15:00:14 <adamw> #startmeeting Fedora QA meeting
15:00:14 <zodbot> Meeting started Mon Jul 11 15:00:14 2022 UTC.
15:00:14 <zodbot> This meeting is logged and archived in a public location.
15:00:14 <zodbot> The chair is adamw. Information about MeetBot at https://fedoraproject.org/wiki/Zodbot#Meeting_Functions.
15:00:14 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:00:14 <zodbot> The meeting name has been set to 'fedora_qa_meeting'
15:00:16 <adamw> #meetingname fedora-qa
15:00:16 <zodbot> The meeting name has been set to 'fedora-qa'
15:00:17 <adamw> #topic Roll Call
15:00:35 <adamw> ahoyhoy folks, who's around for meeting fun?
15:00:41 <lruzicka> .hello2
15:00:42 <andi89gi[m]> Mornings me :)
15:00:43 <zodbot> lruzicka: lruzicka 'Lukáš Růžička' <lruzicka@redhat.com>
15:01:00 <andi89gi[m]> .hello2
15:01:01 <zodbot> andi89gi[m]: Sorry, but user 'andi89gi [m]' does not exist
15:01:38 <nielsenb> I'm here
15:01:59 <adamw> andi89gi: it's okay, we're very open-minded, non-existent folks are welcome here
15:02:18 <andi89gi[m]> adamw: weird - on IRC it may work
15:02:26 <andi89gi[m]> but thx for welcoming :)
15:03:51 <adamw> andi89gi: it works if your irc nick is the same as your FAS account, iirc, or possibly also if your IRC nick is the one registed in FAS
15:03:53 <lruzicka> andi89gi[m], I do not think the [m] is part of your FAS, is it?
15:04:29 <nielsenb> There also appears to be a space in the username zodbot looked up?
15:04:36 <nielsenb> .hello2
15:04:37 <zodbot> nielsenb: nielsenb 'Brandon Nielsen' <nielsenb@jetfuse.net>
15:04:50 * coremodule is here
15:04:51 <Southern_Gentlem> .hello jbwillia
15:04:52 <zodbot> Southern_Gentlem: jbwillia 'Ben Williams' <vaioof@gmail.com>
15:05:00 <andi89gi[m]> nielsenb: have to check it
15:06:55 <andi89gi> .hello2
15:06:56 <zodbot> andi89gi: Sorry, but user 'andi89gi' does not exist
15:07:10 <adamw> morning morning
15:07:31 <andi89gi[m]> weird issues  anyway morning
15:08:27 <adamw> alrighty, let's get rolling
15:08:48 <adamw> #topic Previous meeting follow-up
15:09:36 <adamw> #info "lruzicka to add question about exact required automated test coverage for desktop apps to one of the ongoing discussions with desktop team (mailing list or ticket)" - I believe https://lists.fedoraproject.org/archives/list/desktop@lists.fedoraproject.org/message/LOZ3XKUWX7DZCLCKNK4XVJYWXO2QLAE6/ was this, though we really meant for it to go to an existing thread or ticket
15:10:38 <lruzicka> I sent an email to the Desktop list and the only person who responded to it was Chris Murphy who advised me to look into https://pagure.io/fedora-workstation/issue/310
15:11:02 * kparal is here
15:11:04 <lruzicka> There is a similar discussion going on, adamw is part of it, too.
15:11:27 <adamw> yes, at the time this action item was added it was meant as a follow-up from QA to that existing discussion
15:12:02 <lruzicka> But until now, there has not been any progress. Sumantro promised to start working on those test cases, but I do not have any report on how far he has proceeded.
15:12:03 <adamw> i had brought up the existing discussion in a QA meeting, and 'ask about exact required automated test coverage' was what we decided to do in that discussion :)
15:12:34 <adamw> sumantro: ahoy?
15:14:05 <lruzicka> adamw, looks like a circular strategy then?
15:14:14 <adamw> =)
15:14:28 <adamw> i guess we'll try and follow up in the tickets again
15:14:29 <adamw> meantime we are working on the actual openqa tests, so hey!
15:14:46 <adamw> #topic Fedora 37 check-in and Change review
15:15:06 <lruzicka> adamw, ok ... it would be good if when sumantro works on those tests, that he somehow thinks about automation in the first place.
15:15:08 <adamw> so 37 is currently in 'mostly-working' state after a few hiccups last week
15:15:32 <adamw> composes are happening, tests are mostly passing, i'm not aware of any critical bugs
15:15:34 <adamw> lruzicka: agreed
15:16:49 <adamw> anyone have any major notes about current 37 state?
15:19:19 <nielsenb> I ran into this on a BIOS machine: https://bugzilla.redhat.com/show_bug.cgi?id=2102903#c1
15:19:30 <nielsenb> But I have only had a chance to check that one compose
15:20:01 <nielsenb> Which I see I apparently couldn't even be arsed to note what compose I tested...
15:20:55 <adamw> nielsenb: the sysctl issue was fixed a while back, though not sure if that's actually causing the problem there
15:20:59 <adamw> would be good to check again with a newer one
15:21:03 <adamw> i can try it later, though
15:21:04 <nielsenb> But basically the media check couldn't start, it worked in a VM and UEFI
15:22:16 <nielsenb> I'll check again, if nothing else I'll check the next nominated compose on the problematic machine
15:23:05 <adamw> thanks
15:23:32 <coremodule> I have something for F37
15:23:50 <adamw> fire away
15:24:48 <coremodule> As of F37, Fedora ARM has dropped support for 32-bit images (armhfp). We no longer compose a 32-bit image at all. That said, we still have armhfp columns in the validation matrix. I'd like to remove them, but want to get the official go ahead before I do.
15:25:04 <coremodule> Same for IoT
15:25:43 <coremodule> https://fedoraproject.org/wiki/Changes/RetireARMv7
15:25:58 <coremodule> and
15:25:58 <coremodule> https://pagure.io/fesco/issue/2703
15:26:28 <adamw> coremodule: oh, right. yeah, let's remove them
15:26:48 <coremodule> boom.
15:26:52 <coremodule> alright, I'll get it done
15:26:56 <adamw> do it in the templates, of course...i guess you know how it works?
15:27:01 <coremodule> that's all I had.
15:27:15 <coremodule> yeah, gotcha. I'll modify the templates
15:27:35 <adamw> #action coremodule to remove 32-bit arm columns from validation matrix templates
15:27:45 <adamw> i guess i'll check through the release criteria for anything that should be updated there
15:28:14 <coremodule> I think it's already been modified since we moved all the 32bit testcases to optional, but good idea. would be worth a look
15:29:08 <adamw> so, the other thing i wanted to note in this slot: there's a potential problem with python 3.11 and the f37 release cycle, see this ticket: https://pagure.io/fesco/issue/2825
15:29:41 <adamw> #info there's a potential issue with delays to the python 3.11 release and the fedora 37 release schedule, see https://pagure.io/fesco/issue/2825 for the discussion on that
15:30:29 <adamw> the plan that seems to be getting support is to keep going to branched and hope by then it's clear 3.11 will be abi-stable in time for the f37 release, if not we have to go back to 3.10. none of the choices is great but that seems the least-worst
15:30:31 <adamw> just wanted folks to be aware
15:30:53 <lruzicka> noted
15:31:46 <nielsenb> Sounds reasonable
15:33:11 <adamw> on the Change front, i don't see anything super significant for us since the last check-in. there's https://fedoraproject.org/wiki/Changes/RaspberryPi4 , we do have some pi 4s for testing already right?
15:33:43 <andi89gi[m]> .hello andilinux
15:33:44 <zodbot> andi89gi[m]: andilinux 'Andi Artz' <Kartz@gmx.net>
15:33:57 <nielsenb> Woo!
15:35:00 <andi89gi[m]> yea managed it
15:35:08 <adamw> yay
15:35:18 <nielsenb> I think I have a Pi4 around here I can test against
15:35:34 <andi89gi[m]> nielsenb: I have a Pi4 as well
15:36:33 <adamw> oh, of course, we should also note that it's now proposed to make CentOS an official edition for F37
15:36:42 <adamw> grr
15:36:45 <adamw> CoreOS. not CentOS. :D
15:37:07 <nielsenb> Media check still doesn't work on this BIOS machine, just tried the 20220711 compose
15:37:13 <adamw> #info CoreOS is proposed as an official Edition for F37: https://lists.fedoraproject.org/archives/list/devel@lists.fedoraproject.org/message/OSHZEHWTYSACVKJEPLVL2EVJ3GDY3AE2/
15:37:30 <nielsenb> CentOS Core Edition when?
15:37:40 <adamw> this obviously comes with testing implications, but the current intent is that we will ensure CoreOS' existing CI is sufficient to demonstrate that it complies with the release criteria
15:38:06 <adamw> nielsenb: do the visible messages look different now? i expect they would
15:38:33 <nielsenb> Nope
15:38:58 <nielsenb> "sh: line 1: /sbin/sysctl: No such file or directory"
15:39:03 <adamw> #info the plan for CoreOS testing is to rely on its CI, and ensure that the CI tests cover all of the release criteria. there is a ticket for this cross-check: https://github.com/coreos/fedora-coreos-tracker/issues/1239
15:39:08 <nielsenb> Then the media check failed message
15:39:25 <adamw> nielsenb: that's expected, but there should not be any module load errors any more
15:39:38 <adamw> sysctl not being there is still a thing, but it's not failing anything
15:40:50 <nielsenb> I never saw module load errors, only the media check failing
15:41:16 <adamw> ah k, so you already had the fix for the sysctl thing. so, it's something else causing the media check to fail then
15:41:37 <nielsenb> Unfortunately I see no other relevant diagnostic information
15:41:52 <nielsenb> Just the checkiso job failed, then everything gets garbled
15:43:08 <adamw> fun. i'll take a look at it later
15:43:29 <adamw> anything else on F37?
15:46:35 <adamw> alright
15:46:38 <adamw> @topic Outstanding proposals
15:47:39 <adamw> so, the networking proposal is still technically outstanding. i don't think i'm going to add any dnssec requirements, so i don't think there's anything to change
15:47:53 <adamw> if nobody has any objections i'll go ahead and commit the latest draft to the wiki later today
15:50:24 <adamw> #action adamw to implement the latest draft of the proposed networking criteria
15:50:47 <adamw> matthew's proposed 'default app functionality' changes are still technically pending, but that's kinda tied up in the larger discussion with desktop team
15:51:45 <adamw> #info let's skip the test day / community event item as sumantro's not around
15:51:48 <adamw> #topic Open floor
15:51:51 <adamw> any other business, folks?
15:53:36 <nielsenb> I started implementing "back" in the relval report-results flow, then decided I didn't want to break anything so I've started writing unit tests for that flow instead
15:53:47 <nielsenb> Expect an ugly MR "soon"(tm)
15:54:41 <adamw> nielsenb: thanks a lot! i appreciate it
15:54:48 <adamw> been wanting to do that for years but never got the roundtuits
15:54:58 <adamw> i don't suppose you want to write unit tests for everything else too? ;)
15:55:07 <nielsenb> One (maybe two) things at a time
15:55:18 <nielsenb> I'm less familiar with the other flows
15:55:38 <adamw> it was worth a try
15:55:39 <nielsenb> Then come tests for wikit itself...
15:56:29 <adamw> wikitcms does have tests,
15:56:50 <nielsenb> Well then, I'll consider that done
15:57:04 <adamw> 95% coverage, the remaining bits are a bit tricky iirc.
15:57:28 <adamw> alrighty, anybody else got anything?
15:59:16 <adamw> if not, thanks for coming, everyone!
16:00:31 <adamw> #endmeeting