16:00:55 <adamw> #startmeeting Fedora QA Meeting
16:00:55 <zodbot> Meeting started Mon Jan 28 16:00:55 2019 UTC.
16:00:55 <zodbot> This meeting is logged and archived in a public location.
16:00:55 <zodbot> The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
16:00:55 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
16:00:55 <zodbot> The meeting name has been set to 'fedora_qa_meeting'
16:01:00 <adamw> #meetingname fedora-qa
16:01:00 <zodbot> The meeting name has been set to 'fedora-qa'
16:01:02 <lruzicka> .hello2
16:01:03 <zodbot> lruzicka: lruzicka 'Lukáš Růžička' <lruzicka@redhat.com>
16:01:05 <adamw> #topic Roll call
16:01:12 <adamw> hi from brno, everyone
16:01:16 <tflink> .hello2
16:01:17 <zodbot> tflink: tflink 'Tim Flink' <tflink@redhat.com>
16:01:18 <lruzicka> hello
16:01:25 <lruzicka> .hello2
16:01:26 <zodbot> lruzicka: lruzicka 'Lukáš Růžička' <lruzicka@redhat.com>
16:01:48 <GlenK> .hello2
16:01:49 <zodbot> GlenK: Sorry, but you don't exist
16:01:52 <GlenK> ha
16:02:02 <adamw> it works if your nick is the same as your fas id
16:02:04 <adamw> otherwise you can do
16:02:07 <adamw> .hello adamwill
16:02:08 <zodbot> adamw: adamwill 'Adam Williamson' <awilliam@redhat.com>
16:02:18 <adamw> and it spams your email address on public IRC! yay
16:02:41 <GlenK> .hello gkaukola
16:02:41 <zodbot> GlenK: gkaukola 'Glen Kaukola' <gkaukola@cs.ucr.edu>
16:02:46 <GlenK> k
16:03:29 * kparal is here
16:04:17 <adamw> alrighty
16:04:40 <Southern_Gentlem> .hello jbwillia
16:04:41 <zodbot> Southern_Gentlem: jbwillia 'Ben Williams' <vaioof@gmail.com>
16:05:12 <adamw> hi southern
16:05:18 <adamw> alrighty, let's get rolling :)
16:05:45 <adamw> "lruzicka and kparal to continue trying to get responses to core desktop applications proposal"
16:05:47 <adamw> how's that going?
16:07:04 <kparal> I provided some feedback to lruzicka  and then I don't know :)
16:07:05 <lruzicka> Yeah, so I sent that email out to the KDE and XFCE sigs
16:08:12 <lruzicka> it created some buzz, but that really was not any real outcome, because some people apparently did not understand what was going on, and some suggested we automate.
16:08:23 <adamw> yeah, i see that
16:09:15 <lruzicka> Basically, the question was something like: Is there something important, where we could devote some focus ... perhaps automate as a first step.
16:09:28 <lruzicka> But the replies were not really convincing.
16:09:55 <lruzicka> and they definitely were not what I would call numerous.
16:10:21 <lruzicka> Since then, there has not been any more replies, so I consider this to be a useless shot
16:10:43 <adamw> thanks lruzicka
16:10:50 <adamw> it's never useless, because now we have proof on record that we tried
16:10:58 <adamw> and no-one can yell at us if we just go ahead and come up with our own ideas :)
16:11:06 <adamw> so what would you like to do for next steps?
16:13:36 <lruzicka> Well, we have now those startstop test for complete Gnome "working" ... so this suggests we do the same with KDE and XFCE?
16:13:49 <adamw> that seems like a reasonable first step for sure
16:14:12 <adamw> note that KDE should be rather easier than Xfce, because we have the basic 'boot and log in to desktop' implemented for KDE already, but not for Xfce
16:14:13 <kparal> the question is not whether we want to automate, but whether we want to block on each app in those desktop envs
16:14:28 <adamw> kparal: right, i explained that in a reply to the xfce list just now
16:14:35 <lruzicka> later, we can focus on some important applications from each of the DEs.
16:14:58 * kparal doesn't have any particular plan atm
16:15:34 <adamw> well, personally I'm fine to leave this with lruzicka to move forward if he is happy that he knows what direction he wants to go in
16:15:52 <lruzicka> I can definitely give it a try and start working on it.
16:15:56 <adamw> sounds good
16:16:32 <adamw> #info "lruzicka and kparal to continue trying to get responses to core desktop applications proposal" - lruzicka sent the mail out to xfce and KDE lists and got some feedback but it was not very useful. we discussed among ourselves and agreed on some next steps
16:16:40 <Southern_Gentlem> it not like you cant look at the ks and see what they have for their builds
16:17:03 <adamw> #action lruzicka to look at implementing app start/stop test for KDE and possibly Xfce next, then consider more extensive testing of core apps
16:17:09 <adamw> Southern_Gentlem: it's not a question of knowing what's in the spin
16:17:37 <adamw> Southern_Gentlem: the question we were trying to ask the desktop teams is whether they have any thoughts on how we can narrow down the scope of desktop app testing to some key apps, rather than just 'all the apps'
16:18:22 <lruzicka> adamw, Yeah, I take the action item as is.
16:18:26 <adamw> cool
16:19:33 <adamw> #info "adamw to review core desktop app mail draft for lruzicka and kparal" - I did that before lruzicka sent it out :)
16:19:40 <adamw> any other follow-up items that aren't on the agenda already?
16:20:28 * satellit listening
16:20:33 <adamw> #topic Fedora 30 status and Change review
16:21:11 <adamw> so, F30 status is: most recent successful composes were fine, but composes have been failing for the last few days due to several issues but most lately libreoffice needing a rebuild and failing to build
16:21:34 <adamw> the latest libreoffice attempt is still running, we'll see how it goes
16:22:41 <adamw> there have been poppler and libicu soname bumps recently and a boost soname bump is going on in a side tag, plus we have the mass rebuild coming up soon, so expect a bit of choppy sailing
16:23:01 <adamw> mass rebuild is scheduled to start Wednesday
16:24:24 <adamw> there is also a proposal on devel@ to land gcc 9 in f30, this would be a late change (it's after the system-wide change proposal deadline)
16:24:32 <adamw> anyone have thoughts on that?
16:25:07 <lruzicka> does it mean that the new gcc is going to break everything?
16:26:06 <lruzicka> or will not this gcc used to build the system itself?
16:26:40 <adamw> yes, it means it becomes the new system compiler
16:26:58 <adamw> new gcc doesn't usually break everything, but breaks a fun selection of packages we then get to go and fix up
16:27:14 <lruzicka> I think it is too risky, if it happens late
16:27:29 <adamw> note: it seems a gcc9 build did make it into the most recent rawhide compose, so probably everything built since then was built with gcc9...
16:28:35 <lruzicka> yeah, in my rawhide machine is 9.0.0.3
16:28:56 <lruzicka> in that case, it's going to be branched with 9.0 gcc anyway
16:29:19 <GlenK> the changes they list on the gcc web page don't seem all that major.  but then 9 is still development status too, so whatever that means as far as future changes go.
16:29:49 <adamw> lruzicka: in theory it can be undone (involving an epoch bump, but hey)
16:29:55 <adamw> it sounds like it's quite likely to go ahead, though
16:30:58 <adamw> i am never that happy with gcc team's casual approach to the Change process and dumping things in rawhide, but probably not necessary to push back on this :/ just deal with the breakages in the mass rebuild as usual. does seem like they did a test mass rebuild and sent out mail notifications of failures, at least (i got one of those)
16:31:03 <lruzicka> if it is there now (in Rawhide), then we will have the chance to see bugs from the very first beginning, if they decide to branch it out, this is ok.
16:31:25 <lruzicka> I though that they would sneak it post beta, or something
16:32:40 <adamw> heh
16:32:45 <adamw> no, we wouldn't let them do that :P
16:33:06 <lruzicka> ok, so I wish, too :)
16:33:15 <adamw> #info GCC 9 is in F30 and likely to stay that way, mass rebuild is scheduled to start this week, so expect some fun results from that
16:33:24 <GlenK> I'm curious, has llvm or any alternatives ever been considered?
16:33:42 <adamw> #info three significant soname bumps for Rawhide have recently occurred or are in progress (icu, poppler and boost)
16:34:14 <adamw> GlenK: in practice ~all f/oss is nearly always built with gcc, and switching to anything else would be a gigantic task
16:34:29 <adamw> and there's no particular indication anything else would be significantly better in any way, to justify the work involved
16:34:46 <adamw> there have been discussions though of course :) check devel@ list archives for them
16:37:07 <besser82> Aprospros, soname bump…  Have there been any issues with libcrypt recently?
16:37:11 <adamw> a few other notable changes...https://fedoraproject.org/wiki/Changes/DbusBrokerAsTheDefaultDbusImplementation is implemented now and mostly working OK, https://fedoraproject.org/wiki/Changes/FlickerFreeBoot may be quite significant and would be a great candidate for a test day...
16:37:15 <adamw> besser82: i haven't noted any, at least
16:37:58 <adamw> https://fedoraproject.org/wiki/Changes/RemoveExcessiveLinking is one to be aware of when looking at build failures, though some other distros have been doing it for years so most things have been fixed already
16:39:10 <besser82> ^ anything that links FORTRAN code will be affected by this.
16:39:59 <adamw> https://fedoraproject.org/wiki/Changes/uEFIforARMv7 is obviously significant to ARM users
16:40:10 <adamw> pwhalen: around? do we have testing planned for that?
16:41:14 <pwhalen> adamw, we do, currently a bug with grub2 but changes have landed in anaconda and oz/imagefactory
16:41:14 <adamw> test process says "This process will be further updated and expanded once all the components are in place and the final process is known. " which...yeah, update that already. :P
16:41:28 <adamw> pwhalen: OK, glad to know you're on it
16:41:32 <adamw> i guess another good candidate for a test day
16:41:37 <adamw> sumantro is unfortunately not around
16:42:01 <pwhalen> when its testable, yes we should likely have a test day. I hope to make some progress this week
16:42:28 <adamw> https://fedoraproject.org/wiki/Changes/Make_ambiguous_python_shebangs_error is listed as a self-contained change but *clearly* is not one, i will point this out in the bug
16:42:36 <adamw> pwhalen: of course, after it's testable :)
16:42:37 <adamw> thanks
16:42:53 <adamw> #action adamw to contact sumantro about Changes that are obvious test day candidates
16:45:16 <adamw> there is also a proposal on the list for bash 5.0 as a 'self-contained change' which again it clearly isn't, imho
16:45:20 <adamw> what do others think?
16:45:49 <adamw> fzatloukal already made that point on the mailing list, heh
16:50:05 <adamw> honestly this does make me pretty concerned, to have it show up as a late Change proposal, under the wrong stream
16:50:36 <adamw> i'd propose we at minimum officially state that we're significantly concerned about a major version landing well after the system-wide change deadline
16:50:59 <adamw> upstream announcement with a summary of changes is here: https://lists.gnu.org/archive/html/bug-bash/2019-01/msg00063.html
16:51:50 <adamw> the "known" backwards incompatibilities don't look hugely significant (though i can't say for sure we don't have vital things using namerefs), but i'd be more concerned about...unknown or unintended ones, or flat out bugs introduced by the changes to support new features
16:51:59 <adamw> per devel@ list there was already a bug in globbing discovered in the initial release
16:52:41 <adamw> and stuff like "There are a number of changes to the expansion of $@ and $* in various contexts where word splitting is not performed to conform to a Posix standard interpretation, and additional changes to resolve corner cases for Posix conformance." is at least potentially concerning
16:54:47 <adamw> anyone +1 or -1? :D
16:55:03 <lruzicka> The mail says: There are a few incompatible changes between bash-4.4 and bash-5.0.
16:55:05 * kparal is worried
16:55:39 <lruzicka> which I can read as, watch out, because someone is not going to update their apps to know about those differences and things will start being ugly
16:55:43 <kparal> but I think the timing is still pretty ok
16:56:57 <GlenK> few weeks past the deadline?
16:57:01 <adamw> okay, well, i'll write something vaguer then :)
16:57:41 <adamw> GlenK: deadline was 01-08, so 17 days late
16:58:18 <adamw> well, we're coming up on time, so...for now i can reply to the thread saying we're a bit concerned, and we can continue with this and other topics next week
16:58:26 <adamw> is that OK with everyone? or is there anything urgent?
16:58:56 <kparal> 👍
16:59:03 <lruzicka> yes
16:59:42 <tflink> wfm
17:00:24 <adamw> alrighty then
17:00:35 <adamw> #info we're over time at this point, so we'll continue with all remaining topics next week
17:00:40 * adamw sets fuse
17:00:51 <adamw> thanks for coming everyone
17:02:11 <adamw> #endmeeting