15:00:45 <zbyszek> #startmeeting FESCO (2019-03-04)
15:00:45 <zodbot> Meeting started Mon Mar  4 15:00:45 2019 UTC.
15:00:45 <zodbot> This meeting is logged and archived in a public location.
15:00:45 <zodbot> The chair is zbyszek. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:45 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:45 <zodbot> The meeting name has been set to 'fesco_(2019-03-04)'
15:00:45 <zbyszek> #meetingname fesco
15:00:45 <zbyszek> #chair nirik, bowlofeggs, jforbes, zbyszek, tyll, sgallagh, contyk, mhroncok, otaylor
15:00:45 <zodbot> The meeting name has been set to 'fesco'
15:00:45 <zodbot> Current chairs: bowlofeggs contyk jforbes mhroncok nirik otaylor sgallagh tyll zbyszek
15:00:48 <zbyszek> #topic init process
15:00:50 <jforbes> .heelo2
15:00:52 <zbyszek> .hello2
15:00:53 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl>
15:00:54 <contyk> .hello psabata
15:00:55 <jforbes> .hello2
15:00:56 <mhroncok> hey
15:00:57 <otaylor> .hello2
15:00:57 <zodbot> contyk: psabata 'Petr Šabata' <psabata@redhat.com>
15:00:58 <nirik> morning everyone
15:01:00 <zodbot> jforbes: jforbes 'Justin M. Forbes' <jforbes@redhat.com>
15:01:03 <zodbot> otaylor: otaylor 'Owen Taylor' <otaylor@redhat.com>
15:01:08 <bookwar> .hello
15:01:09 <zodbot> bookwar: (hello <an alias, 1 argument>) -- Alias for "hellomynameis $1".
15:01:13 <bookwar> wow
15:01:34 <bowlofeggs> .hello2
15:01:38 <zodbot> bowlofeggs: bowlofeggs 'Randy Barlow' <rbarlow@redhat.com>
15:01:39 <jforbes> That seems like a pretty accurate description of Monday morning
15:01:50 <bcotton> .hello2
15:01:51 <zodbot> bcotton: bcotton 'Ben Cotton' <bcotton@redhat.com>
15:02:03 <contyk> :)
15:03:32 <zbyszek> So, if I count correctly, we have 4 members present. I'm not counting bookwar for now.
15:03:58 <zbyszek> mhroncok: meeting?
15:04:13 <zbyszek> Oh, you said 'hey', nvm.
15:04:23 <zbyszek> OK, so we do have quorum.
15:04:27 <jforbes> zbyszek: Well, if there are enough votes in ticket, you could
15:04:40 <zbyszek> #2098 Fill Till's FESCo seat
15:04:40 <zbyszek> .fesco 2098
15:04:40 <zbyszek> https://pagure.io/fesco/issue/2098
15:04:42 <zodbot> zbyszek: Issue #2098: Fill Till's FESCo seat - fesco - Pagure.io - https://pagure.io/fesco/issue/2098
15:04:56 <zbyszek> jforbes: I wanted to stick to the letter ;)
15:05:13 <zbyszek> We're +6 in the ticket
15:05:29 <zbyszek> +7 now.
15:05:31 <bowlofeggs> i'm +1 as well (and just added to the ticket)
15:05:54 <mhroncok> for the record, full consensus is needed
15:06:02 <mhroncok> that's 8 pluses
15:06:34 <zbyszek> sgallagh didn't vote.
15:06:56 <zbyszek> Can I treat this as approved?
15:06:56 <contyk> sgallagh is on the go now but might be here in an hour
15:07:16 <sgallagh> +1
15:07:25 <zbyszek> Ah, OK, thanks sgallagh.
15:07:39 <sgallagh> On the road. Will not be responsive.
15:08:14 <zbyszek> APPROVED: bookwar is voted-in to take over tyll's seat. (+8, 0, 0)
15:08:31 <tyll_> bookwar: congrats!
15:08:38 <bookwar> thanks tyll :)
15:08:47 <jforbes> Welcome aboard bookwar
15:08:59 <zbyszek> bookwar: please do either '.hello2' or '.hello <fas-name>'
15:09:04 <bowlofeggs> congrats (and condolences :))
15:09:07 <bookwar> .hello2
15:09:08 <zodbot> bookwar: bookwar 'Aleksandra Fedorova' <alpha@bookwar.info>
15:09:16 <nirik> welcome bookwar
15:09:28 <bowlofeggs> who can add bookwar to the relevant mailing list and FAS groups?
15:09:30 <nirik> and many thanks to tyll
15:09:32 <contyk> bookwar: finally :)
15:09:37 <bowlofeggs> tyll++, we will miss you
15:09:38 <mhroncok> :)
15:09:40 <contyk> tyll++
15:09:42 <zodbot> contyk: Karma for till changed to 3 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:09:49 <zbyszek> tyll++
15:09:50 <zodbot> zbyszek: Karma for till changed to 4 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:09:55 <tyll> thank you all!
15:09:57 <nirik> there's not a fas group, but I can adjust the mailing list and pagure group...
15:10:04 <jforbes> thanks tyll
15:10:12 <contyk> don't forget the badge
15:10:28 <zbyszek> OK, now on to normal business
15:10:29 <zbyszek> = Followups =
15:10:34 <zbyszek> #2091 Add check for co-maintainers to orphan procedure
15:10:34 <zbyszek> .fesco 2091
15:10:34 <zbyszek> https://pagure.io/fesco/issue/2091
15:10:36 <zodbot> zbyszek: Issue #2091: Add check for co-maintainers to orphan procedure - fesco - Pagure.io - https://pagure.io/fesco/issue/2091
15:10:55 <zbyszek> I don't know if jwrgoede is around...
15:11:39 <zbyszek> I made a soft proposal in https://pagure.io/fesco/issue/2091#comment-554301, would everyone be OK with that?
15:11:53 <zbyszek> > add a recommendation to ask the co-maintainers or on fedora-devel before orphaning, iff the owner has time to do this. Otherwise, simply orphan as currently described. This should allow people to perform orphaning without overhead if they don't have time, but still avoid releng involvement in many cases.
15:12:35 <mhroncok> +1 to the recommendation
15:12:52 <otaylor> +1
15:12:57 <jforbes> I can +1 that
15:13:02 <contyk> +1
15:13:09 <bowlofeggs> zbyszek: +1
15:13:45 <nirik> +1
15:14:06 <nirik> but we don't know yet if that satisfies the requestor?
15:14:41 <zbyszek> nirik: we don't, but we don't have an answer. I reached out in the announcement.
15:14:56 <zbyszek> I think we can approve this, and reopen the ticket later if necessary.
15:14:59 <contyk> we could also make one of the comaintainers the primary one
15:15:02 <contyk> randomly
15:15:16 <mhroncok> no
15:15:25 <mhroncok> (not sure if you are serious)
15:15:32 <zbyszek> contyk: people don't like that because it turns out the co-maintainers are often mia
15:15:53 <contyk> mhroncok: kinda am but zbyszek has a point :)
15:16:16 <zbyszek> APPROVED: A recommendation to ask the co-maintainers or on
15:16:16 <zbyszek> fedora-devel before orphaning, iff the owner has time to do
15:16:18 <zbyszek> this is to be added. (+7, 0, 0)
15:16:43 <zbyszek> Oh, I think I missed the hash, let me try again.
15:16:55 <tyll> the advantage of assigning co-maintainers is that the list of comaintainers will be cleaned up eventually
15:17:01 <zbyszek> #APPROVED: A recommendation to ask the co-maintainers or on fedora-devel before orphaning, iff the owner has time to do this is to be added. (+7, 0, 0)
15:17:02 <bcotton> zbyszek: point of order before we get too much further in. please use #topic etc so the minutes come out looking right
15:17:32 <zbyszek> Argh.
15:17:55 <contyk> zbyszek: try #agreed
15:17:57 <bcotton> (also, iirc #approved isn't a command and you have to do #agreed approved)
15:18:08 <zbyszek> OK, let me fix this.
15:18:35 <mhroncok> tyll: after several nonresponsive maintiners yet
15:18:46 <mhroncok> tyll: but even one of them is tedious
15:18:51 <zbyszek> #topic #2098 Fill Till's FESCo seat
15:18:51 <zbyszek> .fesco 2098
15:18:51 <zbyszek> https://pagure.io/fesco/issue/2098
15:18:53 <zodbot> zbyszek: Issue #2098: Fill Till's FESCo seat - fesco - Pagure.io - https://pagure.io/fesco/issue/2098
15:18:56 <zbyszek> #agreed APPROVED: bookwar is voted-in to take over tyll's seat. (+8, 0, 0)
15:19:01 <zbyszek> #topic #2091 Add check for co-maintainers to orphan procedure
15:19:01 <zbyszek> .fesco 2091
15:19:02 <zbyszek> https://pagure.io/fesco/issue/2091
15:19:03 <zodbot> zbyszek: Issue #2091: Add check for co-maintainers to orphan procedure - fesco - Pagure.io - https://pagure.io/fesco/issue/2091
15:19:05 <zbyszek> #agreed APPROVED: A recommendation to ask the co-maintainers or on fedora-devel before orphaning, iff the owner has time to do this is \
15:19:08 <zbyszek> to be added. (+7, 0, 0)
15:19:20 <zbyszek> #undo
15:19:20 <zodbot> Removing item from minutes: AGREED by zbyszek at 15:19:05 : APPROVED: A recommendation to ask the co-maintainers or on fedora-devel before orphaning, iff the owner has time to do this is \
15:19:35 <contyk> Monday
15:19:35 <zbyszek> #agreed APPROVED: A recommendation to ask the co-maintainers or on fedora-devel before orphaning, iff the owner has time to do this is to be added. (+7, 0, 0)
15:19:43 <zbyszek> Apologies.
15:19:50 <bowlofeggs> ha
15:19:57 <zbyszek> #topic #2096 F31 System-Wide Change: BuildRequires generators
15:19:57 <zbyszek> .fesco 2096
15:19:58 <zbyszek> https://pagure.io/fesco/issue/2096
15:19:59 <zodbot> zbyszek: Issue #2096: F31 System-Wide Change: BuildRequires generators - fesco - Pagure.io - https://pagure.io/fesco/issue/2096
15:20:33 <zbyszek> There's some voting int he ticket, but I thought it'd be good to discuss this too.
15:21:03 <zbyszek> If everyone is fine with just discussion in the ticket, we don't really need to do anythign here.
15:21:24 <jforbes> I was watching the discussion in the ticket, but I am generally +1 to the change
15:21:36 <nirik> I think it's kinda odd to approve something before it exists upstream, but you are right that we can just drop it if it's not ready in time.
15:21:53 <contyk> there are too many unknowns still
15:21:59 <zbyszek> The PR is there...
15:22:28 <jforbes> nirik: I don't find it that odd, there is a good bit of time there.
15:23:11 <contyk> I'd rather not vote on a proposal that includes TBD and "we don't know yet"
15:23:33 <contyk> so I'd rather have the proposal owners come with something complete first
15:23:37 <bowlofeggs> i think it can be useful to get approval from FESCo before doing a bunch of work
15:23:46 <bookwar> in the ticket it says focus on Rust only, isn't it better to reduce the scope to Rust then?
15:23:50 <bowlofeggs> because you wouldn't want to do work and then get told it can't be used
15:24:14 <mhroncok> I'm with bowlofeggs here
15:24:28 <zbyszek> Me too.
15:24:30 <mhroncok> if it is implemented, good
15:24:38 <mhroncok> if it is not, well, the change didn't make it
15:24:59 <contyk> I want to know whether the generated SRPMs will include those deps or not
15:25:03 <nirik> well, there's also the case of the implemented version being different from the proposed version...
15:25:50 <zbyszek> Actual use in Fedora will still be gated by FPC agreeing to new stanzas.
15:26:21 <mhroncok> and I'll make sure we cans till get the deps from the source repo
15:26:47 <contyk> "users won't notice differences", I also find that hard to believe :)
15:27:08 <otaylor> At this point, we should constrain ourselves to "this sounds useful" or "this doesn't sound like something Fedora would want" I think
15:27:51 <bowlofeggs> i think we can also say "we like the direction this is going, but please file a new ticket when it is implemented for final approval"
15:28:08 <contyk> I don't actually need it to be implemented
15:28:20 <contyk> but I want a more complete proposal
15:28:38 <otaylor> contyk: A too exact proposal may not survive rpm upstream
15:29:03 <zbyszek> contyk: ok, can you put those questions in the ticket, so that the change owners can respond, and we we can either vote in the ticket or revisit next week?
15:29:12 <contyk> otaylor: I still hope they know what they're trying to achieve, though
15:29:13 * nirik is in general in favor, but we should watch it closely to see whats implemented for sure.
15:29:20 <contyk> zbyszek: yes, will do
15:29:34 <zbyszek> OK, so let's move to the next item.
15:29:41 <zbyszek> #topic #2093 F31 System-Wide Change: Python 3.8
15:29:41 <zbyszek> .fesco 2093
15:29:41 <zbyszek> https://pagure.io/fesco/issue/2093
15:29:42 <zodbot> zbyszek: Issue #2093: F31 System-Wide Change: Python 3.8 - fesco - Pagure.io - https://pagure.io/fesco/issue/2093
15:29:52 <zbyszek> Does anyone know what QA decided in their meeting?
15:30:02 <mhroncok> isn't their meeting after ours?
15:30:11 <bcotton> yes, their meeting starts in 30 minutes
15:30:22 <zbyszek> Oops, my bad then.
15:30:54 <zbyszek> So I think we need to postpone #2093 and #2092.
15:31:08 <bowlofeggs> i'm +1 (just commented in ticket)
15:31:11 <mhroncok> we can vote conditionally on QA I guess if we want to move this forward as it's (kinda) blocking the schedule
15:31:28 <jforbes> We could just vote in ticket once QA responds
15:31:44 <jforbes> Unless people have questions outside of QA's input
15:31:52 <otaylor> mhroncok: how do you see it blocking the schedule? How would you change the schedule if this was approved or not approved?
15:32:22 <mhroncok> if this goes unapproved for being to tight, I'd try to extend the schedule to make this fit
15:33:02 <mhroncok> of course I realize that I might not find enough support for that, but I'd at least try
15:33:32 <bcotton> i'd argue against delaying the schedule because the fall release gets complicated quickly with holidays
15:33:34 * nirik was +1 in ticket, still +1
15:33:49 <mhroncok> bcotton: I knwo you would :)
15:34:04 <bcotton> but i'd also argue that the upstream freeze is sufficiently far in advance that (assuming they stick to it), we should be in good shape
15:34:17 <bcotton> (easy for me to say because i'm not the one doing QA, of course)
15:34:27 <zbyszek> mhroncok: would releasing F30 with python3.8-candidate be an option?
15:34:31 <mhroncok> we (python-main) are fairly confident to ship rc in Fedora
15:34:57 <zbyszek> Would 3.8-final then be released as an update?
15:35:00 <mhroncok> yet I don't want QA to be angry about that
15:35:02 <mhroncok> zbyszek: sure
15:35:39 <zbyszek> I'm still +1, the alternative options of delaying F30 or postponing 3.8 to F31 are not attractive at all.
15:36:14 <nirik> wait, this is a f31 change right?
15:36:22 <nirik> yeah...
15:36:48 <zbyszek> Right, sorry, F31 ... F32 ...
15:37:08 <zbyszek> So we're at +4 now.
15:37:23 <contyk> so, the proposal is to put 3.8rc in f30?
15:37:28 <contyk> I can +1 that
15:37:32 <jforbes> While I am in theory +1, I think it is bad form to ask QA for feedback and then ignore it
15:37:53 <zbyszek> jforbes: yep
15:38:05 <zbyszek> OK, so... Wait for QA feedback and vote in the ticket?
15:38:05 <jforbes> I will vote in ticket when they respond, taking their response into consideration.  Had we not asked, I would have voted already
15:38:28 <nirik> fair point
15:38:31 <mhroncok> note that it's the change owner here who explciitly want QA feedback
15:38:44 <mhroncok> so even if Fesco approves it and QA does not, we would recodnsider
15:38:54 <zbyszek> ok, ok
15:38:55 <jforbes> mhroncok: I don't think it is wrong for them to ask that
15:39:24 <mhroncok> me neither :D
15:39:26 <zbyszek> So I'm skipping #2092 Fedora 31 schedule too.
15:39:31 <jforbes> Still, I abstain from voting until that feedback is provided. I will vote in ticket when it is there
15:39:37 <zbyszek> #topic #1970 Action needed: Orphan packages will be retired if they remain orphaned for six weeks
15:39:39 <mhroncok> sure
15:39:40 <zbyszek> .fesco 1970
15:39:42 <zodbot> zbyszek: Issue #1970: Action needed: Orphan packages will be retired if they remain orphaned for six weeks - fesco - Pagure.io - https://pagure.io/fesco/issue/1970
15:39:43 <zbyszek> https://pagure.io/fesco/issue/1970
15:39:44 <mhroncok> ha
15:39:47 <mhroncok> that is on me
15:39:57 <mhroncok> it's getting tot he top of my TODO after vacation
15:40:24 <mhroncok> I want to draft a policy, specfiiy what needs to be done and reach for volunteers
15:40:31 <zbyszek> mhroncok: you have been sending out orphan warnings...
15:40:49 <zbyszek> This seems to be on track then.
15:40:55 <zbyszek> Do we need to discuss anything here today?
15:41:00 <mhroncok> I also want to make the policy ina  way that even if the automation is broken or doe snot exists yet, peple can open bugs for broken dependencies and FTBFS-like policy would apply
15:41:17 <zbyszek> That'd be excellent.
15:41:49 <zbyszek> Ideally, the scripts could simply be run by volunteers if needs arises, so this even wouldn't require manual opening of bugs.
15:42:14 <zbyszek> #action mhroncok to draft a policy update proposal
15:42:31 <zbyszek> #topic Next week's chair
15:42:36 <zbyszek> Volunteers?
15:42:39 <contyk> can do
15:42:46 <zbyszek> #action contyk to chair next meeting
15:42:50 <zbyszek> Thanks!
15:42:54 <zbyszek> #topic Open floor
15:43:04 <mhroncok> another compose is doomed
15:43:47 <mhroncok> (╯°□°）╯︵ ┻━┻
15:43:51 <zbyszek> I tried to fix as many FTBFS as I could this week, unfortunately I didn't have a lot of time...
15:44:32 <nirik> it's anoying how shifting sands it is... fix one issue and people have pushed two more.
15:44:35 <zbyszek> Any pointers for the most urgent things to fix?
15:44:48 <mhroncok> https://pagure.io/dusty/failed-composes/issue/1574
15:44:52 <mhroncok> Image: Field 'type' has invalid value: vmdk
15:44:58 * mhroncok has no idea this time
15:45:29 <nirik> it's a pungi bug.
15:45:40 <nirik> before that it was kde broken deps
15:46:03 <kalev> I think I asked releng if it would be possible to ignore all the image build failures once, so that we can get one good f31 compose out
15:46:07 <nirik> before that it was zchunk/createrepo_c making zck for comps that dnf couldn't understand
15:46:20 <nirik> before that it was...
15:46:28 <kalev> but releng/nirik said it's fesco policy to block on all the images
15:46:48 <kalev> I wonder if fesco could make an exception here so that we could get one good rawhide out that has f31 content and repos setup?
15:46:54 <nirik> I'm tempted to stop allowing changes so we can get a compose. For example this vmdk thing landed very recently
15:47:44 <nirik> the problem with ignoring failures, is that then they never get fixed...
15:48:17 <mhroncok> ignore them once?
15:48:19 <kalev> I was just imagining doing it once, just so that we could get one f31 content and repos setup out
15:48:23 <kalev> yeah
15:48:24 <nirik> but in any case the current failure isn't images failing. It's a bug in pungi itself... there's nothing we can do to avoid that other than downgrading or upgrading to a fixed version
15:48:51 <zbyszek> nirik: do you have a link to the pungi bug for the meeting logs?
15:49:12 <kalev> zbyszek: mhroncok posted it above
15:49:12 <nirik> no, since it just happened during this meeting. ;)
15:49:24 <zbyszek> :(
15:49:25 <nirik> well, yeah, there's the failed compose ticket...
15:49:53 * nirik can/will file one
15:50:19 <zbyszek> OK, anything else for open floor?
15:50:45 <contyk> is there anything you'd like me to highlight in the council eng report this week?
15:51:06 <zbyszek> A successful compose, surely ;)
15:51:12 <mhroncok> :D
15:51:12 <contyk> ;)
15:52:03 <bowlofeggs> nirik: do we have an easy way to stop changes on rawhide for a bit to get it stabilized for a build (outside of the pungi issue), in general?
15:52:22 <bowlofeggs> hopefully rawhide gating will help in the future, of course, but i mean today
15:52:29 <zbyszek> bowlofeggs: not really, apart from pleading on fedora-devel
15:52:33 <bowlofeggs> ah
15:52:44 <bowlofeggs> i was thinking if we could change koji build targets or something
15:52:44 <mhroncok> disbale koji access to evrybody except nirik :)
15:52:46 <nirik> well, I mean we could turn off koji. ;)
15:52:47 <contyk> we could clone the tag and compose from a semi-frozen set
15:52:47 <bowlofeggs> hahaha
15:53:03 <bowlofeggs> contyk: yeah that's not a bad idea
15:53:21 <bowlofeggs> might be less frustrating
15:53:32 <nirik> well, the packages are just one part.
15:53:43 <bowlofeggs> of course, then the main tag will get worse and worse while we stabilize our side tag
15:53:45 <bowlofeggs> haha
15:53:49 <nirik> pungi changes, pungi-fedora config changes, comps, kickstarts, etc.
15:53:52 <bowlofeggs> sure
15:54:05 <nirik> but those we can control more by just refusing to merge things.
15:55:10 <nirik> anyhow, we will keep plugging away... #fedora-releng to follow the fun
15:55:29 <zbyszek> #info /join #fedora-releng to follow the fun
15:55:35 <bowlofeggs> nirik: if i chew double mint gum, does it double the releng fun?
15:55:41 <zbyszek> OK, anyone else?
15:56:06 <zbyszek> I'll close in one minute.
15:56:07 <nirik> bowlofeggs: you bet
15:56:15 <bowlofeggs> (for those outside the US, double mint brand gum had a series of commercials in the 80/90s(?) with twins saying that they are double the fun)
15:56:30 <bowlofeggs> (or maybe they also sell/market outside the us, i have no idea)
15:56:42 <mhroncok> not here
15:57:04 <bookwar> "mhroncok: disbale koji access to evrybody except nirik" - that's what gating is for :)
15:57:04 <zbyszek> #endmeeting