15:00:03 <ignatenkobrain> #startmeeting FESCO (2019-07-29)
15:00:03 <zodbot> Meeting started Mon Jul 29 15:00:03 2019 UTC.
15:00:03 <zodbot> This meeting is logged and archived in a public location.
15:00:03 <zodbot> The chair is ignatenkobrain. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:03 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:03 <zodbot> The meeting name has been set to 'fesco_(2019-07-29)'
15:00:13 <ignatenkobrain> #meetingname fesco
15:00:13 <zodbot> The meeting name has been set to 'fesco'
15:00:14 <ignatenkobrain> #chair nirik, ignatenkobrain, jforbes, zbyszek, bookwar, sgallagh, contyk, mhroncok, otaylor
15:00:14 <zodbot> Current chairs: bookwar contyk ignatenkobrain jforbes mhroncok nirik otaylor sgallagh zbyszek
15:00:20 <jforbes> .hello2
15:00:21 <zodbot> jforbes: jforbes 'Justin M. Forbes' <jforbes@redhat.com>
15:00:24 <ignatenkobrain> #topic init process
15:00:27 <mhroncok> hey
15:00:28 <contyk> .hello psabata
15:00:29 <zodbot> contyk: psabata 'Petr Šabata' <psabata@redhat.com>
15:00:41 <bcotton> .hello2
15:00:42 <zodbot> bcotton: bcotton 'Ben Cotton' <bcotton@redhat.com>
15:00:47 <nirik> morning
15:01:39 <otaylor> .hello2
15:01:40 <zodbot> otaylor: otaylor 'Owen Taylor' <otaylor@redhat.com>
15:02:19 <zbyszek> .hello2
15:02:19 <zodbot> zbyszek: zbyszek 'Zbigniew Jędrzejewski-Szmek' <zbyszek@in.waw.pl>
15:03:26 <ignatenkobrain> let's give few more minutes for others
15:04:59 <ignatenkobrain> ok, I guess we can start.
15:05:06 <ignatenkobrain> #topic #2179 python27 exception request for ptxdist
15:05:08 <ignatenkobrain> .fesco 2179
15:05:09 <zodbot> ignatenkobrain: Issue #2179: python27 exception request for ptxdist - fesco - Pagure.io - https://pagure.io/fesco/issue/2179
15:05:36 <mhroncok> still -1
15:05:45 <zbyszek> I guess.
15:05:54 <zbyszek> Too bad we don't have a reply from besser82.
15:06:36 <nirik> yeah, it sounds like we could just wait... but not fully clear
15:06:38 <mhroncok> I'm fine if we give him more tiem to reply
15:06:46 <mhroncok> *time
15:06:54 <ignatenkobrain> I'm 0 here. it does not require any dependencies outside of stdlib and it is not just this package blocking python2 removal
15:07:06 <zbyszek> OK, so wait one more week, deny if still no answer?
15:07:16 <jforbes> I would be good with one more week
15:07:17 <nirik> zbyszek: +1
15:07:23 <mhroncok> for my sake it can wait even a month
15:07:24 <contyk> +1
15:07:29 <mhroncok> but +1, why not
15:07:29 <contyk> to zbyszek, that is
15:07:44 <ignatenkobrain> +1 to zbyszek
15:09:29 <ignatenkobrain> otaylor: jforbes are you +1 to zbyszek 's proposal?
15:10:35 <jforbes> I already answered that, _1
15:10:37 <ignatenkobrain> #agree Wait for one more week, deny if no answer from besser82 (+5, 0, -0)
15:10:37 <otaylor> sure, +1
15:10:37 <jforbes> err +1
15:10:45 <ignatenkobrain> #undo
15:10:45 <zodbot> Removing item from minutes: AGREED by ignatenkobrain at 15:10:37 : Wait for one more week, deny if no answer from besser82 (+5, 0, -0)
15:10:53 <ignatenkobrain> #agree Wait for one more week, deny if no answer from besser82 (+7, 0, -0)
15:11:05 <ignatenkobrain> #topic #2180 F31 System-Wide Change proposal (late): No i686 Repositories
15:11:21 <ignatenkobrain> .fesco 2180
15:11:22 <zodbot> ignatenkobrain: Issue #2180: F31 System-Wide Change proposal (late): No i686 Repositories - fesco - Pagure.io - https://pagure.io/fesco/issue/2180
15:11:46 <zbyszek> Hmm, was the on the schedule?
15:11:51 <zbyszek> *this
15:11:57 * nirik happy to try and answer any questions...
15:12:07 <ignatenkobrain> zbyszek: nirik asked me to bring it
15:12:08 <nirik> I can also abstain unless there's a tie or something.
15:12:44 <mhroncok> contyk, sgallagh, ignatenkobrain: any of you would eb abel to answer my 2 questions there?
15:12:46 <zbyszek> Next time, please reply to the schedule. I'm not against, but I need to re-read the ticket.
15:12:55 <contyk> looking
15:13:13 <contyk> note modules do support multilib
15:13:24 <mhroncok> for reference: 1) How does one locally debug a i686 modular build failure?
15:13:28 <nirik> they do? Thats unfortunate.
15:13:29 <mhroncok> contyk: noted
15:13:40 <mhroncok> 2) How does this collide with Modules in buildroot enablement?
15:14:20 <contyk> if there are no repos, local builds and debugging will be complicated
15:14:36 <nirik> there is the koji buildroot repo.
15:14:45 <contyk> re:2) we could still compose the buildroot modules for the urse prime case
15:15:21 <contyk> local module builds support two modes -- "online" and "offline"
15:15:38 <contyk> the online mode pulls RPMs directly from koji tags and "composes" them locally, that would still work
15:15:49 <contyk> the offline mode relies solely on repodata and repo content
15:15:53 <contyk> that would not
15:16:18 <contyk> note the offline mode is the default now, afaik; however, there would still be a way
15:16:46 <mhroncok> contyk: so to confirm, if I had a i686 modular build failure, I would be able to debug it locally by using the online mode even if the modular repos ar enot composed
15:17:18 <contyk> yes, the online mode + the koji repos for the non-modular content
15:17:45 <mhroncok> can we get that documented?
15:17:58 <contyk> it might need some tweaks to the tools, I'm not sure how exactly it's invoked
15:18:22 <contyk> besides us not liking i686, what's the reason for the rush?
15:19:01 <mhroncok> do we have somebody who would do the tweaks or is it one of those "this is possible but doesn't work now and nobody is going to work on it" situations?
15:19:08 <nirik> composing and publishing i686 repos means: a) people can/will upgrade to it with old kernels and b) we waste time/bw composing and publishing them.
15:20:35 <contyk> I suppose this will be both an MBS and fedpkg change; so it's all about getting their maintainers to identify what needs to be changed and then do it
15:20:51 <contyk> chances are it could be all fixed within half an hour at Flock but you never know
15:20:56 <nirik> no, there's no changes there...
15:21:05 <nirik> we would still build i686, koji would still have it.
15:21:11 <nirik> we would just not compose and mirror it out
15:21:13 <contyk> for the local builds, nirik
15:21:44 <nirik> ok... so that expects a local mirror ?
15:22:08 <contyk> no, but it just needs to know where to look
15:22:25 <mhroncok> and as for 2) the ursa prime document at https://docs.google.com/document/d/1NxBbTPVJ17Ax4k-NbtorzxYspJ_CsjQ9m3In2wqWqVU/edit?pli=1# now says: "Koji needs to be updated so that build tag repositories can be created from more than one source -- the tag contents and zero or more external repositories." will those "external repositories" work when we don't compose them?
15:22:27 <nirik> ok, so could use koji buildroot....
15:23:03 <contyk> mhroncok: that's why I said we could still compose the buildroot modules for this case
15:23:12 <contyk> it's supposed to be yet another compose anyway
15:23:30 <contyk> as it might include modules not shipped in the normal repos and vice versa
15:23:56 <mhroncok> "we could still compose the buildroot modules" for i686, while we don't compose classic "Modular Repository" or i686
15:24:07 <contyk> yep
15:24:09 <mhroncok> ack
15:24:26 <mhroncok> so I'm OK with 2), but with 1) I'm still not sure
15:25:08 <contyk> it might already work, I don't know :)
15:25:13 <mhroncok> can somebody actually check how to build modules locally if the i686 Modular and Everything composes ar enot available?
15:25:34 <contyk> I can play with it
15:26:25 <mhroncok> nirik: there's no hurry dropping the composes, we could investigate first, correct?
15:26:51 <nirik> sure... but we should decide before beta freeze.
15:27:57 <nirik> end of aug...
15:27:59 <zbyszek> That's 2019-08-29, still quite a bit of time.
15:28:25 <jforbes> Well, it would be better to decide before branch wouldn't it?
15:28:33 <mhroncok> proposal: we try to get answer for the "modularity local build" question and we defer this decision until we know the answer or until 2 weeks before the beta freeze (whatever happens first)
15:28:47 <contyk> ack
15:28:51 <zbyszek> +1
15:29:18 <jforbes> +1
15:29:21 <nirik> jforbes: yeah, possibly so...
15:29:36 <nirik> just more work, but would be nice to do it once.
15:30:12 <jforbes> Though I would like that answer to be more in depth than "it doesn't work" if it doesn't, what would need to change to make it work?
15:30:16 <zbyszek> "2 weeks" is exactly one day before the branch point.
15:31:13 <ignatenkobrain> sorry, had to disappear due to broken glass around me
15:32:09 <nirik> ignatenkobrain: yikes. stay safe.
15:32:25 <nirik> anyhow, +1 to waiting a bit more for more info.
15:32:31 <zbyszek> contyk, ignatenkobrain, otaylor: vote on mhroncok's proposal to postpone?
15:32:47 <contyk> s/ack/+1/
15:32:54 <otaylor> +1
15:33:01 <zbyszek> ignatenkobrain: I hope it's not the laptop screen ;)
15:33:04 <ignatenkobrain> mhroncok: +1
15:34:37 <ignatenkobrain> #agree Defer decision until we know answer for the "modularity local build" question or until 2 weeks before the beta freeze (whatever happens first) (+7, 0, -0)
15:34:49 <ignatenkobrain> #topic Next week's chair
15:35:10 <contyk> sgallagh was volunteering last week ;)
15:35:13 <nirik> I think sgallagh signed up for it?
15:35:14 <nirik> yeah
15:36:14 <jforbes> <sgallagh> I'll be out next week, but I can take the week after
15:36:19 <jforbes> From the 7/22 logs
15:36:29 <ignatenkobrain> zbyszek: no, just some glass after drinking wine in the office. Accidentally dropped it from the table :)
15:37:30 <zbyszek> oh
15:39:23 <zbyszek> ?
15:39:58 <zbyszek> action and openfloor?
15:41:23 <mhroncok> #action sgallagh to chair next week
15:41:34 <mhroncok> #topic Open Floor
15:41:43 <ignatenkobrain> oh, so sgallagh will do it?
15:41:44 <ignatenkobrain> I'm blind
15:41:57 <ignatenkobrain> I thought that he can't do next meeting
15:41:59 <ignatenkobrain> sorry
15:42:11 <contyk> it's this one
15:42:16 <ignatenkobrain> I don't have anything for the open flor
15:42:32 <nirik> I have just a few quick things...
15:42:54 <nirik> #info mass rebuild was started last week. Took about 3.5 days and is all done now.
15:43:12 <nirik> #info single package rawhide gating is hooked up and ready for people who want to opt-in.
15:43:37 <ignatenkobrain> cool
15:43:47 <nirik> #info after some builder tweaks, rawhide composes are taking now about 4.5 hours. Down from 8-12.
15:43:56 <zbyszek> nirik: what are the failure stats?
15:44:05 <contyk> \o/
15:44:12 <mhroncok> zbyszek: https://kojipkgs.fedoraproject.org/mass-rebuild/f31-failures.html
15:44:13 <nirik> https://kojipkgs.fedoraproject.org/mass-rebuild/f31-failures.html
15:44:16 <nirik> 1469 failed
15:44:33 <mhroncok> but half of failures I've seen in my packages were cpio failures on random arches
15:44:50 <nirik> I resubmiited all the failed ones yesterday
15:45:01 <mhroncok> nirik: does this page reflect that?
15:45:04 <nirik> before that there was 1760 failures
15:45:12 <mhroncok> nirik: cool! thanks
15:45:12 <nirik> yes
15:45:23 <mhroncok> where is this going to be merged to f31?
15:45:27 <mhroncok> *when
15:45:35 <nirik> later today.
15:46:11 <mhroncok> One thing from me. The nonresponive policy is blcoked by https://bugzilla.redhat.com/show_bug.cgi?id=1731835
15:46:52 <nirik> oh, can we add a link to the non responsive maintainer policy/process in tickets in the fesco tracker?
15:47:09 <mhroncok> if you could 1) vote the severity up (not sure if it helps) and/or 2) give a friendly nudge to the bugzilla people if you know them personally?
15:47:09 <nirik> ie, if someone just gets an email they don't know whats going on... a link might help there.
15:47:27 <mhroncok> I'll add it to the template
15:48:30 <nirik> mhroncok: thanks!
15:48:51 <ignatenkobrain> mhroncok: sure
15:49:20 <ignatenkobrain> something else?
15:50:45 <mhroncok> nirik: added
15:50:57 <ignatenkobrain> then I'll close in 10 seconds if nobody has anything :)
15:51:41 <ignatenkobrain> #endmeeting