17:00:55 #startmeeting F24-blocker-review 17:00:55 Meeting started Mon Feb 22 17:00:55 2016 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:55 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:55 The meeting name has been set to 'f24-blocker-review' 17:00:55 #meetingname F24-blocker-review 17:00:55 #topic Roll Call 17:00:55 The meeting name has been set to 'f24-blocker-review' 17:01:01 ahoyhoy folks, who's around to review some blockers? 17:01:06 * pschindl is here 17:01:08 * kparal is here 17:01:14 * satellit_e listening 17:01:16 #chair kparal pschindl 17:01:16 Current chairs: adamw kparal pschindl 17:01:18 morning satellit 17:01:25 hi 17:01:48 adamw: btw, happy birthday, old man! 17:02:01 * tflink pops his head up from the spiky rabbit hole 17:02:26 .moar adamw cake 17:02:26 here cake, have some more adamw 17:02:32 umm 17:02:35 damn, I always mess it up 17:02:49 .moar cake adamw 17:02:51 here adamw, have some more cake 17:02:58 * pwhalen is here 17:03:20 * tflink thought of http://tng.trekcore.com/gallery/albums/s7/7x06/phantasms119.jpg 17:03:21 happy birthday adamw! 17:03:24 * linuxmodder stumbles in 17:03:30 my god, this cake is trying to eat me 17:03:44 thanks pwhalen! how did you know?! are you spying on me with lasers? 17:03:57 oh, kparal said it first 17:04:00 same question to him 17:04:09 google knows everything 17:04:39 and it happened to tell me 17:05:20 blame the tentacle 17:05:41 * rdieter waves with popcorn 17:06:17 wow, I sound incredibly random today :-/ 17:06:50 I got it 17:07:31 can't escape that tentacle 17:07:35 who wants to secretarialize? 17:07:41 yeah, just wondering exactly what kind of a day it's going to be with spiky rabbit holes, people cake and tentacles 17:07:41 ahoy rdieter 17:07:49 tflink: just another day at the office! 17:08:05 * adamw is writing an incredibly existential post on the compose process, so everyone look forward to that 17:08:18 adamw: i suspect that the norm in your office may be slightly different than mine 17:08:18 * adamw waves pipe, strokes beard, dons turtleneck 17:08:21 another one? oh no 17:08:25 kparal: gird your loins 17:09:08 pschindl: I'm waiting on you to volunteer, step up! 17:09:57 he's playing a rock today, so I guess I'll secretarialize 17:11:16 kparal: I'll do it. I just missed the call :) 17:11:42 pschindl: no problem, it's yours :) 17:12:42 #info pschindl will secretarialize 17:12:45 #topic Introduction 17:12:46 Why are we here? 17:12:46 #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. 17:12:46 #info We'll be following the process outlined at: 17:12:46 #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 17:12:47 #info The bugs up for review today are available at: 17:12:49 #link http://qa.fedoraproject.org/blockerbugs/current 17:12:51 #info The criteria for release blocking bugs can be found at: 17:12:53 #link https://fedoraproject.org/wiki/Fedora_24_Alpha_Release_Criteria 17:12:55 #link https://fedoraproject.org/wiki/Fedora_24_Beta_Release_Criteria 17:12:59 #link https://fedoraproject.org/wiki/Fedora_24_Final_Release_Criteria 17:13:05 we have: 17:13:31 #info 3 Proposed Blockers (Alpha) 2 Proposed Blockers (Beta) 2 Proposed Blockers (Final) 17:13:52 #info 2 Proposed Freeze Exceptions (Alpha) 17:14:03 #info 1 Accepted Blocker (Alpha) 17:14:15 (I figure there's not much point doing Beta or Final FEs or accepted blockers at present) 17:14:37 everyone ready to dive into Alpha proposed blockers? 17:15:23 * kparal puts on a diving suit 17:15:39 and a snorkel 17:18:18 :) 17:18:19 #topic (1309754) firewall-offline-cmd aborts if a service is already enabled 17:18:19 #link https://bugzilla.redhat.com/show_bug.cgi?id=1309754 17:18:19 #info Proposed Blocker, firewalld, NEW 17:18:37 sorry, someone slap me every time i get distracted writing another paragraph of waffle about what a compose really *is* 17:20:25 this looks fairly straightforward, +1 17:20:28 direct violation of the criteria 17:21:05 +1 17:22:01 well, it should be resolved somehow, I'm just not sure if we can immediately blame firewalld 17:22:17 but +1 doesn't work now, either firewalld or anaconda needs to come up some with fix 17:22:55 yeah, since it's reported by anaconda to firewalld i figure they'll work something out :) 17:23:01 +1 17:23:39 proposed #agreed 1309754 - AcceptedBlocker - clear violation of "...Supported install-time firewall configuration options must work correctly.", with the footnote that kickstart-specified options are covered 17:25:34 ack 17:25:48 ack 17:26:20 ack 17:27:08 #agreed 1309754 - AcceptedBlocker - clear violation of "...Supported install-time firewall configuration options must work correctly.", with the footnote that kickstart-specified options are covered 17:27:14 #topic (1308773) Very dark grey used for text in recent Rawhide, makes terminal almost unreadable 17:27:15 #link https://bugzilla.redhat.com/show_bug.cgi?id=1308773 17:27:15 #info Proposed Blocker, gnome-terminal, NEW 17:27:32 this could've been a bit of a tricky one, but fortunately it seems to be fixed now, so we can probably close it and move on 17:28:29 also, for some reason, the gtk style was changed to white 17:28:39 ? 17:28:41 the borders 17:28:45 window borders 17:28:48 oh, yeah 17:29:09 I noticed it as well, +1 from me 17:29:20 +1 17:29:46 +1 from me 17:29:48 +1 17:30:00 its fixed, id say close it and move on as well 17:30:22 (in and out today -- also suffering from shitty connect today) 17:32:01 OKie dokie 17:32:14 #info this bug is reported fixed by multiple users, so we will close it and move on 17:32:20 #topic (1308771) Current Rawhide Workstation live image does not reach GDM due to mislabelled /run/systemd/inhibit and /run/user/1000 17:32:21 #link https://bugzilla.redhat.com/show_bug.cgi?id=1308771 17:32:21 #info Proposed Blocker, systemd, NEW 17:33:52 clear +1 17:34:33 yeah, this one also seems fairly obvious 17:34:34 agreed, +1 17:34:44 this bug prevents KDE live from working properly either, and by some reports might affect installed systems too 17:36:09 proposed #agreed 1308771 - AcceptedBlocker (Alpha) - clear violation of "All release-blocking images must boot in their supported configurations." 17:36:16 +1 17:36:17 ack 17:36:18 ack 17:36:28 ack 17:36:31 * danofsatx is on irssi now, hopefully it's stable 17:36:33 ack 17:36:36 ack 17:37:02 #agreed 1308771 - AcceptedBlocker (Alpha) - clear violation of "All release-blocking images must boot in their supported configurations." 17:37:07 I like this enthusiasm! 17:37:19 keep it up and I may prevent the carnivorous cake from devouring you all 17:37:47 that's all the Alpha proposed blockers, let's move on to Beta proposed blockers 17:37:55 #topic (1306462) _ped.PartitionException: Unable to satisfy all constraints on the partition. 17:37:55 #link https://bugzilla.redhat.com/show_bug.cgi?id=1306462 17:37:55 #info Proposed Blocker, anaconda, NEW 17:38:00 ...aaand we're back to the cmurf specials 17:38:59 i'll see if i can possibly get anaconda folks to give us an opinion 17:40:34 -1, unless it can be dup'd with something other than chromiumOS 17:40:58 I think it will behave the same for any kind of disk with invalid gpt table 17:41:04 right, this is kind of an interesting one 17:41:06 invalid backup gpt table, to be exact 17:41:21 on the one hand, anaconda probably shouldn't crash; but what can we practically achieve if it doesn't? 17:41:26 and how do you get one of those without deliberatly doing it? 17:41:36 dd to a larger disk 17:41:49 as cmurf says, installing to the unpartitioned space might actually be kind of a bad idea, because we'd render the chromium install inoperable, we can't do dual boot with chromium 17:42:07 so say we fix this, what does it achieve? make it easier to shoot yourself in the foot? 17:42:51 I was trying to convince anaconda folks in the past that it would be nice to at least say "your primary and backup gpt table don't match, please fix it and come back" 17:42:51 yeah, it removes the safety and permanently aims at the bullseye on your big toe. 17:43:04 then people would at least know something and could search for a solution 17:43:23 point 17:43:32 I'm not saying this is exactly the same case as I reported in the past, but seems to be close 17:43:57 yeah, the crash here is ultimately an inconsistent gpt table case, i think, the chromium scenario is kind of a separate thing 17:44:08 can someone give me a valid use case where you would be dd'ing to the hard drive in the first place? 17:44:41 danofsatx: well, that's what you're *supposed* to do with the recovery image, i think? 17:44:46 * danofsatx is of the opinion that it is _very_ ucommon to hit this corner case 17:44:55 yeah, that seems to be the consensus 17:45:05 okay, so we all seem to be fairly solidly -1 17:45:12 right, -1 here too 17:45:27 proposal: let's reject it and we can think about how to explain clearly why at our leisure (or pschindl can) 17:45:46 danofsatx: move OS installation to a bigger disk 17:46:12 the thing i'd be most inclined to +1 is 'anaconda crashes when dealing with broken GPT tables', under the 'reject invalid configurations without crashing' criterion, but even that's arguable and i believe has been considered before (so *that* part of this bug is a dupe) 17:47:15 proposed #agreed 1306462 - RejectedBlocker - this is a somewhat complex case and hard to reduce to a summary, but at a very high level, we don't see any way in which making this bug a blocker would produce a significantly improved outcome. See bug for further details 17:47:34 ack 17:47:36 ack 17:47:46 (if anyone has a better snappy summary, by all means, throw it at me) 17:48:04 ack 17:49:01 nope? okay, then, next cmurf special 17:49:04 #agreed 1306462 - RejectedBlocker - this is a somewhat complex case and hard to reduce to a summary, but at a very high level, we don't see any way in which making this bug a blocker would produce a significantly improved outcome. See bug for further details 17:49:06 heh 17:49:17 #topic (1306808) blivet.errors.FSError: mount failed: 32 17:49:17 #link https://bugzilla.redhat.com/show_bug.cgi?id=1306808 17:49:18 #info Proposed Blocker, anaconda, NEW 17:50:03 imagine what would happen if cmurf started using ext4 17:50:41 all our systems would magically break 17:51:09 we seem to be short on anaconda devs feedback 17:52:16 yeah, but this one *does* seem fairly legit 17:52:26 yes 17:52:42 yeah, from the write up, I'm +1 on this one 17:53:05 +1 for me 17:53:22 +1 17:53:27 +1 17:53:44 +1 as well I guess, anaconda folks will come complaining if it is not as it seems 17:58:14 proposed #agreed 1306808 - AcceptedBlocker (Beta) - this seems to be a clear violation of "When using the custom partitioning flow, the installer must be able to: ... Assign mount points to existing storage volumes" , where an earlier section of the criterion makes clear that btrfs "storage volumes" are in the set expected to be supported 17:58:30 ack 17:58:51 ack 17:59:13 ack 18:00:30 #agreed 1306808 - AcceptedBlocker (Beta) - this seems to be a clear violation of "When using the custom partitioning flow, the installer must be able to: ... Assign mount points to existing storage volumes" , where an earlier section of the criterion makes clear that btrfs "storage volumes" are in the set expected to be supported 18:00:48 OK, that's all the beta blockers! onto final. 18:00:50 #topic (1304681) Journal spam: Gdk-WARNING **: gdk-frame-clock: layout continuously requested, giving up after 4 tries 18:00:50 #link https://bugzilla.redhat.com/show_bug.cgi?id=1304681 18:00:50 #info Proposed Blocker, gnome-shell, NEW 18:01:39 I believe I saw this even in my VM 18:02:02 and I saw how unusable pschindl's laptop is with this 18:02:09 I'm more likely to give it +1 now :) 18:02:23 +1 18:02:33 it seems to be either universal or very common 18:03:20 but it would be nice to have more evidence 18:04:03 +1 18:05:09 if we're +1ing it, on what grounds, exactly? 18:05:12 we need to cite a criterion 18:05:50 * adamw tries to get someone from #fedora-desktop to join 18:06:14 pschindl: you were saying the even basic desktop operations are sometimes hard to perform when it gets real bad, right? 18:07:53 kparal: There were moments when whole desktop freeze for a second, but I'm not 100% sure that it is because of this bug. 18:08:29 Gnome-shell consumed a lot of processor time, but it could be something else. 18:08:56 let's introduce a new "we don't like it" criterion 18:09:57 adamw: here now 18:10:04 hi fmuellner! 18:10:08 kparal: don't we have one of those? 18:10:29 so we're trying to decide if it's appropriate for this bug to block Fedora 24 Final release (and if so, what criterion it breaks) 18:10:53 final release is a way down the road but we try to get on top of blockers early these days 18:11:03 do you have a feeling on whether it's appropriate as a release blocker? 18:12:43 https://bugzilla.redhat.com/show_bug.cgi?id=1304681 in case you didn't get the link 18:13:07 i'm kinda struggling because it "feels" like it's a blocker but it's hard to justify why - everything basically *works* 18:16:12 * adamw wonders if everyone got eaten by the cake 18:16:33 is it too early to invoke the "works well" creterion? 18:17:03 so digging in my logs, that message appears a lot until dec-22 18:17:05 which criterion is that again? 18:17:17 (where "a lot" means several times a second) 18:17:20 fmuellner: i'm still getting it here on today's rawhide, it definitely seems to still be a bug 18:17:41 after that I see it twice (jan-29 and feb-09) 18:17:53 i ran journalctl -f for a few seconds and got http://fpaste.org/327382/45616506/ 18:18:29 fmuellner: I'm getting it everyday too on rawhide. 18:21:59 oh, it could be a wayland issue (embarrassingly I'm still running it nested most of the time so I can make changes to gnome-shell without restarting the session) 18:22:32 * adamw is on X 18:23:49 mmh, now that's odd then 18:23:58 so, bleh. um. 18:24:07 since everyone's kinda +1ish, let me try to come up with an excuse 18:24:36 * kparal watching the pro in action 18:24:47 let's use the good old get-out-clause 18:24:59 3.19.90 should be in rawhide by now, so no big difference to master 18:25:36 fmuellner: I'm running 3.19.90 and still seeing the gdk-frame-clock messages all the time 18:25:55 I'm also running Wayland, FWIW 18:27:49 jeez, i'm coming up empty here 18:28:00 * adamw cannot criteria judo the get-out clause into covering this 18:28:05 okay, let's go with something really half-assed 18:28:18 what do we use at go/no-go for the general "must work correctly" criteria? 18:28:27 there is no such thing. 18:28:41 we can always punt and wait until it's fixed 18:28:44 ^^ 18:28:55 a "it must work correctly criterion" would basically nullify the entire criteria and put us back on subjective decisions for everything 18:29:02 nah, i got something half-assed 18:29:02 hold on 18:29:58 proposed #agreed 1304681 - AcceptedBlocker (Final) - this is considered a violation of "All elements of the default panel (or equivalent) configuration in all release-blocking desktops must function correctly in typical use", in that the Shell is the core of the "default panel" for Workstation and constantly spamming the system logs and consuming significant resources constitutes failing to "function correctly" 18:30:14 FOURTH DAN CRITERIA JUDO right there, folks 18:30:32 go adamw! its your birthday.. (sung like 50 cent) 18:30:33 (actually that's not as bad as I thought it would be) 18:30:44 are you sure that this is better than "we don't like it" criterion? 18:30:55 i actually don't think it's bad... 18:31:17 but it's your birthday, so 'ack' :) 18:31:24 i like it 18:31:25 ack 18:31:29 ack 18:33:28 #agreed 1304681 - AcceptedBlocker (Final) - this is considered a violation of "All elements of the default panel (or equivalent) configuration in all release-blocking desktops must function correctly in typical use", in that the Shell is the core of the "default panel" for Workstation and constantly spamming the system logs and consuming significant resources constitutes failing to "function correctly" 18:33:42 alright, last cmurf special, i think 18:33:43 #topic (1033778) installer considers encrypted Apple Core Storage volumes as resizeable 18:33:43 #link https://bugzilla.redhat.com/show_bug.cgi?id=1033778 18:33:43 #info Proposed Blocker, python-blivet, NEW 18:33:58 and one last cmurf 18:34:16 ah, you already said that, damn 18:34:17 that's an old one, innit? 18:34:29 yeah 18:34:48 my biggest objection to this is cmurf is kinda trying the thing where he uses the blocker process to try and gain leverage in a technical discussion 18:35:04 this seems to be a difficult thing to fix with some back-and-forth between anaconda and libblkid folks about what to do with it 18:35:08 that has a long history 18:36:19 it does bother me a bit that real users are hitting it - https://bugzilla.redhat.com/show_bug.cgi?id=1033778#c25 18:36:30 but seriously, who the hell goes into the installer, sees an "unknown" partition and decides to resize it 18:36:36 how do you think that is going to end well 18:37:02 lol, right.. :) 18:38:23 lol 18:38:23 I might go so far as to get myself kicked out of the fedora project before "fixing" that 18:40:19 so let's see about a rational proposal 18:42:22 hmm, rejecting this is actually quite hard 18:43:13 it really is a pretty explicit criterion violation 18:43:26 i'm gonna propose we punt it and discuss the problem with anaconda in a low-pressure way 18:45:18 proposed #agreed 1033778 - punt (delay decision) - this bug does quite clearly violate the criteria as written, but anaconda devs are strongly of the opinion that it does not qualify as a blocker on technical/architectural grounds. we will discuss this in the coming days and decide its status later 18:46:48 ack 18:48:16 any other acks? 18:48:29 ack 18:50:32 ack 18:50:53 ack 18:51:18 #agreed 1033778 - punt (delay decision) - this bug does quite clearly violate the criteria as written, but anaconda devs are strongly of the opinion that it does not qualify as a blocker on technical/architectural grounds. we will discuss this in the coming days and decide its status later 18:51:19 alrighty! 18:51:35 so, i was thinking Alpha freeze was next week, but i got mixed up, it's only F24 branch point 18:51:42 so we probably don't need to do Alpha FEs yet still 18:52:54 nope 18:54:06 OK, so I guess we can check in on the one accepted Alpha blocker 18:54:08 #topic (1283348) Black screen on KDE live session (with qemu-kvm) 18:54:08 #link https://bugzilla.redhat.com/show_bug.cgi?id=1283348 18:54:08 #info Accepted Blocker, plasma-workspace, NEW 18:54:33 this one hasn't really moved a lot... 18:54:39 i guess we should poke it again 18:55:06 sorry, i've been waiting and hoping for a successful live image to be composed, for testing purposes 18:55:15 (it's been awhile, for various reasons) 18:55:32 current ones should be usable with enforcing=0 i think 18:55:48 though, hum 18:55:52 oh, ok, I'll try to find one 18:56:33 now i look at it 18:56:48 the fact that current KDE lives boot basically fine with enforcing=0 i think means this got fixed? 18:57:58 pardon my ignorance, but where do I find that latest successful image? koji? 18:59:46 rdieter, http://koji.fedoraproject.org/koji/tasks?state=all&view=tree&method=livecd&order=-id 19:00:29 muchas gradias 19:00:33 gracias even 19:01:26 rdieter: 'fedfind images -t live -s kde' 19:01:29 ;) 19:01:47 (which more or less does what that koji search URL does, with a bit of extra magic in there) 19:01:55 excellent tool 19:02:16 so anyhow, i guess we can discuss further offline 19:02:25 #info we suspect this bug has actually been fixed, will discuss it further in-bug 19:02:26 agreed, use it daily 19:03:19 #topic Open floor 19:03:25 OK, anyone have anything else blocker-y to discuss? 19:03:55 just as a heads up, hughsie told me today that anaconda doesn't seem to start on wayland at all 19:04:10 (we have recently fixed the bug that caused Live images to boot on X11 and not on Wayland) 19:04:25 so we might have some nice blocker ahead of us 19:04:41 * kparal haven't tested it yet 19:07:04 funsies 19:07:19 yeah, people have been discussing subsequent bugs in the live session in the systemd/selinux bug 19:07:31 i'm trying to squelch that but we *should* separate them out into their own reports, not ignore them 19:07:44 there appears to be an issue where anaconda doesn't start, indeed, and also an issue where GDM appears (autologin doesn't work)( 19:08:40 the anaconda bug will be related to non-working sudo with graphical apps under wayland 19:10:00 #info there are troubles ahead with Wayland in the Workstation live session, we should identify and file and nominate them 19:12:31 sounds like that's all we got! 19:12:34 thanks for coming along, folks 19:12:45 see you next week, same bat-time, same bat-channel 19:12:46 thanks adamw! 19:13:25 and don't forget, bat-shaped cake for everyone 19:20:42 It seems that adamw fell asleep :) Or went to bake some cake. 19:21:11 pschindl: the cake is a lie, but don't tell anyone 19:21:38 we have chair, right? let's end this 19:21:48 thanks for coming 19:21:52 #endmeeting