofeggs> i'm +1
15:39:08 <bookwar> can we maybe ask directly owners of 5 or ten last changes to provide explicit feedback?
15:39:18 <bcotton> i have a bias toward that point of view because it's mine, but i have tried to make it minimally impactful to change owners, who will generally interact with the process a lot less
15:39:22 <otaylor> I feel there's a *general* move away from the wiki, so I'm not sure mediawiki competency is a core Fedora skill these days
15:39:27 <mhroncok> can we get a demo for the change owners to try?
15:39:30 <bcotton> otaylor++
15:39:31 <zodbot> bcotton: Karma for otaylor changed to 2 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:39:34 <zbyszek> We're essentially "buying a cat in a bag", because the implementation is yet to be written.
15:39:36 <contyk> otaylor +1
15:39:47 <contyk> I'm +1, it's just a different tool
15:40:02 <bowlofeggs> zbyszek: heh, i haven't heard that idiom before
15:40:16 <smooge> its a pig in a poke with claws
15:40:20 <bcotton> zbyszek: yes and no. we could use taiga without the tool that's being written, but it's definitely easier with the tool
15:40:25 <zbyszek> bowlofeggs: it's a medieval thing, about buying a pig and returning to the farm with a cat
15:40:26 <mhroncok> Czechs buy a rabbit in a bag
15:40:29 <bcotton> zbyszek++ for "cat in a bag"
15:40:29 <zodbot> bcotton: Karma for zbyszek changed to 1 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:40:31 <bowlofeggs> imo, this doesn't have to be a permanent change - we can try it for f31 and see how it goes?
15:40:37 <bowlofeggs> the wiki isn't going away any time soon
15:41:05 <bcotton> bowlofeggs: s/1/2/ but that's a good point. if this is awful, we can either fix it or go back to the wiki. there's a very easy reversion available
15:41:07 <mhroncok> bowlofeggs: who decides how it goes?
15:41:22 <bowlofeggs> mhroncok: fesco, i suppose?
15:41:25 <bcotton> (unless the wiki shuts off between now and then, which is an entirely different problem we'll have to tackle)
15:41:44 <bowlofeggs> i haven't heard anyone suggest turning off the wiki - it's still used for a lot of things
15:42:07 <mhroncok> sounds fair. we test it on f32, once that is over, we reevaluate it on fesco level and the contindency is to put everything back on the wiki
15:42:22 <bcotton> yeah, i think the wiki will continue to exist for a long time, but the council in particular and the community generally seems to be pushing toward its use as a scratch space, not for permanent information
15:42:24 <mhroncok> *contingency
15:43:02 <bowlofeggs> i honestly think taiga will work nicely for this
15:43:14 <mhroncok> I hope so
15:43:48 <zbyszek> proposal: We ask for the new tooling to be implemented, and once it's functional, we'll evaluate whether it's ready to be used for the F32 changeset.
15:43:52 <bowlofeggs> plus, if it doesn't, bcotton owes each of us $5
15:44:02 <bookwar> i honestly would prefer to have pull requests into git repo with markdown, but i guess this is not an option nowdays
15:44:03 <bowlofeggs> (that's in the tiny print that nobody reads!)
15:44:14 <bookwar> so let's try taiga
15:44:25 <otaylor> bcotton: is there taiga work to be done to prepare for this? in particular, is it be possible given current setup to have a project in taiga that any Fedora contributor can easily enter an issue into?
15:44:50 <bcotton> not asking for discussion now, but... i designed this change with the idea that it's changing no policies, just changing how we implement them. but it's also a good time for fesco to think about if it *wants* some policy changes
15:45:09 <bowlofeggs> zbyszek: +1, though i would also +1 a more lenient proposal that just says "the proposal is approved"
15:45:32 <bcotton> bookwar: there are some advantages to that idea, but it also has some drawbacks. happy to have that conversation later if you'd like :-)
15:46:09 <mhroncok> zbyszek: +1
15:46:12 <bcotton> otaylor: it just requires some ACL changes on the project
15:46:39 <zbyszek> bowlofeggs: we generally have been wary of accepting proposals for solutions that are yet to be written. In this case, pac23 is a new contributor, and it's quite likely that they underestimate the coding work required.
15:46:57 <bcotton> zbyszek: pac23 has previous fedora contribution experience
15:47:04 <zbyszek> I'd very much prefer to have an explicit sign-off on the new tooling before we put it into production.
15:47:14 <bcotton> zbyszek: but i would also argue that fesco should not be voting based on the tooling
15:47:35 <jforbes> bcotton: I agree with that as well
15:47:39 <otaylor> bcotton: OK - adding people to the silverblue project has been really confusing/hard, but maybe that's because we've been going through an invitation code path
15:47:42 <mhroncok> bcotton: and I woild dissagree
15:47:43 <jforbes> my +1 in ticket stands for a general approval
15:47:45 <mhroncok> *would
15:47:50 <bcotton> the tooling isn't fesco's concern, really, because i'm mostly the one impacted by the tool. the question for fesco is whether or not using taiga is appropriate
15:47:58 <mhroncok> if the tooling is horrible, we don't want our contributors to be forced to use it
15:47:59 <bowlofeggs> yeah i personally think it's ok for us to approve the general concept
15:48:15 <bowlofeggs> we can file new tickets later if there are problems with implementation
15:48:20 <mhroncok> bcotton: so the tool is only for you?
15:48:21 <bcotton> mhroncok: the contributors won't necessarily be using the script pac23 is writing
15:48:31 <mhroncok> necessarily?
15:48:50 <bcotton> mhroncok: the goal is to have a contributor-facing part of the tool which will allow CLI-based submission of changes to taiga
15:49:03 <bcotton> but the core functionality of the tool is in the copy/paste work that i do to manage changes now
15:49:21 <zbyszek> bcotton: but how would that work? A change is many paragraphs of text... Are we suppposed to provide that on the commadn line?
15:49:28 <bcotton> i guess i haven't made that clear enough
15:49:54 <bcotton> zbyszek: it will be interactive with a call to $EDITOR for the big blobs of text
15:50:07 <mhroncok> bcotton: can we sid down with vondruch and you next week and try to file an example change in taiga?
15:50:20 <bcotton> mhroncok: sure
15:50:25 <mhroncok> I think having something specific would help here
15:50:47 <mhroncok> bcotton: can we postpone the decision for next meeting?
15:50:48 <bcotton> zbyszek: but there will be no *requirement* for the contributors to use the tool
15:50:49 <otaylor> bcotton: But you will be able to file a change through taiga, right? the cli thjing is optional?
15:51:06 <bcotton> otaylor: exactly. i expect most people will use the taiga web interface. the cli is just a nice-to-have
15:51:19 <bcotton> mhroncok: i'm okay with postponement
15:51:43 <zbyszek> bcotton: do you have an example issue?
15:53:04 <bcotton> #link https://teams.fedoraproject.org/project/bcotton-test-changes-tracker/issue/3
15:53:43 <bcotton> ^^ is an example issue
15:53:56 <jforbes> If we want to postpone that is fine, but I am counting plenty of +1s in ticket. Are you expecting to change people's votes with the postponement?
15:54:07 <bcotton> #link https://teams.fedoraproject.org/project/bcotton-test-changes-tracker/us/1?kanban-status=29
15:54:20 <zbyszek> I see there's history... which is good.
15:54:21 <bcotton> ^^ example user story (basically after a change is announced)
15:54:26 <mhroncok> This Issue has been promoted to US:#4 Switch to Taiga for Change proposals
15:54:30 <mhroncok> what does that mean?
15:55:19 <zbyszek> But I also see that the history is wrong :( I changed systemd-wide-change from false to true to false, and "6 activities" only captured the second step :(
15:55:20 <bcotton> mhroncok: the way it works is that the contributor would submit an issue and it gets converted to a user story when it gets announced to the mailing list. this makes it easier to do back and forth if the change isn't complete (e.g. doesn't have a check with releng first)
15:56:26 <zbyszek> Never mind, it probably was "true" in the beginning.
15:56:28 <bcotton> zbyszek: race condition, maybe? you probably set it to true right after i did
15:57:09 <mhroncok> Oops, something happened...
15:57:09 <mhroncok> Our Oompa Loompas are sad, your changes were not saved!
15:57:46 <mhroncok> from the first glance, it appears very broken
15:57:57 <mhroncok> migth be the demo effect
15:58:11 <contyk> it is
15:58:15 <contyk> Taiga generally works alright
15:58:15 <bcotton> but to echo jforbes's question: is the expectation that a postponement might change some +1s to -1s or is the expectation that it would make the decision unanimous
15:59:14 <bcotton> yeah, and taiga has a longer login timeout, which means you don't get bitten by the "you got distracted while editing a wiki page and hahah i didn't actually save your edits" which i have encountered more times than i am willing to admit :-)
15:59:20 <jforbes> Side note, we waste a *lot* of time trying to make decisions unanimous. There is no need for FESCo decisions to be unanimous, and I don't think most people care if they are or not
15:59:49 <mhroncok> I'm trying to play this safe. if you'd rather vote me over, go
16:00:36 <mhroncok> I think it is a good idea to sit down and see how would a change be filed before we approve this
16:00:37 <zbyszek> I'd be much happier to have mhroncok and bcotton and vondruch go over an issue to test this.
16:01:28 * mhroncok got an email: You've been Taigatized!
16:01:52 <bookwar> zbyszek: +1
16:02:10 <bookwar> we can overvote it, but we rather address the feedback we got
16:02:34 <zbyszek> OK, so we have at least three votes for waiting.
16:02:37 <mhroncok> bcotton: one thing that bothers me: the US is immutable for me
16:02:38 <bowlofeggs> i'm ok with further inspection, and i'm also ok with just approving the change ☺
16:02:50 <mhroncok> now normally when we receive feedback on changes, we adapt
16:04:27 <otaylor> I'm +1 for just approving, but people would rather discuss in more detail with bcotton, then let's move on, and revisit in a few weeks
16:04:33 <bcotton> mhroncok: that's intentional, but also changeable. i can either open up the permissions or make it so that the conversion to a user story happens *after* fesco approves it
16:04:48 <bookwar> let's not delay it to much, if you can organize this try-out next week - let's wait, if it is not done by next friday - let's vote as it is
16:05:05 <mhroncok> Yes, we can do this next week.
16:05:13 <zbyszek> bookwar: +1
16:05:20 <mhroncok> if vondruch cannot, we'll do without him
16:05:33 <mhroncok> assuming bcotton can
16:06:03 <zbyszek> action bcotton to test the issue filing process with mhroncok and vondruch.
16:06:15 <mhroncok> zbyszek:  #
16:06:15 <bcotton> mhroncok: for sure. by EOD monday i'll make sure it's ready for a trial run. it doesn't necessarily have to be interactive (and maybe it's better if it *isn't*)
16:06:21 <zbyszek> #action bcotton to test the issue filing process with mhroncok and vondruch.
16:06:43 <zbyszek> #info We'll revisit next week.
16:06:52 <zbyszek> OK, let's move on.
16:06:53 <zbyszek> #topic Next week's chair
16:07:09 <contyk> I can do next week
16:07:17 <zbyszek> action contyk will chair next meeting
16:07:22 <zbyszek> #action contyk will chair next meeting
16:07:30 <zbyszek> Great, thanks.
16:07:36 <zbyszek> #topic Open Floor
16:08:20 <zbyszek> Anyone? If not, I'll close in a minute…
16:08:26 <mhroncok> don't forget to vote I guess
16:08:47 <contyk> already did
16:08:50 <bcotton> #link http://elections.fedoraproject.org/
16:09:04 <zbyszek> #info Voting closes in 13 days
16:10:07 <zbyszek> OK. Thanks you for attending.
16:10:09 <zbyszek> #endmeeting