17:00:52 <adamw> #startmeeting F27-blocker-review
17:00:52 <zodbot> Meeting started Mon Nov  6 17:00:52 2017 UTC.  The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
17:00:52 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
17:00:52 <zodbot> The meeting name has been set to 'f27-blocker-review'
17:00:52 <adamw> #meetingname F27-blocker-review
17:00:52 <adamw> #topic Roll Call
17:00:52 <zodbot> The meeting name has been set to 'f27-blocker-review'
17:01:01 <adamw> morning folks, who's around for some blocker review funtimes?
17:01:45 * siddharthvipul1[ is here
17:01:46 <Kohane> Morning...?
17:01:51 * lbrabec is here
17:01:55 * Kohane is here
17:01:59 * pwhalen is here
17:02:15 * kparal is here
17:02:15 <Kohane> adamw: is evening almost night for me....
17:02:22 * coremodule is here!
17:02:41 <sgallagh> .hello2
17:02:43 <zodbot> sgallagh: sgallagh 'Stephen Gallagher' <sgallagh@redhat.com>
17:02:52 * nirik is lurking around
17:03:03 <adamw> Kohane: PST is the one true timezone
17:03:32 <Kohane> adamw: PST? Is that a new chemical product?
17:03:42 <adamw> .fire kohane
17:03:42 <zodbot> adamw fires kohane
17:03:51 <Kohane> LOL
17:04:07 <adamw> #chair sgallagh lbrabec
17:04:07 <zodbot> Current chairs: adamw lbrabec sgallagh
17:04:22 <adamw> impending boilerplate alert!
17:04:23 <adamw> #topic Introduction
17:04:23 <adamw> Why are we here?
17:04:23 <adamw> #info Our purpose in this meeting is to review proposed blocker and nice-to-have bugs and decide whether to accept them, and to monitor the progress of fixing existing accepted blocker and nice-to-have bugs.
17:04:24 <adamw> #info We'll be following the process outlined at:
17:04:24 <adamw> #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting
17:04:26 <adamw> #info The bugs up for review today are available at:
17:04:28 <adamw> #link http://qa.fedoraproject.org/blockerbugs/current
17:04:30 <adamw> #info The criteria for release blocking bugs can be found at:
17:04:32 <adamw> #link https://fedoraproject.org/wiki/Basic_Release_Criteria
17:04:34 <adamw> #link https://fedoraproject.org/wiki/Fedora_27_Beta_Release_Criteria
17:04:36 <adamw> #link https://fedoraproject.org/wiki/Fedora_27_Final_Release_Criteria
17:04:38 <adamw> who wants to secretarialize?
17:04:45 <coremodule> I'll do it!
17:04:57 <frantisekz> .hello2
17:04:57 <zodbot> frantisekz: frantisekz 'František Zatloukal' <fzatlouk@redhat.com>
17:05:19 * sumantro is here
17:06:18 <adamw> #info coremodule will secretarialize
17:06:24 <adamw> hi sumantro and frantisekz
17:06:52 <frantisekz> hey adamw
17:07:02 <sumantro> hi adamw
17:08:01 <adamw> #info starting with the one proposed Final blocker
17:08:03 <adamw> #topic (1494586) [abrt] gnome-shell: g_type_check_instance_is_fundamentally_a(): gnome-shell killed by SIGSEGV
17:08:04 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1494586
17:08:04 <adamw> #info Proposed Blocker, gnome-shell, NEW
17:09:11 * Kohane is reading the bug
17:09:24 <adamw> oh goody, more shell crashing
17:09:58 * adamw goes looking for a Shell folk
17:10:14 <coremodule> Petrol!
17:10:26 <coremodule> whoops,wrong Shell
17:10:41 <Kohane> LOL
17:10:45 <sgallagh> coremodule: Time for a new password
17:12:22 <adamw> upstream report: https://bugzilla.gnome.org/show_bug.cgi?id=789484
17:14:46 <Kohane> adamw: Same bug?
17:15:49 <adamw> i think so, yes.
17:17:11 * mattdm is here
17:17:22 <mattdm> (and in another meeting, but there's life)
17:18:19 <sgallagh> GNOME Shell bugs are unfortunate, but I'm not sure we want to set a precedent on blocking for every random crash.
17:18:36 <sgallagh> Though the fact that it takes the whole session with it is particularly ugly.
17:19:21 <Kohane> TBH...  I don't like it crashing but....  I'm not sure is a blocker.
17:20:16 <Kohane> I mean, it happens. And Gnome Wayland doesn't restart the session as Gnome with Xorg does, but...  it's something that from to time happens.
17:20:48 <adamw> sgallagh: *all* shell crashes on wayland take out the whole session, which is a big issue since we switched to wayland by default
17:20:50 <sgallagh> Sure; at the same time, for the people to whom it's happening it is frequent
17:20:54 <frantisekz> ... so, let's switch to Xserver by default and those crashes won't matter that much... :)
17:21:04 <adamw> that makes me want to block on sufficiently-common shell crashes.
17:21:17 <sgallagh> Yeah, that's why I'm fence-riding here
17:21:19 <adamw> but that can open a can of worms.
17:21:20 * Kohane likes frantisekz idea...
17:21:47 <adamw> the bug also references https://bugzilla.redhat.com/show_bug.cgi?id=1469129 , which looks similar but not quite the same
17:21:56 <adamw> that one has 26 CCs
17:22:09 <adamw> this one has 27 CCs
17:23:32 <Kohane> So... what do we do with this?
17:23:43 <kparal> and FAF numbers seem to be broken
17:23:58 <adamw> 4 come from assigness etc. , so that looks like ~22 and 23 affected folks
17:24:10 <adamw> which seems quite a lot for a beta
17:24:55 <Kohane> Yes, it is.
17:25:12 <Kohane> Well, I suppose it happens too often to ignore.
17:25:15 <adamw> i'm going to see if the alleged reproducer in https://bugzilla.redhat.com/show_bug.cgi?id=1494586#c26 works for me
17:25:22 <Kohane> ok
17:25:45 <adamw> yeah, it'd help if faf was more reliable here
17:25:49 <adamw> or if we just knew how the heck to use it
17:25:55 <adamw> (if that's the problem)
17:26:01 <mattdm> there *is* a gnome 3.26.2 update in the wings....
17:26:21 <adamw> mattdm: the upstream bug hasn't been closed or given any attention, though, so not sure if i'd expect that to fix anything
17:26:29 <sgallagh> mattdm: Are you legitimately suggesting pulling in a megaupdate during Freeze?
17:26:34 <frantisekz> but no mention about fixing this issue in 3.26.2
17:26:38 <sgallagh> I just want to make sure I don't need to go to the optometrist.
17:26:50 <mattdm> sgallagh: no. I am not suggesting that
17:28:40 * sgallagh tries to round up cschalle
17:28:57 <adamw> i'm already talking to mclasen in #fedora-desktop
17:29:05 <adamw> <mclasen> we discussed it earlier
17:29:05 <adamw> <mclasen> seems another instance of 'unclear bug'
17:29:12 <adamw> <mclasen> I've pointed fmuellner at it
17:29:31 <sgallagh> ah
17:29:41 <adamw> also from earlier:
17:29:42 <adamw> <fmuellner> 07:46:37> the backtrace isn't very enlightening I'm afraid
17:29:42 <adamw> <fmuellner> 07:46:55> other than that it's something entirely different than the one about maximized windows
17:29:52 <adamw> <fmuellner> 07:47:27> (it looks like something is trying to set an invalid gobject property on a StLabel, but it's too generic to tell anything more)
17:29:52 <adamw> <fmuellner> 07:48:38> maybe we should consider turning on https://bugzilla.gnome.org/show_bug.cgi?id=789237 though
17:29:52 <adamw> <fmuellner> 07:49:00> (that bug adds an env variable to dump the JS backtrace on segfaults)
17:31:37 <sgallagh> So in short, this is serious but we have no idea what's causing it or how long it might take to fix it?
17:32:08 <otaylor> sgallagh: I don't think we can fix it from the abrt stack traces
17:32:10 <adamw> that sounds...approximately correct, yeah.
17:32:24 <sgallagh> This is fine.
17:32:44 <adamw> otaylor: that commit to get a gjs trace when shell crashes would presumably help?
17:33:28 <otaylor> sgallagh: really needs a reproducer, or someone who sees it frequently  getting some detailed debugging help
17:33:41 <otaylor> yes, getting a gjs stack trace would go a long ways
17:33:58 <sgallagh> I'm going to see if I can repro it; one sec.
17:34:44 <adamw> man, this hot shiny ball in the sky is really annoying.
17:34:58 <sgallagh> Ah, wait. I'm on X11 on this sytem.
17:35:01 <sgallagh> crud
17:35:21 <Kohane> adamw: what shiny ball? are you alright?
17:35:36 <otaylor> sgallagh: is the crash wayland specific, or just the crash taking down the session?
17:35:47 <sgallagh> otaylor: Good question
17:35:51 * sgallagh checks his ABRT history
17:35:51 <adamw> Kohane: i hear it's some star we're orbiting or something like that.
17:36:05 <otaylor> But I'm not expecting it to be as simple as "notification crashes gnome-shell"
17:36:10 <Kohane> adamw: Oh. I heard that too.
17:36:22 <adamw> otaylor: there's a claimed reproducer in https://bugzilla.redhat.com/show_bug.cgi?id=1494586#c26
17:36:30 <sgallagh> OK, I haven't seen it on this system.
17:36:43 <Kohane> otaylor: I think it happens in X too, but while Xorg restarts the session Wayland just crashes.
17:36:47 <sgallagh> Does it happen on Live sessions?
17:36:48 <adamw> https://bugzilla.redhat.com/show_bug.cgi?id=1494586#c13 claims it happens on X too.
17:37:06 <sgallagh> If so, I'll throw it on this spare laptop I just found under a rock
17:37:13 <adamw> i'm doing that atm
17:37:14 <Kohane> Yeah, the first link has a comment saying it happens with Xorg as well
17:37:19 <adamw> (trying from a live session on a laptop)
17:37:34 <kparal> I tried it in VM, but it seems suspend doesn't place nice with libvirt. can't be resumed
17:38:00 <otaylor> If it's not actually reproducible for others, probably just need to get fmuellner talking to Berend who seems to (for whatever reason) be seeing it frequently and able to reproduce it.
17:39:05 <adamw> not crashing it here.
17:39:31 <adamw> i wish we had a better framework for evaluating this kinda bug. :/
17:39:55 <sgallagh> adamw: I'm slightly leaning towards "Not a blocker because Fedora QA can't reproduce it" :)
17:40:14 <adamw> heh
17:41:03 <adamw> i guess i'm reluctantly leaning towards -1 on the basis that it's clearly not affecting everybody and we don't have any kinda confidence at this point that it's fixable in a reasonable time frame (per that wording we introduced last cycle)
17:41:22 <adamw> it's really hard to say just how many people need to be seeing something like this to block, though
17:41:32 <adamw> 20+ seems like...a lot of people? but is it too many? i dunno.
17:41:43 <adamw> otaylor: oh, btw, do you think https://bugzilla.redhat.com/show_bug.cgi?id=1469129 is related?
17:41:54 <Kohane> I think over 20 people is a lot.
17:42:22 <otaylor> adamw: I see no reason to think it is related
17:42:40 <adamw> okay
17:42:49 <adamw> so that's just another crash with a bunch of people hitting it...whee.
17:43:12 <sgallagh> I wish we had a hardware inventory.
17:43:30 <sgallagh> So we could at least see if it was common to Nvidia or something.
17:43:31 <adamw> i wish i had a solid gold toilet.
17:43:35 <adamw> yeah, that'd help.
17:44:33 <adamw> so, i guess we talked this through about as hard as we can...votes?
17:45:02 <kparal> -1 atm
17:45:02 <sgallagh> I'm a weak -1 unless we get more information that it's going to hit a big percentage of users
17:45:36 <Kohane> -1
17:45:40 <lbrabec> yea, -1 for now
17:45:44 <frantisekz> -1
17:46:41 <pwhalen> agreed, -1
17:46:47 <sgallagh> I'll continue trying to get it to fail on my laptop as well
17:47:13 <adamw> proposed #agreed 1494586 - RejectedBlocker (Final) - this is concerning as it's hitting quite a few people for a Beta. However, clearly not everyone is affected, and neither a reliable reproducer nor a clear cause analysis have been found yet. As this is being raised very late in the cycle, we feel we can only block the release on it with definition information indicating it will affect a broad range of users
17:47:46 <kparal> ack
17:47:49 <lbrabec> ack
17:47:50 <sumantro> ack
17:47:56 <sgallagh> ack
17:48:05 <sgallagh> wait, patch
17:48:11 <sgallagh> s/definition/definite/
17:48:16 <Kohane> ack
17:48:42 <adamw> #agreed 1494586 - RejectedBlocker (Final) - this is concerning as it's hitting quite a few people for a Beta. However, clearly not everyone is affected, and neither a reliable reproducer nor a clear cause analysis have been found yet. As this is being raised very late in the cycle, we feel we can only block the release on it with definite information indicating it will affect a broad range of users
17:48:46 <adamw> thanks sgallagh
17:48:48 <adamw> sgallagh++
17:48:49 <zodbot> adamw: Karma for sgallagh changed to 20 (for the f26 release cycle):  https://badges.fedoraproject.org/tags/cookie/any
17:48:54 <adamw> cookie for paying attention :P
17:48:58 <sgallagh> heh
17:49:11 <adamw> we have a couple of proposed FEs
17:49:13 <adamw> #topic (1509996) Install color emoji fonts by default
17:49:13 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1509996
17:49:13 <adamw> #info Proposed Freeze Exceptions, gnome-themes-standard, NEW
17:49:56 <adamw> seems pretty late...
17:50:03 <kparal> but emoji!
17:50:10 <sgallagh> With no more blockers on the list, it seems unlikely another respin will happen
17:50:34 <Kohane> Freeze Exception for an emoji?
17:50:44 <kparal> 💩
17:50:48 <adamw> lots of emoji, i think.
17:50:49 <kparal> lovely!
17:50:58 <Kohane> I know but...
17:51:06 <sgallagh> Also, -1 for this.
17:51:10 <adamw> someone argued in the update thread that it's a bad idea to use a hard requirement here, which seems a reasonable argument too.
17:51:13 <sgallagh> There's no reason it needs to be on the media
17:51:21 <adamw> sgallagh: workstation live?
17:51:22 <Kohane> kparal: how did you send that? I didn't know IRC has emojis....
17:51:29 <adamw> Kohane: they're unicode.
17:51:32 <Kohane> Ah
17:51:43 <adamw> you can send any unicode character over IRC. well, with a unicode-enabled client and server. which most are these days.
17:51:48 <Kohane> adamw: thanks, I didn't know
17:52:07 <adamw> i think i'm -1 here too, bit late for mucking around with this.
17:52:12 <Kohane> anyway, I'm -1 on this
17:52:15 <frantisekz> -1
17:52:20 <sumantro> -1
17:52:23 <lbrabec> -1
17:52:28 <sgallagh> adamw: Fine, I don't like the idea of making any non-blocker changes at this time (excepting maybe "FEs that would be blockers except it's a non-blocking arch")
17:52:34 <sgallagh> -1
17:52:36 <kparal> 0
17:53:45 <adamw> proposed #agreed 1509996 - RejectedBlocker (Final) - this seems like a relatively minor improvement (that only affects Workstation live sessions really) and we felt it's a bit late to be fiddling with comps and core package requirements for something like this
17:53:48 <pwhalen> -1
17:53:51 <adamw> grr
17:54:01 <adamw> proposed #agreed 1509996 - RejectedFreezeException (Final) - this seems like a relatively minor improvement (that only affects Workstation live sessions really) and we felt it's a bit late to be fiddling with comps and core package requirements for something like this
17:54:13 <sgallagh> ack
17:54:15 <lbrabec> ack
17:54:17 <Kohane> ack
17:54:26 <kparal> ack
17:54:29 <sumantro> ack
17:54:40 <pwhalen> ack
17:55:09 <frantisekz> ack
17:55:26 <adamw> #agreed 1509996 - RejectedFreezeException (Final) - this seems like a relatively minor improvement (that only affects Workstation live sessions really) and we felt it's a bit late to be fiddling with comps and core package requirements for something like this
17:55:35 <adamw> #topic (1507676) AArch64 disk images include '/etc/sysconfig/network-scripts/ifcfg-enp1s0'
17:55:36 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1507676
17:55:36 <adamw> #info Proposed Freeze Exceptions, spin-kickstarts, MODIFIED
17:55:41 <adamw> i almost feel like we already did this one?
17:55:51 <adamw> oh, i did it in-bug.
17:56:58 * adamw really not sure what to say to this one.
17:58:12 <kparal> why is that file problematic?
17:58:57 <Kohane> adamw: give me a moment to read...
17:58:58 <pbrobinson> kparal: it confuses people
18:00:02 <Kohane> This might be a silly question but... what is AArch64 ?
18:00:19 <pwhalen> Kohane, 64bit arm
18:00:20 <pbrobinson> Kohane: ARM 64
18:00:35 <Kohane> Oh, okay, thanks
18:00:37 <adamw> it's a non-blocking arch at present.
18:00:43 <Kohane> Yeah.
18:01:03 <pbrobinson> it's also a completely self contained change
18:01:05 <sgallagh> I guess I'm okay with an FE on the off-chance that another respin happens
18:01:25 <adamw> pbrobinson: i'm still not convinced by that.
18:01:37 <adamw> pbrobinson: don't the same kickstarts get used for 32-bit arm and cloud disk images?
18:01:44 <adamw> some of which are release-blocking.
18:01:48 <pbrobinson> adamw: nope, no they don't
18:01:55 <adamw> hmmmm. /me fry eyes
18:02:39 <pbrobinson> as I explained in the bug report the fedora-disk* are currently just aarch64, the plan is eventually to move ARMv7 to consume them too, and maybe a x86 disk image but it doesn't happen yet
18:03:04 <pbrobinson> you can verify that by checking the includes and the pungi config if you don't believe me
18:03:38 <adamw> okay.
18:03:46 <adamw> then i guess fine
18:03:53 <adamw> seems kinda trivial, but if it doesn't affect any blocking image, me
18:03:54 <adamw> h
18:03:55 <adamw> +1
18:05:49 <adamw> so that's +2
18:05:51 <pwhalen> +1 as well
18:05:51 <adamw> any other votes?
18:06:02 <Kohane> +1
18:06:15 <coremodule> Dang, sorry guys, not trying to spam you with disconnect/reconnect messages...
18:06:28 <frantisekz> +1
18:06:30 <Kohane> coremodule: No worries.
18:06:48 <sumantro> +1
18:07:40 <adamw> proposed #agreed 1507676 - AcceptedFreezeException (Final) - this seems pretty minor, but it can't be fixed with an update and pbrobinson swears on his mother's life that it can't affect any images other than aarch64, so we'll grant an FE
18:07:54 <adamw> pbrobinson: hope your mom has her affairs in order...:P
18:08:27 <Kohane> These meetings are fun...
18:08:31 <pbrobinson> adamw: LOL.... just you leave my mother out of this ;-)
18:08:34 <adamw> okey dokey, moving onto beta
18:08:36 <adamw> er, server beta
18:08:39 <kparal> ack
18:08:47 <pwhalen> heh, ack
18:08:49 <frantisekz> ack
18:08:50 <Kohane> ack
18:08:56 <adamw> #agreed 1507676 - AcceptedFreezeException (Final) - this seems pretty minor, but it can't be fixed with an update and pbrobinson swears on his mother's life that it can't affect any images other than aarch64, so we'll grant an FE
18:09:10 <adamw> #info moving onto Modular Server Beta proposed blockers
18:09:18 <pbrobinson> there's a glitch in the matrix adamw
18:09:26 * adamw hears some sort of faint buzzing
18:09:39 <adamw> a sort of whiny, australian buzzing
18:09:40 <adamw> how strange
18:10:08 <adamw> #chair pwhalen
18:10:08 <zodbot> Current chairs: adamw lbrabec pwhalen sgallagh
18:10:09 * pbrobinson goes to dig out some real blockers for adamw to lose sleep over ;-)
18:10:12 <adamw> #topic (1509659) Modular Server is missing PackageKit
18:10:13 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1509659
18:10:13 <adamw> #info Proposed Blocker, fedora-modular-release, ON_QA
18:10:51 <adamw> +1 blocker (though allegedly fixed already)
18:11:16 <sgallagh> +1 blocker
18:11:18 <pwhalen> +1
18:11:35 <kparal> beta blocker, right?
18:11:43 <adamw> proposed #agreed 1509659 - AcceptedBlocker (Server Beta) - clear violation of "It must be possible to log in to the default Cockpit instance and use it to [...] Enrol the system to a FreeIPA or Active Directory domain"
18:11:44 <adamw> yeah
18:11:45 <sgallagh> kparal: Yes
18:11:46 <adamw> criterion is beta
18:11:50 <sumantro> +1 blocker
18:11:52 <kparal> +1
18:11:59 <Kohane> +1 blocker
18:12:27 <sgallagh> Well, it's not 100% clear. But I won't argue the minutiae
18:12:45 <sgallagh> (technically, cockpit can still enroll if you manually install all the packages ahead of time)
18:12:56 <sgallagh> But realistically it violates the spirit of the criterion
18:13:24 <adamw> oh great, now i'm hearing a faint whiny american buzzing
18:13:25 <adamw> :P
18:13:33 <adamw> ack/nack/patch?
18:13:34 <Kohane> LOL
18:13:40 <sgallagh> ack
18:13:49 <sumantro> ack
18:13:50 <pwhalen> ack
18:13:54 <Kohane> ack
18:13:56 <frantisekz> ack
18:13:56 <adamw> #agreed 1509659 - AcceptedBlocker (Server Beta) - clear violation of "It must be possible to log in to the default Cockpit instance and use it to [...] Enrol the system to a FreeIPA or Active Directory domain"
18:14:05 <adamw> #topic (1509671) Modular Server is missing python-psycopg2
18:14:05 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1509671
18:14:05 <adamw> #info Proposed Blocker, fedora-modular-release, ON_QA
18:15:32 <adamw> +1, obv.
18:15:38 <pwhalen> +1
18:15:41 <sgallagh> +1
18:15:42 <kparal> +1
18:15:51 <Kohane> +1
18:16:12 <Kohane> This is plain and simple to decide. I wish all bugs were like this.
18:16:15 <frantisekz> +1
18:16:36 <sgallagh> adamw: Well, I just don't file any of the complicated ones.
18:16:37 * sgallagh runs
18:16:47 <sumantro> +1
18:16:57 <adamw> hehe
18:17:27 <adamw> proposed #agreed 1509671 - AcceptedBlocker (Server Beta) - clear violation of "The core functional requirements for all Featured Server Roles must be met, without any workarounds being necessary" for the database server role
18:18:06 <sgallagh> ack
18:18:22 <Kohane> ack
18:18:24 <frantisekz> ack
18:18:30 <pwhalen> ack
18:18:39 <kparal> ack
18:18:59 <adamw> #agreed 1509671 - AcceptedBlocker (Server Beta) - clear violation of "The core functional requirements for all Featured Server Roles must be met, without any workarounds being necessary" for the database server role
18:19:21 <adamw> #info moving onto proposed Server Beta freeze exceptions
18:19:22 <adamw> #topic (1508545) AArch64 requires shim-signed-13-0.6
18:19:22 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1508545
18:19:22 <adamw> #info Proposed Freeze Exceptions, fedora-modular-release, ASSIGNED
18:19:38 <sgallagh> This is one of those that would be a blocker if it was x86_64
18:19:43 <adamw> +1, yeah.
18:19:50 <sgallagh> So in my mind, that's pretty much an automatic +1 FE
18:20:34 <frantisekz> +1
18:20:35 <pwhalen> +1. this was a beta FE
18:20:35 <kparal> can a new build be submitted as an update just for one arch?
18:20:48 <Kohane> +1
18:20:52 <kparal> or this will also change shim on blocking arches?
18:21:43 <kparal> just saying we could break it on blocking arches
18:21:47 <sgallagh> kparal: That... is a good question.
18:22:01 <sgallagh> We actually CAN restrict it to just one arch, where we couldn't in traditional Fedora
18:22:09 <sgallagh> Though in this case I'm not sure that's a good idea
18:22:30 <adamw> oh, i was misunderstanding this a bit
18:22:40 <adamw> i thought a newer shim just hadn't been pulled into modular server for some reason
18:22:59 <adamw> sgallagh: it's messy. which version do you put in the SRPMs?
18:23:03 <pwhalen> adamw, right
18:23:08 <sgallagh> .bug 1497854
18:23:08 <zodbot> sgallagh: Bug 1497854 – upgrade F26 - F27 fails on reboot due to missing shim.efi - https://bugzilla.redhat.com/1497854
18:23:14 <sgallagh> This is what it originally fixed.
18:23:15 <pwhalen> this is the same FE we had for classic beta
18:23:42 <sgallagh> adamw: no, you're right. We basically just forgot to pull this in.
18:23:43 <adamw> so we already sent 13-0.7 to stable for vanilla fedora?
18:23:52 <sgallagh> Because when I went through the list of Beta blockers, I didn't check the FEs
18:23:57 <adamw> okay. i'm less worried about this breaking other arches for Server, then.
18:23:58 <sgallagh> yes
18:23:59 <Kohane> I'm lost...
18:24:01 <adamw> since we know it works fine in vanilla.
18:24:04 <adamw> so, still +1.
18:24:09 <pwhalen> right, thanks adamw
18:24:41 <adamw> sgallagh: it'd make sense to pull in 0.7 then, right? not 0.6?
18:24:54 <sgallagh> Yes, I'll do that.
18:24:56 <adamw> if that's what's in stable
18:24:57 <adamw> okay
18:25:39 <adamw> proposed #agreed 1508545 - AcceptedFreezeException (Server Beta) - this prevents installs on an entire (non-blocking) arch from working, so clearly worth an FE. the relevant shim-signed build is in 'regular' 27 Final and stable already, so we know it works
18:25:51 <kparal> ack
18:26:13 <sumantro> ack
18:26:14 <pwhalen> ack
18:26:21 <Kohane> ack
18:27:00 <adamw> #agreed 1508545 - AcceptedFreezeException (Server Beta) - this prevents installs on an entire (non-blocking) arch from working, so clearly worth an FE. the relevant shim-signed build is in 'regular' 27 Final and stable already, so we know it works
18:27:11 <adamw> #topic (1509934) mongodb fails installation because of libboost_chrono
18:27:11 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1509934
18:27:11 <adamw> #info Proposed Freeze Exceptions, fedora-modular-release, NEW
18:28:11 <sgallagh> Not a blocking role, but an important module.
18:28:17 <sgallagh> That just turned up overnight
18:28:31 <kparal> can this be installed from netinst as addon?
18:28:40 <sgallagh> kparal: Not without fixing this.
18:28:54 <sgallagh> The problem is that one of its deps is missing from the module, which means it's missing from the repository.
18:28:55 <kparal> I mean if this module can actually be selected as part of some addon
18:29:14 <kparal> otherwise 0day update is fine, no?
18:29:31 <sgallagh> kparal: "It's complicated" (There's another bug we'll discuss a bit later on that)
18:29:36 <kparal> ok :)
18:30:13 <kparal> I guess +1 then
18:31:24 <Kohane> I'm 0 on this. Sorry.
18:31:35 * adamw gives it the sgallagh +1
18:32:15 <sgallagh> kparal: I just mentioned that other bug in #fedora-qa, if you want to sneak an early peek at it
18:33:17 <sgallagh> +1 from me
18:33:45 <sgallagh> I think the fix is unlikely to destabilize anything else (since it's adding a missing package, not changing an existing one)
18:34:00 <pwhalen> +1
18:34:09 <sumantro> +1
18:34:54 <adamw> proposed #agreed 1509934 - AcceptedFreezeException (Server Beta) - sgallagh says there's a reason this can't be done as a post-release update, and it fixes installation of something that could pretty plausibly be needed in a Server product
18:35:06 <kparal> ack
18:35:22 <kparal> nice justification
18:35:23 <sgallagh> Well, the reason is "it's selectable in Anaconda"
18:35:45 <sgallagh> Which would result in it not working properly after installation, since anaconda can't install its deps
18:36:12 <sgallagh> That said, we might opt to make it unselectable when we fix the other bug that's causing all of comps.xml to show up instead of just the available stuff
18:36:22 <Kohane> sgallagh: That's a good point
18:36:41 <pwhalen> ack
18:36:49 <Kohane> ack
18:36:50 <sgallagh> But I'm fine with this resolution
18:36:50 <sgallagh> ack
18:37:26 <adamw> #agreed 1509934 - AcceptedFreezeException (Server Beta) - sgallagh says there's a reason this can't be done as a post-release update, and it fixes installation of something that could pretty plausibly be needed in a Server product
18:37:26 <adamw> <kparal> ack
18:37:36 <adamw> #topic (1509971) Modular Server is missing sscg
18:37:36 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1509971
18:37:36 <adamw> #info Proposed Freeze Exceptions, fedora-modular-release, NEW
18:38:31 <sgallagh> Mostly this just happened because mod_ssl grew a new dep on sscg after the module was created and no one updated the module. Trivial to fix, useful for the same reason as the previous BZ
18:38:38 <sgallagh> +1 FE from me
18:38:50 <pwhalen> +1
18:38:51 <adamw> sure, fine.
18:39:07 <kparal> +1
18:39:09 <Kohane> +1 FE
18:39:36 <adamw> btw, i have a question here
18:39:46 <sgallagh> Shoot me
18:39:51 <adamw> what does the FE actually achieve here?
18:39:52 <sgallagh> err, shoot
18:40:05 <adamw> i mean, a regular FE means i send a push request that lets a bodhi update go through the freeze
18:40:10 <sgallagh> adamw: We implemented a Freeze procedure this morning.
18:40:14 <adamw> haha. okay.
18:40:24 <sgallagh> We won't let updated module builds through to pungi-fedora without a Blocker or FE at this point.
18:40:29 <adamw> do we have to do anything to actually get the changes 'pushed'? or are you taking care of it?
18:40:37 <sgallagh> I'm gatekeeping it
18:41:04 <sgallagh> I'll only send PRs to pungi-fedora with associated blocker/FE BZs
18:41:15 <adamw> proposed #agreed 1509971 - AcceptedFreezeException (Server Beta) - this again fixes install of useful software on Beta and per sgallagh cannot be fully fixed with a post-release update
18:41:15 <sgallagh> pungi-fedora decides which module builds make it to a compose
18:41:16 <adamw> okay.
18:41:29 <sgallagh> specifically, the variants-modular.xml file
18:42:01 <sgallagh> For the record: https://pagure.io/pungi-fedora/pull-request/437 is how we switched into "Freeze"
18:42:32 <sgallagh> So now it's got exact versions specified instead of "latest on the branch"
18:42:33 <adamw> ack/nack/patch?
18:42:41 <sgallagh> ack
18:42:49 <kparal> ack
18:42:51 <pwhalen> ack
18:43:18 <sumantro> ack
18:43:31 <adamw> #agreed 1509971 - AcceptedFreezeException (Server Beta) - this again fixes install of useful software on Beta and per sgallagh cannot be fully fixed with a post-release update
18:43:42 <adamw> #topic (1509973) fonts/everything profile installation fails because of missing dependencies
18:43:42 <adamw> #link https://bugzilla.redhat.com/show_bug.cgi?id=1509973
18:43:42 <adamw> #info Proposed Freeze Exceptions, fedora-modular-release, NEW
18:44:35 <sgallagh> Short version is that the module was created with an "everything" profile that was self-conflicting.
18:44:40 <sgallagh> So we need to back that out.
18:45:01 <sgallagh> That said; I suppose this doesn't strictly have to be on the media, since I don't think this profile is going to be visible.
18:45:28 <kparal> what is the everything profile?
18:45:28 <kparal> or, what is a profile?
18:45:30 <sgallagh> So I can back this change out if preferred (it was the last to go in before we formally Froze, though I had mostly just expected it to get the nod)
18:45:45 <Kohane> kparal: Thanks, I was thinking the same question.
18:46:05 <sgallagh> OK, so when you enable a module, you specify a "stream", which is basically like a major version. e.g. Node.js 8 vs 6
18:46:16 <sgallagh> After that, you have profiles, which are a set of packages out of that module to install.
18:46:34 <sgallagh> There's a special-case one named "default" which will get picked if you don't specify another one.
18:46:56 <kparal> so something like hard and soft dependencies for rpms
18:46:59 <sgallagh> Profiles are used for modules like Samba so you can pick the server, the client, etc.
18:47:20 <kparal> or perhaps like subpackages
18:47:25 <Kohane> Mmmh....  now I understand a bit better...
18:47:27 <kparal> ok, just wanted to get the idea
18:47:31 <sgallagh> In this particular case, karsten got ambitious and included an "everything" (non-default) profile that installs all available fonts
18:47:43 <sgallagh> However it turns out that for whatever reason, they can't actually ALL be installed properly.
18:48:03 <sgallagh> (I think some of them conflict with each other, but I didn't double-check that this is the case)
18:48:27 <sgallagh> So the fix is to drop the ones that conflict from the "everything" profile; we've already checked that none of those are needed by Anaconda.
18:48:51 <adamw> kparal: i think of it more as like the 'mandatory / default /optional' thing in comps.
18:49:07 <sgallagh> Well, it's closer to metapackages than anything else
18:49:39 <adamw> when you say "So I can back this change out if preferred" what do you mean?
18:49:48 <adamw> what 'change'? back it out how? prefer to what?
18:50:23 <sgallagh> adamw: The fix for this had already made it into the builds before I snapshotted for the Freeze on pungi-fedora
18:50:36 <sgallagh> But I can manually revert it to the version that had this bug if we don't want to give it an FE
18:50:50 <kparal> sgallagh: what is your preferred approach?
18:50:52 <sgallagh> Mostly because karsten starts work a few hours before I do :)
18:50:52 <adamw> oh. no. that seems silly.
18:51:00 <adamw> sgallagh: just tell us what to vote
18:51:11 <pwhalen> :)
18:51:15 <kparal> adamw: +1
18:51:17 <sgallagh> It's minimal risk, so I'm in favor of it
18:51:20 <adamw> alrighty
18:51:20 <adamw> +1
18:51:20 <kparal> +1
18:51:21 <adamw> :P
18:51:24 <Kohane> +1
18:51:25 <pwhalen> +1
18:51:27 <sumantro> +1
18:51:28 <adamw> this is excellent bureaucracy btw
18:51:36 <Kohane> I noticed lol
18:51:39 <sgallagh> The only risk I can see is that it's possible karsten was wrong and one of the anaconda fonts is missing. But I doubt that.
18:51:53 <kparal> sgallagh: next time, you can save the explanation and just say the "magic line"
18:51:54 <adamw> sgallagh invents a freeze process, then proposes all the freeze exceptions and we all vote how he tells us to
18:51:56 <sgallagh> And it's easy to revert if that's what happens.
18:52:02 <sgallagh> hahahaha
18:52:08 <adamw> +10 bureaucracy!
18:52:12 <sgallagh> Well, we *do* need a freeze process.
18:52:26 <sgallagh> So I work with what I have, since no one else implemented a better one
18:52:29 <adamw> sgallagh: how is that a 'risk', if this is only about a non-default profile?
18:52:41 <adamw> i thought the fix here was just to take the affected fonts out of the 'everything' profile
18:52:46 <adamw> is the everything profile used to construct the installer images?
18:53:09 <sgallagh> adamw: Actually, right. This will have no impact on anaconda.
18:53:15 <adamw> okay, then we're good.
18:53:17 <sgallagh> Sorry, confused myself slightly
18:53:35 <adamw> proposed #agreed 1509973 - AcceptedFreezeException (Server Beta) - sgallagh sez so
18:53:37 <adamw> :P
18:53:41 <adamw> okay fine
18:53:52 <sgallagh> I'm working on almost no sleep and no days off for like two and a half weeks here. Cut me some slack, please :-/
18:53:56 <kparal> adamw: patch, grammar
18:54:16 <adamw> proposed #agreed 1509973 - AcceptedFreezeException (Server Beta) - this will ensure an available profile for the fonts module can actually be installed without errors
18:54:32 <sgallagh> ack
18:54:36 <sumantro> ack
18:54:39 <kparal> ack
18:54:41 <Kohane> ack
18:54:45 <adamw> #agreed 1509973 - AcceptedFreezeException (Server Beta) - this will ensure an available profile for the fonts module can actually be installed without errors
18:54:57 <adamw> okay, that looks like everything
18:55:02 <adamw> #topic Open floor
18:55:07 <sgallagh> adamw: One more, possibly
18:55:09 <adamw> any other business related to Final or Server releases?
18:55:17 <sgallagh> We should discuss the one I was just mentioning in #fedora-qa
18:55:23 <adamw> 1510140 ?
18:55:24 <sgallagh> See if it fits somewhere
18:55:33 <sgallagh> yes
18:56:05 <adamw> i can't really find a way to judo it into the criteria. i think probably FE is fine
18:56:11 <adamw> i'd certainly be +1 FE for beta
18:56:29 <sgallagh> As long as it's NOT a Beta Blocker, I can work with that.
18:56:35 <sgallagh> I'm just certain I can't fix that this week
18:56:51 <adamw> yeah, don't think it needs to block beta.
18:57:07 <adamw> if you don't intend to fix it for beta, granting it a beta fe doesn't make much sense...
18:57:20 <adamw> and it presumably doesn't need a final FE yet
18:57:51 <sgallagh> Right, I just wanted to make sure the blocker crew agreed that it's not a Beta blocker
18:58:23 <Kohane> I do agree is not a Beta blocker
18:58:42 <sgallagh> I'll do whatever I can to make sure it doesn't get to the point we need to decide on Final Blocker status :)
18:58:58 <adamw> yeah, i'm okay with it not being a beta blocker.
18:59:10 <adamw> so in that case, sounds like we don't need to topic it formally as there's nothing to vote on.
18:59:52 <adamw> #info for the record, we informally considered #1510140 and agreed it'd definitely be best to fix it before final, but it doesn't obviously hit any criteria and no-one seemed worried that it should block beta
18:59:56 <sgallagh> ack
19:00:19 <sumantro> ack
19:01:04 <pwhalen> ack
19:01:07 <adamw> nothing to ack, there
19:01:09 <adamw> it wasn't proposed
19:01:10 <adamw> :P
19:01:17 <adamw> okay, so anything else, or can i go get some breakfast?
19:01:20 <sgallagh> I was acking your statement of "nothing to vote on"
19:01:23 <adamw> aha
19:01:24 <adamw> heh
19:01:26 <sgallagh> Sorry for confusion
19:01:50 <sgallagh> I think I'm good.
19:02:04 <Kohane> I'm good too
19:02:08 <sgallagh> Anyone who wants to put the Modular Server through its paces, please download and do so :)
19:02:23 * sgallagh is really trying to hit Go for Thursday
19:02:27 <pwhalen> sgallagh, which is the best to look at?
19:02:50 <sgallagh> 2017_11_05.n.3 is the one I Froze
19:02:54 <pwhalen> ok, thanks
19:03:39 <adamw> i'll try to adjust openqa for differences in modular as best as i can
19:03:44 <adamw> i'll need to talk to you about the repo setup
19:04:00 <sgallagh> Sure
19:04:09 <sgallagh> If I'm not around, tdawson also has that information
19:04:58 <adamw> roger
19:05:01 <adamw> thanks for coming, folks
19:05:03 * adamw sets fuse
19:05:32 <Kohane> thanks for chairing adamw
19:05:56 <Kohane> bye everyone, see you later
19:05:56 <kparal> thanks
19:06:05 <sumantro> adamw, thanks for chairing :)
19:08:18 <adamw> #endmeeting