15:00:38 <zbyszek> #startmeeting FESCO (2018-04-13)
15:00:38 <zodbot> Meeting started Fri Apr 13 15:00:38 2018 UTC.  The chair is zbyszek. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:38 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:38 <zodbot> The meeting name has been set to 'fesco_(2018-04-13)'
15:00:38 <zbyszek> #meetingname fesco
15:00:38 <zodbot> The meeting name has been set to 'fesco'
15:00:38 <zbyszek> #chair maxamillion dgilmore nirik jsmith sgallagh bowlofeggs tyll jwb zbyszek
15:00:38 <zodbot> Current chairs: bowlofeggs dgilmore jsmith jwb maxamillion nirik sgallagh tyll zbyszek
15:00:41 <zbyszek> #topic init process
15:00:50 <jsmith> .hello2
15:00:51 <zodbot> jsmith: jsmith 'Jared Smith' <jsmith.fedora@gmail.com>
15:00:53 <jwb> i'm partially here
15:01:54 <nirik> I'm sort of here also.
15:02:11 <maxamillion> .hello2
15:02:12 <zodbot> maxamillion: maxamillion 'Adam Miller' <maxamillion@gmail.com>
15:03:22 <zbyszek> bowlofeggs said he cannot join today
15:03:30 <zbyszek> sgallagh said he'll be late
15:03:51 <zbyszek> So we have quorum, sort of, if jwb and nirik are counted in
15:03:55 <zbyszek> Let's start.
15:04:28 <zbyszek> = Followups =
15:04:28 <zbyszek> #topic #1877 large number of packages FTBFS in F28
15:04:28 <zbyszek> .fesco 1877
15:04:28 <zbyszek> https://pagure.io/fesco/issue/1877
15:04:30 <zodbot> zbyszek: Issue #1877: large number of packages FTBFS in F28 - fesco - Pagure - https://pagure.io/fesco/issue/1877
15:04:43 <zbyszek> Let me provide a short status update
15:04:47 <dgilmore> hi all
15:04:53 <zbyszek> dgilmore: hi
15:05:22 <zbyszek> I started hacking on this, and fired off builds for all bugs in NEW state that don't have build in f28-build ending with .fc28.
15:05:54 <zbyszek> It seems that there is a number of packages which were build and updates have been made for them, but the bugs have not been closed (~5% maybe)
15:06:12 <zbyszek> Then there's a number of packages which were built but no bodhi update was created
15:06:25 <tyll> .hello till
15:06:26 <zodbot> tyll: till 'Till Maas' <opensource@till.name>
15:06:33 <zbyszek> But a majority of packages actually still fail to build now
15:06:46 <dgilmore> zbyszek: yeah, sometimes people ignore the bugs
15:07:08 <zbyszek> I'll be closing the bugs for packages which were built successfull and have a bodhi update
15:07:11 <jsmith> I provided some points in the ticket.
15:07:49 <zbyszek> jsmith: yes, but that's all longer term
15:07:49 <jsmith> In short -- it would be nice to know which packages are still FTBFS, and whether it's because of dependencies.... and then focus on packages that are breaking the highest number of other packages.
15:08:56 <zbyszek> Indeed, that would help. But so far the failures I have seen have been in build, and not in the dependency install phase.
15:10:07 <zbyszek> My impression is that a majority of packages is are real failures in the build phase... so any sort of automated processing is not going to help much.
15:10:54 <zbyszek> Then, there's a number of packages for which the last build is .fc24 – .fc26.
15:11:04 <zbyszek> I think we should consider retiring those.
15:11:55 <zbyszek> So far:   20 free  112 open  372 done  443 failed
15:13:05 <zbyszek> Shit, I need to step out for a moment.
15:13:26 <zbyszek> plan is at https://paste.fedoraproject.org/paste/ZyfiYxQB2punYZtNRDGQwQ
15:13:31 <zbyszek> somebody needs to take over sorry
15:13:36 <tyll> IIRC we used to retire pkgs that FTBFS for two releases
15:13:49 <tyll> zbyszek: no problem, take care
15:14:47 <nirik> it would be nice if we could automate that
15:15:58 <tyll> we could, but we need someone to do it :-)
15:16:46 <nirik> right
15:17:53 <jsmith> Agreed.
15:18:43 <tyll> What can we do now? Usually rel-eng handled the mass rebuilds and cleanup, IMHO the next steps would be to cleanup the bug reports to get a better picture
15:18:58 <pjones> Also, most failures in the build phase are still going to be because of dependencies.
15:20:29 <tyll> Probably all failures are because of dependencies, somehow, if the package built in the past and was not changed ;-)
15:20:41 <maxamillion> likely so, yes
15:21:17 <pjones> tyll: well, some are going to be "gcc changed behavior", which aren't dependencies in the same sense, at least... :)
15:23:59 <tyll> how about I create a list of current FTBFS packages and post it to devel, ask them to be fixed or they will be retired in 6 weeks?
15:24:12 <tyll> not sure, how to move this forward otherwise :-)
15:24:25 <jwb> good
15:25:14 <nirik> sure. +1
15:26:38 <tyll> proposal: till will create a list of FTBFS packages, post it to devel and announce that they will be retired in 6 weeks if not being worked on
15:27:12 <tyll> maxamillion: jsmith  dgilmore sgallagh ^
15:28:05 <jsmith> +1
15:28:15 <maxamillion> +1
15:28:18 <nirik> who will retire them?
15:28:48 <tyll> I can also retire them
15:28:53 <maxamillion> we could ask releng? (or at least let them know)
15:30:20 <zbyszek> I'm back. I forgot to pick up my daughter from school ;(
15:31:18 <tyll> zbyszek: do you want to vote and take over? proposal: till will create a list of FTBFS packages, post it to devel and announce that they will be retired in 6 weeks if not being worked on
15:31:29 <tyll> we are at +5 so far?
15:31:38 <zbyszek> +1
15:31:39 <tyll> I can finish this vote
15:31:49 <tyll> # agree till will create a list of FTBFS packages, post it to devel and announce that they will be retired in 6 weeks if not being worked on (+5, 0, -0)
15:31:57 <tyll> #agree till will create a list of FTBFS packages, post it to devel and announce that they will be retired in 6 weeks if not being worked on (+6, 0, -0)
15:32:13 <tyll> #action till will send a list of FTBFS packages to devel and retire them in 6 weeks
15:32:25 <zbyszek> OK, I'll take over again.
15:32:26 <zbyszek> #topic #1874 Package Anki has a Non-responsive Maintainer
15:32:26 <zbyszek> .fesco 1874
15:32:26 <zbyszek> https://pagure.io/fesco/issue/1874
15:32:28 <zodbot> zbyszek: Issue #1874: Package Anki has a Non-responsive Maintainer - fesco - Pagure - https://pagure.io/fesco/issue/1874
15:33:11 <zbyszek> I think we can close this, based on the last comment.
15:33:42 <tyll> +1
15:33:56 <jsmith> +1
15:34:04 <jsmith> Glad to have some resolution to that ticket :-)
15:34:14 <nirik> yeah, +1 to close
15:34:25 <tyll> maxamillion: jwb dgilmore ^
15:34:32 <jwb> +1
15:35:22 <tyll> #ifno anki maintainer is responsive again
15:35:27 <tyll> #info anki maintainer is responsive again
15:35:47 <zbyszek> +1 to my proposal ftr
15:36:11 <zbyszek> #agreeed No additional action is required from FESCo (+5, 0, 0)
15:36:15 <zbyszek> #agreed No additional action is required from FESCo (+5, 0, 0)
15:36:18 <zbyszek> Pfff.
15:36:29 <zbyszek> #topic #1872 Disable Test Gating requirements until more UI is enabled
15:36:29 <zbyszek> .fesco 1872
15:36:29 <zbyszek> https://pagure.io/fesco/issue/1872
15:36:31 <zodbot> zbyszek: Issue #1872: Disable Test Gating requirements until more UI is enabled - fesco - Pagure - https://pagure.io/fesco/issue/1872
15:36:48 <jsmith> @bowlofeggs: Any status update here?
15:37:01 <zbyszek> bowlofeggs is afk today
15:37:08 <jsmith> Oh :-(
15:37:25 <jwb> i'm -1 to the overall idea of disabling gating right now
15:37:45 <maxamillion> I'm on the fence
15:37:49 <jsmith> For the record, I'm still +1 to disabling gating until we have a better UI/developer experience
15:37:49 <maxamillion> +0 from me
15:38:04 <nirik> well, we made some improvements this week
15:38:17 <jsmith> nirik: That's what I was hoping to hear from bowlofeggs :-)
15:38:18 <nirik> we got the bodhi 'waive' button working (but I don't think it's in prod yet)
15:38:36 <nirik> and we fixed the topic/messages so they were not confusing
15:39:40 <zbyszek> nirik: that's good to hear
15:40:19 <zbyszek> Is more work planned?
15:40:35 <nirik> yeah, definitely.
15:40:41 <jsmith> Proposal: Wait another week to see how the Bohdi changes shake out before deciding whether to disable gating
15:40:48 <zbyszek> +1 yeah
15:41:00 <maxamillion> +1
15:41:03 <nirik> oh, the ui also shows waivers and what tests were missing...
15:41:12 <nirik> so when we get that bodhi out it should help
15:41:32 <jsmith> I'd be very inclined to change my mind on disabling gating if these changes go to production :-)
15:41:33 <tyll> +1
15:42:07 <pingou> and we'll be able to run the cron job more frequently
15:42:27 <zbyszek> jwb, dgilmore?
15:43:25 <zbyszek> nirik, also please vote formally
15:43:41 <nirik> +1 to wait a week... we will try and get these changes landed.
15:43:57 <zbyszek> OK.
15:43:58 <zbyszek> #agreed Wait another week to see how the Bohdi changes shake out before deciding whether to disable gating (+5, 0, 0)
15:44:09 <zbyszek> #topic #1868 Change the Changes template
15:44:09 <zbyszek> .fesco 1868
15:44:09 <zbyszek> https://pagure.io/fesco/issue/1868
15:44:11 <zodbot> zbyszek: Issue #1868: Change the Changes template - fesco - Pagure - https://pagure.io/fesco/issue/1868
15:45:54 <maxamillion> I'm +1 to the change, I think it makes sense and it will help the Fedora Marketing group
15:46:18 <jsmith> I'm also +1 to the change
15:46:33 <jsmith> (while acknowledging the mistakes of the past with "Features")
15:46:48 <zbyszek> maxamillion, jsmith: you mean as proposed in https://pagure.io/fedora-marketing/issue/raw/files/b212833149dad842515b2e928dc535f0129fc4dc973606dc06c64b616165c251-change2changes.patch?
15:46:59 <zbyszek> or just in general?
15:47:18 <jsmith> zbyszek: Yes
15:47:25 <jsmith> zbyszek: As proposed
15:48:00 * jsmith encouraged the marketing group to give examples, rather than just mandate a strict formula to follow
15:48:00 <maxamillion> yes, as proposed
15:49:19 <zbyszek> jsmith: but before you said " Let's show people how to do the right thing, but not make the template so rigid that it causes frustrations."
15:49:54 <zbyszek> I think that the proposed format is just that, too rigid, and also a bit artificial.
15:52:50 <zbyszek> OK, what about the following: this seems stuck. What about I prepare a counter-proposal along the lines I think are better, and then we can discuss if it is any better?
15:53:44 <nirik> sure
15:54:52 <zbyszek> maxamillion, jsmith, ok?
15:55:12 <jsmith> Sure, I'm OK with that counter-proposal
15:55:28 <zbyszek> #action zbyszek to generate a counter-proposal to spur discussion
15:55:34 <zbyszek> = New business =
15:55:34 <zbyszek> #topic #1878 Please change "Everything" directory to something less inaccurately comprehensive
15:55:37 <zbyszek> .fesco 1878
15:55:39 <zodbot> zbyszek: Issue #1878: Please change "Everything" directory to something less inaccurately comprehensive - fesco - Pagure - https://pagure.io/fesco/issue/1878
15:55:40 <zbyszek> https://pagure.io/fesco/issue/1878
15:55:48 <zbyszek> another bikeshedding exercise ;)
15:56:14 <jsmith> The bike shed should obviously be a tool shed, and should be painted blue.
15:56:39 <jsmith> That being said, I'm fine with "Traditional" or "Classic" or "RPM_Base" or anything along those lines.
15:56:55 <jsmith> I'd even be ok with "AlmostEverything"
15:57:31 <tyll> \o/ mostly my proposals
15:57:36 <zbyszek> rpm-collection or RPMCollection is nice
15:57:50 <jsmith> "There are three hard problems in Computer Science.... naming things, cache invalidation, cache invalidation, and off-by-one errors"
15:57:54 * nirik is just thinking of all the work, but sure.
15:58:45 <zbyszek> Right, your comment indicated that there's quite a number of places to touch.
15:58:50 <jsmith> nirik: Yeah, in my mind the work (and changes on the mirrors) are more work than the name change is worth
15:59:05 <zbyszek> Can you estimate how many person-hours the work would be?
15:59:17 <nirik> I have no exact estimate...
15:59:41 <zbyszek> pungi, mm, sync scripts sound like >few hours each
15:59:59 <jsmith> Can I propose we make the change *after* F28?
16:00:02 <nirik> probibly yeah.
16:00:03 <maxamillion> I have no desire to bikeshed on this, if someone has a proposed name that isn't just outrageous, I'll +1 it
16:00:14 <jsmith> Any reason it needs to be done immediately?
16:00:19 <nirik> jsmith: I would be very much in favor of that.
16:00:35 <tyll> maybe we can ask Matthew if there was some feedback that it actually causes confusion or is it something that he fears it might?
16:00:38 <nirik> doing it right before final is pretty crazy
16:00:47 <jsmith> nirik: I couldn't agree more!
16:01:10 <jsmith> It also gives us the chance to take a broader view and say "Are there any other naming/layout changes we want to make at the same time"
16:01:26 * zbyszek is leaning towards not changing this at all, since there is no clearly superior name and the amount of work involved is nontrivial
16:01:34 <jsmith> In other words, let's be deliberate and thoughful about it, rather than being rash to make a quick change.
16:01:48 <tyll> IMHO it would be nice to have a name in general for the traditional artifacts in general, just to make it easier to talk about them
16:02:18 <nirik> oh, other places that may need changes: fedfind, bunch of stuff on the wiki/docs, websites.
16:02:28 <jsmith> brb
16:02:32 <tyll> but maybe the stuff will go away anyhow in five years and we will be all using modular OSTrees
16:03:49 <zbyszek> Let's move on...
16:04:06 <zbyszek> #info We'll return to this after F28 is released
16:04:09 <zbyszek> OK?
16:04:37 <x3mboy> !
16:04:51 <tyll> +1
16:04:53 <nirik> +1
16:05:05 <zbyszek> #topic #1876 Please orphan all LXQt packages / nonresponsive maintainer heliocastro
16:05:08 <zbyszek> .fesco 1876
16:05:09 <zodbot> zbyszek: Issue #1876: Please orphan all LXQt packages / nonresponsive maintainer heliocastro - fesco - Pagure - https://pagure.io/fesco/issue/1876
16:05:11 <zbyszek> https://pagure.io/fesco/issue/1876
16:06:05 <jsmith> I don't know that there's any debate on this one... seems like a formality
16:06:18 <nirik> x3mboy: just chime in if you want to add something.
16:06:27 <x3mboy> 2 stuff
16:06:35 <x3mboy> First, I'm a little late, but in the name of the Fedora Marketing group, thanks for taking the Changes to changes ticket and discuss it
16:06:59 <x3mboy> Second, there is a lot of move in the LXQt upstream ML to retake the project, but it will take a lot of time
16:07:06 <zbyszek> x3mboy: we can return to the Changes change later on
16:07:11 <zbyszek> (if you want to)
16:07:19 <x3mboy> zbyszek, I'm ok if there is time
16:07:48 <zbyszek> proposal: wait another week for the maintainer to respond and orphan packages if there's no reponse
16:08:01 <jsmith> zbyszek: +1 from me
16:08:13 <tyll> +1
16:09:16 <nirik> +1
16:09:21 <zbyszek> maxamillion, dgilmore, jwb, sgallagh ?
16:09:30 <jwb> +1
16:09:55 <maxamillion> +1
16:10:09 <zbyszek> #agreed Wait another week for the heliocastro to respond and orphan packages if there's no reponse (+6, 0, 0)
16:10:18 <zbyszek> #topic Next week's chair
16:10:32 <zbyszek> Any takers?
16:11:31 <zbyszek> import random; random.choice('maxamillion dgilmore nirik jsmith sgallagh bowlofeggs tyll jwb zbyszek'.split()) ?
16:11:40 <jsmith> I can take it if nobody else volunteers
16:12:02 <jsmith> I *will* have a hard stop at an hour next week, though
16:12:16 <maxamillion> I can take it, I've not done it in a while.
16:12:26 <jsmith> Thanks maxamillion -- i'm happy to take it after that
16:12:33 <maxamillion> +1
16:12:37 <zbyszek> #action maxamillion will chair next meeting
16:12:39 <zbyszek> thanks maxamillion
16:13:07 <zbyszek> #topic #1868 Change the Changes template
16:13:19 <zbyszek> x3mboy: do you want to say something?
16:13:53 <x3mboy> Yes, the marketing team use the changes to try to expand the information that will use Ambassadors to reach people in conferences
16:14:33 <x3mboy> So the idea is to include a "non-technical" section that can be more readable to users in how a change is good for the OS
16:15:13 <x3mboy> In this way, we can simplify our Talking Points to talk to people
16:15:57 <x3mboy> The new sections are intended to simplify the way we explain this to our users and new users
16:15:59 <x3mboy> eof
16:17:00 <zbyszek> Yeah, I can live with that, but I don't like the section headers in the patch ;)
16:17:06 * nirik has to leave.
16:17:16 <zbyszek> I'll try to reword this and maybe we can discuss in the ticket?
16:17:22 <x3mboy> Sure!
16:17:24 <x3mboy> Thanks a lot
16:17:28 <x3mboy> zbyszek++
16:17:28 <zodbot> x3mboy: Karma for zbyszek changed to 7 (for the f27 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
16:17:35 <x3mboy> Also, I need to run
16:17:40 <x3mboy> Thanks again for taking this
16:17:48 <zbyszek> x3mboy: thanks for coming
16:17:54 <zbyszek> #topic Open Floor
16:18:18 <tyll> about the FTBFS packages, there are about 2944 of them
16:18:50 <tyll> tehrefore I plan to try to rebuild all of them and then do the reporting for the remaining pkgs
16:18:51 <zbyszek> tyll, no I don't think so
16:19:07 <zbyszek> There's only 817 bugs open
16:19:23 <zbyszek> And I rebuilt a bunch today, and some can be closed.
16:19:34 <zbyszek> 0 free  34 open  386 done  529 failed
16:20:17 <zbyszek> I think there's about 10 tasks per package, so 386 done means ~40 packages built successfully.
16:20:41 <tyll> zbyszek: ok, my bad, let me rephrase: there are 2944 pkgs that were not rebuild in the F28 mass rebuild
16:21:16 <zbyszek> 2944 binary or source?
16:21:17 <tyll> https://koji.fedoraproject.org/koji/packageinfo?packageID=19167 for example does not have a build for f28
16:21:21 <tyll> source
16:21:32 <tyll> I guess there is also not a FTBFS bug for it
16:21:37 <tyll> something went wrong there
16:21:52 <tyll> or I am missing something
16:22:15 <zbyszek> Yes, no bug.
16:22:24 <zbyszek> And it should have been rebuilt.
16:22:48 <zbyszek> Should we try to build those packages too for F28?
16:22:59 <zbyszek> It seems pretty late.
16:23:57 <zbyszek> tyll: I'll post my script to connect bugzilla, koji, and bodhi to avoid rebuilding packages that dont' need to be rebuilt.
16:24:15 <tyll> not sure yet, it seems that koji does not show a failed build since the pkg fails early: https://kojipkgs.fedoraproject.org//work/tasks/8161/26348161/build.log
16:25:16 <tyll> zbyszek: even if there is a bug report, we would need to fix the bug reports to contain the build logs
16:25:47 <tyll> at least some bugs did not contain build logs and the builds are too old now for koji to still have them
16:25:50 <zbyszek> 26348161/build.log looks like a hiccup in the build system
16:26:07 <zbyszek> error: File /builddir/build/SOURCES/e1c259484b495133836706f46319f5897f1e9bf6.tar.gz: No such file or directory
16:27:24 <tyll> zbyszek: no, the sources file is wrong
16:27:58 <zbyszek> Oh.
16:29:25 <tyll> nevertheless, the issue is a little bigger than expected, I will file a releng ticket about the missing bug reports, maybe we can also coordinate after the meeting how to proceed
16:29:37 <zbyszek> Yes, sounds like a plan.
16:29:51 <zbyszek> Let's wrap this up.
16:30:02 <zbyszek> Anything else for Open Floor?
16:30:07 <zbyszek> I'll close in 1 minute.
16:30:08 <jsmith> Thanks zbyszek for making an effort to get those cleaned up :-)
16:30:14 <jsmith> And for hosting the meeting today.
16:30:17 <jsmith> zbyszek++
16:30:17 <zodbot> jsmith: Karma for zbyszek changed to 8 (for the f27 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
16:30:59 <zbyszek> #endmeeting