16:01:34 <adamw> #startmeeting Fedora QA Meeting
16:01:34 <zodbot> Meeting started Mon Mar  2 16:01:34 2020 UTC.
16:01:34 <zodbot> This meeting is logged and archived in a public location.
16:01:34 <zodbot> The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:01:34 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:01:34 <zodbot> The meeting name has been set to 'fedora_qa_meeting'
16:01:38 <adamw> #meetingname fedora-qa
16:01:38 <zodbot> The meeting name has been set to 'fedora-qa'
16:01:42 <adamw> #topic Roll call
16:01:43 <bcotton> .hello2
16:01:44 * kparal is here
16:01:44 <zodbot> bcotton: bcotton 'Ben Cotton' <bcotton@redhat.com>
16:01:45 <adamw> morning morning
16:01:47 <adamw> how's everyone doing?
16:02:09 <frantisekz> .hello2
16:02:10 <zodbot> frantisekz: frantisekz 'František Zatloukal' <fzatlouk@redhat.com>
16:02:22 <onuonga> It is my first time here.
16:02:32 <onuonga> So I am great
16:02:40 <coremodule> .hello2
16:02:41 <zodbot> coremodule: coremodule 'Geoffrey Marr' <gmarr@redhat.com>
16:02:41 <frantisekz> welcome then onuonga :)
16:02:49 <onuonga> Thank you.
16:02:52 <coremodule> good morning/afternoon/evening!
16:03:27 <lruzicka> .hello2
16:03:28 <zodbot> lruzicka: lruzicka 'Lukáš Růžička' <lruzicka@redhat.com>
16:03:50 * pwhalen is here
16:03:59 <cmurf> .hello chrismurphy
16:04:00 <zodbot> cmurf: chrismurphy 'Chris Murphy' <bugzilla@colorremedies.com>
16:05:18 <jlanda> .hello2
16:05:19 <zodbot> jlanda: jlanda 'Julen Landa Alustiza' <julen@landa.eus>
16:05:40 <jlanda> heya, but I gtg in a few :S
16:06:02 <adamw> we'll miss you
16:07:09 <adamw> #topic Previous meeting follow-up
16:07:39 <adamw> so we have:
16:07:40 <adamw> "frantisekz, lruzicka, coremodule to review and ack/nack/patch kparal's
16:07:40 <adamw> proposal"
16:07:53 * coremodule responded on the list
16:07:59 <adamw> coremodule, tip for next time - try and write action items that are comprehensible without context :P
16:08:09 <adamw> (i mess this up all the time, but i try!)
16:08:11 <adamw> what was the proposal?
16:08:14 <kparal> coremodule: but you didn't reply to my reply!
16:08:17 <lruzicka> adamw, which one, the one with the application?
16:08:21 <kparal> the proposal for reducing desktop requirements
16:09:06 <adamw> lruzicka: i don't know, i didn't run the meeting last week :P
16:09:08 <adamw> ah, that one
16:09:50 <lruzicka> I replied to the list already.
16:10:10 <lruzicka> But I did not explicitely say +1 or something, do you want me to?
16:10:29 <lruzicka> I fully support it (for the record here).
16:10:45 <coremodule> I'm trying to find it in hyperkitty for context
16:11:15 <lruzicka> coremodule, look for proposal: Default application functionality criterion reduction
16:11:17 <kparal> there was a bit of a discussion in the kde list. perhaps not everyone saw that
16:11:30 <kparal> test list: https://lists.fedoraproject.org/archives/list/test@lists.fedoraproject.org/thread/QEYDNIJQYPXZTXDMSMLF2JQLLZMZMCTM/
16:11:32 <lruzicka> oh, i did not
16:11:33 <tflink> coremodule: if you have the email, there is an archive link in the headers
16:11:36 <kparal> kde list: https://lists.fedoraproject.org/archives/list/kde@lists.fedoraproject.org/thread/HNXF7WCYQ7RAKMNGCIOXCBXERBAK44TK/
16:11:44 <coremodule> https://lists.fedorahosted.org/archives/list/kde@lists.fedoraproject.org/message/HNXF7WCYQ7RAKMNGCIOXCBXERBAK44TK/
16:11:57 <kparal> coremodule: ninja'd
16:12:09 <coremodule> haha
16:12:37 <lruzicka> oh, it seems I responded in the kde list then :D
16:12:48 <adamw> kparal: so where do you think we stand on this?
16:12:54 <adamw> we do have a 'proposals' topic later
16:12:57 <adamw> so for now, just the action item
16:13:12 <kparal> it looks like nobody cares expect for Kevin Kofler
16:13:34 <lruzicka> and Kevin stopped arguing, so he might not care either.
16:14:02 <kparal> cmurf mentioned on IRC that desktop team is OK with the current proposal because it doesn't reduce anything from Workstation coverage
16:14:17 <kparal> but no reply on the list, so you have to trust me
16:14:44 <kparal> I might have caused kkofler a heart-attack or something :/
16:15:03 <kparal> but yes, he stopped talking to me
16:15:07 <lruzicka> kparal, he comes from nothern italy, you may well be innocent
16:15:18 <adamw> #info "frantisekz, lruzicka, coremodule to review and ack/nack/patch kparal's proposal" - they have replied, and discussions are ongoing! we will follow up on this topic later in the meeting
16:15:20 <adamw> how's that
16:15:26 <coremodule> +!
16:15:29 <coremodule> +1
16:15:32 <lruzicka> +?
16:15:38 <lruzicka> +1
16:15:41 <kparal> frantisekz was lazy and didn't reply, let's fire him instead
16:16:02 <kparal> but sure, let's move to the next topic
16:16:15 <frantisekz> +1
16:16:21 <frantisekz> didn't have time for it
16:16:42 <frantisekz> got more pressing things to solve first
16:17:46 <adamw> alrighty
16:18:02 <adamw> #topic Fedora 32 status
16:18:12 <adamw> so, we seem to be significantly less busted than the last two weeks, which is good news!
16:18:40 <cmurf> indeed
16:20:05 <adamw> #info Fedora 32 is currently composing and broadly working, there are no complete showstopper bugs. We have tabs on various blockers which will be discussed in the upcoming blocker review meeting
16:20:20 <adamw> the modularity upgrade mess is getting 'fixed' with a big dumb hammer
16:20:50 <adamw> can anyone think of any other systemic issues we have to think about, aside from blocker whack-a-mole?
16:21:10 * kparal doesn't have anything
16:22:55 <frantisekz> maybe we can test grub2 a little bit more
16:22:59 <frantisekz> dual boot , etc.
16:23:15 <frantisekz> there was grub2 upgrade from 2.02 to 2.04 with a lot of changes
16:23:25 <adamw> yep, that'd be good
16:23:28 <adamw> also testing EFI stuff would be good
16:23:48 <frantisekz> yeah... with 20k+ line diff backport as downstream patch in efivar
16:23:48 <adamw> as the 'fix' for the EFI-installs-on-VMs issue was basically 'here's two years of grubenv development we didn't quite get around to making a release for yet'
16:23:50 <frantisekz> ... :D
16:23:55 <adamw> thanks pjones!
16:24:00 <cmurf> i've heard that some EFI do goofy things with USB sticks
16:24:31 <cmurf> like, oh  you wanna boot off a USB stick, here are some extra options
16:24:36 <lruzicka> cmurf, that will be a difficult thing to test though, we do not have that many machines
16:24:46 <adamw> #info bootloader testing (especially UEFI) is suggested as an area of focus due to recent changes to grub2 and efivar
16:25:01 <cmurf> sure, it could be useful to find out the range of how bad it is (?) and get screenshots to docs team?
16:25:02 <lruzicka> so, we should probably make it a Test Day.
16:25:22 <adamw> lruzicka: it's no problem, just go down to the lobby and tell people they have to let you test on their laptop or else you'll talk to them about jenkins for an hour
16:25:35 <lruzicka> adamw, haha :D
16:25:35 <coremodule> sumantro, ^^^
16:25:37 <adamw> ...or we could do that, *fine*
16:26:11 <coremodule> cmurf, I will talk to sumantro about a EFI/USB test day
16:26:30 <cmurf> adamw haha :D
16:27:19 <pjones> adamw: efivar, not grubenv, and more like 5 months?  I mean, it's a fix I wrote in October but hadn't put in Fedora before the release because we weren't seeing the bug.
16:27:46 <adamw> pjones: well, the thing is you put all the other changes in too, not just that fix :P
16:28:13 <pjones> well, yes.
16:28:19 <adamw> the previous build had 10 patches, the new one has 86, iirc.
16:28:21 <pjones> except the ones I left out
16:28:40 <adamw> it's a lot of change, is the point i'm making
16:28:42 <pjones> sure.
16:29:20 <adamw> alrighty
16:29:23 <adamw> moving along
16:29:40 <adamw> I'm gonna split up the next topic on the fly because it's really two topics
16:29:43 <adamw> #topic Outstanding proposals
16:30:39 <adamw> #info the proposal to replace Xfce-on-armv7hfp with Workstation-on-aarch64 as the release-blocking ARM desktop has been approved and implemented
16:30:48 <adamw> i updated the criteria and the matrices for that last week
16:31:01 <adamw> pwhalen, does everyone look okay there to you?
16:31:04 <pwhalen> many thanks adamw
16:31:44 * pwhalen is looking
16:32:01 <kparal> I saw the diffs, looked ok
16:32:58 <pwhalen> The install template might need to be adjusted to include workstation rather than Xfce as basic
16:33:06 <adamw> pwhalen: oh yes, good call
16:33:07 <pwhalen> and.. another arch.
16:33:09 <adamw> didn't think about that one
16:33:28 <adamw> #action adamw to update installation matrix template for workstation-on-aarch64 change
16:33:30 <pwhalen> and Base
16:33:34 <adamw> #undo
16:33:34 <zodbot> Removing item from minutes: ACTION by adamw at 16:33:28 : adamw to update installation matrix template for workstation-on-aarch64 change
16:33:40 <adamw> #action adamw to update installation and base matrix templates for workstation-on-aarch64 change
16:33:43 <adamw> :P
16:34:17 <pwhalen> Maybe just add a (aarch64) beside workstation. Otherwise.. looks good. thanks
16:34:25 <adamw> #info the proposal to reduce the coverage of the default application functionality criterion is still open with ongoing discussion
16:35:47 <kparal> should I re-ping the thread again?
16:36:10 <kparal> currently nothing happens there
16:36:18 <cmurf> i have a suggestion :)
16:36:34 <cmurf> somehow make it consequential instead of inviting discussion
16:36:34 <adamw> yeah, i'd maybe do the 'summarize current state and say what's going to happen in a week if no-one yells too hard' thing
16:37:18 <kparal> cmurf: what do you mean exactly?
16:37:40 <kparal> adamw: will do
16:38:21 <adamw> does anyone here have any concerns/ideas/thoughts about that proposal or are we happy just for kparal to take it forwards? i'm happy with that plan
16:38:24 <cmurf> maybe a dozen single line sentences instead of a dozen paragraphs?
16:38:28 * cmurf ducks and runs
16:39:14 <cmurf> like I read it, but i'm not sure many people read it, or if they pretty much only use Workstation and thus don't have much input?
16:39:56 <kparal> I'll make a short summary then
16:40:28 <cmurf> i realize that's hard because there's also choices
16:40:28 <kparal> I might add a tldr version next time to the top, good idea
16:40:34 <adamw> #action kparal to move default app functionality reduction proposal forward by summarizing current state and declaring intent to make changes shortly
16:40:40 <adamw> sound good?
16:40:44 <kparal> yep
16:40:44 <lruzicka> yes
16:41:27 <adamw> ok
16:41:42 <adamw> #topic IoT and CoreOS
16:41:56 <adamw> so, i figured we should check in on where we are with handling this two for Fedora 32
16:42:23 <adamw> it looks like coremodule and pwhalen are making progress with IoT?
16:42:47 <coremodule> aye, we will have some test cases specific for IoT in the next week or so
16:42:57 <coremodule> and optional test cases as time permits
16:43:16 <coremodule> we have release criteria for IoT
16:43:38 <coremodule> https://pagure.io/fedora-qa/issue/618
16:43:39 <kparal> we should probably link IoT criteria from our QA release criteria
16:43:47 <coremodule> https://pagure.io/fedora-qa/issue/620
16:43:55 <kparal> I might create another ticket for that
16:44:13 <adamw> yeah, agreed
16:44:17 <kparal> ok
16:44:20 <coremodule> kparal, maybe we just lump that into the existing IoT ticket
16:44:27 <coremodule> its all one-in-the-same
16:44:41 <coremodule> and the ultimate goal anyway
16:45:00 <kparal> I think I prefer a separate one. It's a simple change that can be done immediately
16:45:13 <adamw> i don't really love the IoT criteria being an entirely separate thing in an entirely separate system, but if that's what people want...
16:45:33 <kparal> adamw: do you want to move their source to our wiki?
16:45:45 <coremodule> as the person who will be working on it, I prefer less tickets. less clutter.
16:45:54 <lruzicka> If IoT is part of Fedora, it should be all in one place, I believe.
16:45:55 <adamw> kparal: i dunno
16:46:02 <pwhalen> adamw, doesnt have to be. it was always considered an addition to the existing criteria. It add a few testcases
16:46:35 <adamw> we can probably discuss it outside the meeting, i'll comment on the tickets after i think it through a bit more
16:46:44 <lruzicka> sure
16:46:57 <pwhalen> sounds good
16:47:22 <adamw> #info initial IoT release criteria have been written and published at https://docs.fedoraproject.org/en-US/iot/release-criteria/
16:47:43 <adamw> #info work on writing IoT test cases is being tracked at https://pagure.io/fedora-qa/issue/618
16:48:04 <adamw> the other thing we're gonna need is result tracking
16:48:28 <adamw> i can see two ways to do that, broadly speaking: a new parallel set of matrices for Fedora-IoT, or a Shiny New System
16:48:47 <adamw> we cannot just add to the existing matrices, AIUI, because IoT is going to have a separate release process
16:49:16 <adamw> we are not building IoT images in the main 'Fedora' composes, but in a separate 'Fedora-IoT' compose
16:49:26 <kparal> does that mean separate timeline?
16:49:32 <kparal> I thought it's just built differently
16:49:46 <adamw> separate timeline is at least potentially a thing, aiui
16:49:56 <adamw> and even if it isn't, the composes are still separate
16:50:10 <pwhalen> +1 for parallel set of matrices, seems a little late for new shiny
16:50:21 <adamw> we literally will not have a single compose that contains both the IoT images and the other images for 'Fedora 32 Beta' or 'Fedora 32'
16:50:29 <adamw> even if we expect to sign them off and release them together
16:50:36 <kparal> well, we're going to have a separate timeline with coreos, so having IoT separate as well doesn't really matter much
16:50:45 <adamw> and the design of wikitcms is fundamentally tied to the concept that 'we're testing a compose'
16:51:05 <adamw> kparal: with coreos it's a bit different again because we don't even have 'composes' for coreos, in the pungi sense at least
16:51:08 <adamw> everything is fun!
16:51:32 <pwhalen> thankfully.. lots of time :P
16:51:43 <adamw> for New Shiny, i was vaguely thinking that IoT might make a good target for sumantro's plan to do a test deployment of...uh...which one was it you were going to do, sumantro? :)
16:51:52 <adamw> pwhalen: har
16:52:00 <adamw> kiwi?
16:52:45 <adamw> #info work on CoreOS test cases is ongoing at https://pagure.io/fedora-qa/issue/617
16:52:51 <tflink> yeah, he was talking about looking at kiwi tcms
16:52:59 <adamw> #info a CoreOS Test Day is also planned: https://pagure.io/fedora-qa/issue/616
16:53:08 <adamw> so yeah, IoT might make a decent target for that
16:53:13 <cmurf> does the "in a separate 'Fedora-IoT' compose" look the same as Silverblue?
16:53:18 <adamw> it would be much less work than trying to redo the entire validation process right away
16:53:56 <adamw> cmurf: no? we build Silverblue as part of the main compose
16:54:11 <adamw> cmurf: it's more like the post-release Cloud composes we do and the post-release Atomic composes we used to do
16:54:32 <adamw> but the difference is it's being used for Rawhide and Branched as well, it's not just a post-release thing
16:54:44 <adamw> if you check https://kojipkgs.fedoraproject.org/compose/cloud/ there are only F31 builds there
16:55:11 <adamw> but in https://kojipkgs.fedoraproject.org/compose/iot/ there's 31 (i think they might've stopped building 31 now?), 32 and 33
16:55:23 <adamw> and the main Branched and Rawhide composes contain no IoT images
16:55:40 <lruzicka> so there won't be any IoT composes throughout the year? just to make sure I understand.
16:55:42 <cmurf> adamw: ok, how does Fedora CoreOS compare? Totally different too?
16:55:53 <adamw> cmurf: CoreOS is doing something conceptually the same, technically different
16:55:59 <adamw> CoreOS doesn't build as Pungi composes at all
16:56:06 <adamw> it has a whole separate build process i don't really know the details of yet
16:56:11 <cmurf> I see...
16:56:14 <adamw> (nor can I remember offhand where the actual bits live)
16:57:06 <adamw> #action adamw to file a ticket and figure out how to handle IoT test result tracking
16:57:17 <cmurf> I don't know the entire implication of what it means to "rebase X on CoreOS" but I think that's the idea for IoT
16:57:24 <adamw> the *good* thing about CoreOS is the whole test/release loop is highly automated
16:57:38 <adamw> so it matters a bit less whether we (QA) have it all nicely tied into the formal validation process
16:57:57 <cmurf> and expressly stated for Silverblue, but I don't know the time frame.
16:58:00 <adamw> there's less, you know, icky physical hardware involved
16:58:25 <adamw> cmurf: that's probably the ostree thing where they sort of decide that everyone gets a major version upgrade?
16:58:27 <pwhalen> LOL :_
16:58:42 <adamw> like, 'ok today's the day everything goes to 31' or whatever
16:58:50 <adamw> oh hey look, we're out of time!
16:58:52 <adamw> i am good at this
16:58:58 <cmurf> yeah to be continued
16:59:19 <adamw> #topic Community event status
16:59:22 <adamw> so very briefly!
16:59:29 <cmurf> I'm just wondering what New Shiny looks like if there's a bunch of other cattle coming along for the ride
17:00:22 <adamw> #info i18n Test Day is scheduled for 2020-03-10
17:00:26 <adamw> so put that in yer calendars
17:00:31 <adamw> I don't see anything else coming up right away
17:00:36 <adamw> #topic Open floor
17:00:41 <adamw> anything else super-urgent, folks?
17:01:05 * kparal shakes his head
17:01:07 <frantisekz> blocker review :D
17:02:41 <cmurf> wharrcoffeetimefirst!
17:02:52 <adamw> okay, thanks for coming everyone
17:02:57 <adamw> sorry for standard awful time management :P
17:03:04 <cmurf> worked out ok i think
17:03:25 <onuonga> It was nice to attend. I got a feel of how things run :-)
17:04:23 <bcotton> adamw: don't worry. standard time ends sunday in the US :p
17:04:25 <lruzicka> take care, I am moving to the other room.
17:05:38 <adamw> =)
17:05:39 <adamw> #endmeeting