15:00:06 <adamw> #startmeeting Fedora QA Meeting
15:00:06 <zodbot> Meeting started Mon May 27 15:00:06 2019 UTC.
15:00:06 <zodbot> This meeting is logged and archived in a public location.
15:00:06 <zodbot> The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:00:06 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:06 <zodbot> The meeting name has been set to 'fedora_qa_meeting'
15:00:10 <adamw> #meetingname fedora-qa
15:00:10 <zodbot> The meeting name has been set to 'fedora-qa'
15:00:14 <adamw> #topic Roll call
15:00:19 <adamw> morning, folks - how's everyone doing?
15:00:22 <Lailah> Hello
15:00:26 <Lailah> .fas lailah
15:00:27 <zodbot> Lailah: lailah 'Sylvia Sánchez' <BHKohane@gmail.com>
15:00:50 <lruzicka> .hello2
15:00:51 <zodbot> lruzicka: lruzicka 'Lukáš Růžička' <lruzicka@redhat.com>
15:02:36 <adamw> well, at least all the best people showed up :P
15:03:40 * kparal is hungry, might not last till the end of the meeting
15:03:49 * adamw eats kparal
15:04:25 * coremodule is here
15:05:14 <lruzicka> kparal should eat adamw, this did not help him much
15:05:34 <adamw> i got there first, though
15:05:36 <adamw> it was self defence
15:05:37 <adamw> :P
15:05:47 <adamw> alrighty, let's get rolling
15:05:49 <Lailah> LOL
15:05:54 <Lailah> adamw is right
15:06:00 <adamw> #topic Previous meeting follow-up
15:06:06 <adamw> we have just one mysterious item from the last meeting:
15:06:18 <adamw> "sgallagh to add lruzicka to the invite list"
15:06:25 <adamw> lruzicka, did you get into the...party?
15:06:30 <lruzicka> Yes, this happened indeed.
15:06:55 <Lailah> There was a party and nobody told me? How rude...
15:07:23 <lruzicka> We had one meeting where we discussed how modules will behave from the user's perspective ... we could agree on some points, but we were not able to finish everything on the agenda. Next such meeting is planned on Thursday-
15:07:27 <adamw> Lailah: i know, right?
15:07:33 <adamw> we'll have our own party! screw them!
15:07:41 <Lailah> Yeah!
15:07:44 <adamw> #action adamw and lailah to have a much better party
15:08:21 <adamw> #info "sgallagh to add lruzicka to the invite list" - this was about module behaviour issues. there was one meeting about this with some progress, not all points are decided, there will be another such meeting on thursday
15:08:50 <lruzicka> Yes, I will report everything we this is over.
15:08:52 <adamw> any other follow-up from last time? i wasn't here, so speak up if there's anything else we should check in on
15:08:54 <adamw> thanks lruzicka
15:08:55 <lruzicka> we -> when
15:09:18 <lruzicka> no problem
15:09:32 <coremodule> there was the stuff from the meeting before the last one... i didnt make the things we couldnt accomplish last meeting action items so they didnt show up here...
15:09:33 <Lailah> adamw  I wasn't here either, so I don't know
15:09:44 <coremodule> hang on...
15:10:15 <coremodule> this "adamw to revise F30 Common Bugs for Final"
15:10:23 <coremodule> and this "adamw to resurrect the 'late blocker waiver' proposal with a third
15:10:24 <coremodule> draft or re-send of the second draft"
15:10:29 <coremodule> we couldn't complete with you adamw
15:10:36 <adamw> aha
15:10:37 <adamw> thanks
15:10:42 <adamw> #info some action items from meeting before last
15:10:46 <coremodule> *without
15:11:01 <adamw> #info "adamw to revise F30 Common Bugs for Final" - this was done, ongoing revisions will continue until morale improves
15:12:02 <adamw> #info "adamw to resurrect the 'late blocker waiver' proposal with a third draft or re-send of the second draft" - I did not manage to do this before I went on holiday, but Patrick Kelly has done it instead and the decision is ongoing (thread "Last-minute blocker bugs")
15:12:08 <adamw> that look good>?
15:12:09 <Lailah> adamw: Who's morale, software's?
15:12:28 <adamw> just morale in general!
15:12:35 <Lailah> Oh, okay
15:13:21 <adamw> (it's just me being silly, don't worry about it)
15:13:35 <Lailah> I know :-)
15:13:39 <coremodule> *two thumbs up
15:15:06 <adamw> #topic Fedora 31 status / Change review
15:15:25 <adamw> so, Rawhide is currently actually testing out pretty well
15:15:35 <adamw> we had a couple compose issues last week but the composes that got out do mostly work
15:15:47 <adamw> anyone hit any big problems in manual testing?
15:16:15 <coremodule> i have been doing arm testing lately, no big issues, but I do have a question about silverblue...
15:17:08 <coremodule> what is the "appropriate" channel to go about getting an armhfp/armv7 compose of silverblue...? all my boards are armv7 minus a few exceptions, so i think not catering to this market is hurting the overall userbase
15:17:23 <lruzicka> And I have a question, if the KDE payload bug is having progress.
15:17:45 <Lailah> Uhm... I had a few issues with System Settings (KDE), but not in Rawhide.
15:18:29 <Lailah> lruzicka  what was that problem with KDE?
15:18:50 <adamw> coremodule: um. i think silverblue is not built for armhfp...
15:18:52 <lruzicka> Lailah, the installation fails with an Anaconda error, there is a bug about it
15:18:59 <adamw> i'm guessing that's on purpose...
15:19:18 * Lailah doesn't know if it's a L or an I the first letter of lruzicka...
15:19:28 <coremodule> adamw, right, I'm wondering how in the heck I could go about starting an inquiry to get it built for armhfp... or at least get more info as to why its not
15:19:32 <lruzicka> Lailah, its L
15:19:37 <Lailah> Ah.
15:19:38 <adamw> lruzicka: https://bugzilla.redhat.com/show_bug.cgi?id=1713109 ? I fixed it
15:19:40 <Lailah> Okay, thanks
15:20:19 <adamw> coremodule: i think https://discussion.fedoraproject.org/c/desktop/silverblue
15:20:40 <adamw> silverblue is using discourse as their primary communication method as an experiment aiui
15:20:55 <coremodule> adamw, thanks, I'll try there.
15:20:57 <coremodule> ahhhh
15:21:03 <Lailah> lruzicka & adamw I remember this bug...  well, seeing the report, that is
15:21:35 <Lailah> Yes, there's a Slack channel as well.
15:21:36 <adamw> it should be gone in the most recent compose (the kde install bug)
15:21:51 <Lailah> Okay, that's good
15:22:05 <lruzicka> adamw, good, as it was breaking a bunch of KDE related tests
15:24:11 <adamw> yeah.
15:24:35 <adamw> #info Fedora 31 is generally in decent state right now, a bug affecting install of KDE package sets was recently resolved
15:24:41 <adamw> looking at the changes so far...
15:25:20 <adamw> the obvious ones that jump out to me are https://fedoraproject.org/wiki/Releases/31/ChangeSet#F31_Mass_Python_2_Package_Removal , https://fedoraproject.org/wiki/Releases/31/ChangeSet#Retire_YUM_3 and https://fedoraproject.org/wiki/Releases/31/ChangeSet#Firefox_Wayland_By_Default_On_Gnome
15:27:05 <Lailah> The mass Python2 removal is affecting a number of packages. I know of Bleachbit that was going unmaintained.
15:27:38 <Lailah> I think someone picked it up, but still AFAIK it has Python2 dependencies, so I don't know what is going to happen with it
15:27:43 <adamw> the main impact for us may be on dependencies and image composes
15:27:51 <Lailah> Yes
15:30:09 <adamw> #info several Changes look like they may be significant risk factors for Fedora 31: mass Python 2 package removal, YUM 3 retirement, Firefox defaulting to native Wayland backend. we will continue to monitor the Change list as more changes are approved
15:30:13 <adamw> anything else on F31?
15:32:46 <lruzicka> not from me
15:33:10 <adamw> alrighty
15:33:13 <adamw> #topic Release criteria / blocker process proposals
15:33:39 <adamw> so, we got some response from xen folks on the proposal to drop the xen criterion
15:33:43 <adamw> what do folks think about that?
15:33:55 <lruzicka> keep it?
15:34:25 <lruzicka> or which folks do you mean, I was talking about the xen folks.
15:35:42 <kparal> what was the reason back then to block on something that's not in mainline kernel?
15:36:33 <adamw> lruzicka: i mean what do we folks think about what those folks said :P
15:36:37 <adamw> too many folks going around, i guess, hehe
15:36:43 <adamw> kparal: i don't honestly recall?
15:36:43 <Lailah> LOL
15:36:50 <adamw> kparal: i feel like it was to do with ec2 or something
15:36:58 * adamw sees if he can find the history
15:37:10 <kparal> adamw: xen is mandatory for ec2?
15:37:30 <adamw> at some point it was suggested that it was
15:37:33 <adamw> i don't know if it is any more
15:37:43 <adamw> there's a footnote " EC2 is by far the most significant use case here"
15:37:45 <Lailah> Uhm. I don't remember either
15:38:10 <lruzicka> adamw, yeah. I am thinking they want us to keep the criteria, right?
15:38:16 <kparal> my feeling is that we shouldn't block on it, the same way we don't block on e.g. virtualbox. but cloud people might have a different opinion here
15:38:26 <adamw> https://en.wikipedia.org/wiki/Amazon_Elastic_Compute_Cloud#Instance_types
15:38:35 <adamw> lruzicka: yeah, the question is, do we buy it :)
15:38:42 <adamw> kparal: yeah, that's still my instinct
15:38:54 <adamw> if the intent is to make sure we work on ec2, that should just be a direct requirement in itself
15:38:57 <adamw> (if it isn't already)
15:39:03 <kparal> adamw: so it looks like ec2 now supports kvm
15:39:20 <adamw> we do have this in the preamble: "Supported cloud environments
15:39:20 <adamw> Release-blocking cloud images must boot in the Fedora OpenStack Cloud and in Amazon EC2."
15:39:29 <adamw> we don't provide any more detail in terms of what 'instance types' we mean or anything liek that
15:39:35 <adamw> perhaps this would bear a bit more investigation
15:39:43 <Lailah> kparal:  I agree with you, I don't see it as something that should block.
15:39:50 <coremodule> I'm still +1 to removing the blocking criteria. It worries me that if we drop the criteria, we might lose some support from the xen team, the support they're offering here
15:40:47 <adamw> none of the types listed at https://aws.amazon.com/ec2/instance-types/ specifically mentions xen
15:40:57 <adamw> several of them specifically say they are based on nitro, which aiui is *not* xen
15:40:59 <coremodule> but I think that's a moot point, if they want their hypervisor in Fedora, they'll have to provide at least *some* support, akin to what they've been providing
15:41:07 <adamw> coremodule: right
15:41:34 <lruzicka> how do we want to test is, when we do not have access to the cloud? or we do?
15:42:04 <lruzicka> if the folks are able to test it and provide results in time, we could keep the criteria
15:42:18 <lruzicka> if not, I will say, drop it.
15:43:37 <adamw> i kinda feel like we should drop the criterion even if testing is available
15:43:44 <adamw> there are plenty of other things we test but don't block on
15:43:56 * kparal agrees with adamw
15:44:13 * Lailah agrees with both, kparal & adame
15:44:16 <Lailah> adamw
15:44:18 <Lailah> Sorry
15:44:37 <coremodule> agreed. ^^
15:44:57 <adamw> alrighty
15:44:59 <lruzicka> well, I was trying to think about a solution which would prevent them from stop their support
15:45:08 <lruzicka> as coremodule was suggesting
15:45:13 <adamw> i don't think we can make any decisions here, but i'll record the opinions and bump the thread along...
15:45:37 <Lailah> Okay
15:45:50 <Lailah> I'm good with that
15:45:59 <adamw> #info several folks agree that the criterion is not sufficiently justified even if xen project provides testing; we should only block on our primary virtualization stack, and we already have a specific requirement that we boot on ec2, so blocking on xen is not needed for that purpose either
15:46:04 <adamw> d'oh
15:46:05 <adamw> #undo
15:46:05 <zodbot> Removing item from minutes: INFO by adamw at 15:45:59 : several folks agree that the criterion is not sufficiently justified even if xen project provides testing; we should only block on our primary virtualization stack, and we already have a specific requirement that we boot on ec2, so blocking on xen is not needed for that purpose either
15:46:10 <adamw> #info several folks agree that the Xen criterion is not sufficiently justified even if xen project provides testing; we should only block on our primary virtualization stack, and we already have a specific requirement that we boot on ec2, so blocking on xen is not needed for that purpose either
15:46:39 <adamw> #info however, there was also some concern that dropping the criterion might lead to xen project not providing the testing they promised if the criterion remains
15:47:22 <adamw> anyone have any thoughts/notes on the 'last minute blocker' discussion?
15:47:56 <Lailah> Uhm... not from my side.
15:48:03 <Lailah> Not that I remember at least
15:48:15 <Lailah> I've been a bit absent in the last days
15:48:21 <lruzicka> I still support the quorum thing
15:49:27 <Lailah> The quorum thing?
15:49:31 <lruzicka> it does not have to be huge quorum, but at least a couple of people should be discussing that
15:50:07 <adamw> lruzicka: btw, a note there, we always try to avoid bare majority blocker votes
15:50:15 <lruzicka> Lailah, I suggested that a decision should be made by at least X people
15:50:17 <adamw> i don't know if we've ever had any, in fact
15:50:37 <Lailah> lruzicka Oh. Sounds sensible to me.
15:50:56 <adamw> i'm a bit worried that a specific number might be a bit inflexible
15:50:56 <lruzicka> adamw, and I am not suggesting, we are, but those last day blockers are a dangerous thing and I can imagine some situations.
15:52:12 <lruzicka> well, I have suggested 10 (like an example), those important meetings have almost all the time more than 10 in attendace.
15:52:59 <lruzicka> But I think even a lesser number will be just fine. I did not want to have a situation, where somebody decides that this bug will not be blocked on without a previous discussion.
15:53:32 <Lailah> Well, if there is only one person, is not a meeting, right?
15:53:45 <Lailah> Or a meeting with oneself counts as meeting?
15:54:27 <lruzicka> Lailah, if nobody comes to an announced meeting and only three people (with huge interest in waiving the bug) show up?
15:54:48 <lruzicka> or if some people are late for that meeting and that thing is decided before?
15:55:07 <jlanda> the essence of the last minute bug makes it difficult (even impossible) to be droped without discussion if we take bcotton's definition as valid
15:55:13 <adamw> #info there is ongoing discussion about exactly how the last-minute blocker bugs provisions should be handled and if a quorum should be required for waiving a blocker
15:55:15 <jlanda> Since there is already an ongoing meeting
15:55:20 <jlanda> :)
15:55:23 <adamw> let's go with that :P
15:55:26 <Lailah> Uhm... okay, I guess that's something that could happen lruzicka
15:56:22 <adamw> #topic Open floor
15:56:27 <adamw> so, anything for the last three minutes?
15:56:29 <lruzicka> Lailah, it happened in my life before Fedora ...
15:56:54 <jlanda> heya :) I'll read you later
15:56:58 <Lailah> lruzicka is there life before Fedora?
15:57:31 <lruzicka> Lailah, yeah, I am not that young :)
15:57:43 <Lailah> Heh, neither I
15:57:55 <adamw> what is this strange 'life outside fedora' idea
15:58:01 <adamw> sounds highly suspect to me
15:58:19 <Lailah> I don't know, it was lruzicka 's idea
15:58:50 <lruzicka> adamw, I did not say "outside", I said "before"
15:59:08 <adamw> that implies outside though
15:59:14 <adamw> were you not born in a fedora like the rest of us?
15:59:33 <Lailah> adamw: About the Open Floor...  I would just like to point out that System Settings is breaking way too often, and so far it didn't get any attention. This is on KDE BTW
15:59:51 <adamw> Lailah: is there a bug report for it?
15:59:55 <Lailah> Yup
15:59:58 <adamw> crashing system settings sounds like at least a candidate for blocker to me
16:00:02 <adamw> is it proposed?
16:00:11 <Lailah> No, I didn't propose it.
16:00:20 <lruzicka> adamw, something like this? https://zabav-deti.cz/foto/bob-a-bobek.jpg
16:00:26 <adamw> Lailah: i'd say go ahead and propose it...
16:00:32 <Lailah> I thought someone would take a look and at least send some feeback
16:00:35 <adamw> that's a top hat. but close!
16:01:12 <Lailah> adamw: Okay, I'll propose it then.
16:01:24 <adamw> #action lailah to propose KDE system settings crash as F31 blocker
16:01:26 <adamw> alrighty
16:01:30 <lruzicka> adamw, you are far too clever, thats a problem
16:01:35 <adamw> we're over time, so i think we'd better wrap it :)
16:01:37 <adamw> thanks for coming, everyone!@
16:01:39 <Lailah> It happened to me to use System Settings and it would crash two or three times before I would give up and close it
16:02:02 <adamw> #endmeeting