17:02:26 <tflink> #startmeeting f17-final-blocker-review-meeting-6 17:02:26 <zodbot> Meeting started Fri May 18 17:02:26 2012 UTC. The chair is tflink. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:02:26 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:02:33 <tflink> #meetingname f17-final-blocker-review-meeting-6 17:02:33 <zodbot> The meeting name has been set to 'f17-final-blocker-review-meeting-6' 17:02:39 <tflink> #topic roll call 17:02:48 * nirik is lurking around 17:02:52 <tflink> who's ready for some blocker review awesomeness? 17:02:56 * l0ft1369 is here 17:03:04 <tflink> l0ft1369: welcome! 17:03:08 <l0ft1369> thank you 17:05:12 <l0ft1369> sooo blockers? 17:05:46 <tflink> waiting for more people 17:05:50 <l0ft1369> oh oh ok 17:05:53 <tflink> hopefully some will show up :) 17:06:00 <tflink> right now, it's just the 2 of us 17:06:07 <l0ft1369> sorry first time... a bit over energetic I think 17:06:14 <tflink> no worries 17:06:22 <l0ft1369> it is? if I dump who it shows like 12? 17:06:48 <tflink> yeah but not everyone in the channel is here for blocker review 17:07:00 <tflink> adamw: you around? 17:07:00 * satellit_Tris55R listening 17:07:23 <l0ft1369> oh oh I get it now 17:07:41 <cpuobsessed> another round? 17:08:10 <cpuobsessed> still learning how to use irssi 17:08:11 * l0ft1369 nods 17:08:31 <tflink> cpuobsessed: yep, should be a quicker one this time 17:08:44 * tflink uses irssi as well 17:09:04 <cpuobsessed> tflink: how do i leave other rooms without disconnecting 17:09:04 <l0ft1369> only 4 blockers right? 17:09:11 * l0ft1369 uses BitchX 17:09:25 <tflink> 4 accepted blockers, 7 proposed NTH, 0 accepted blockers 17:09:44 <tflink> cpuobsessed: /part to leave channels, /window close for PM 17:09:48 <jwb> might want to look at 822981 17:09:55 <jwb> as a proposed blocker 17:10:27 <cpuobsessed> that should clean up some of the noise 17:10:31 <jwb> just got opened about 10min ago 17:12:37 <tflink> jwb: so the effect is that current binutils may result in non-bootable kernels on EFI? 17:12:42 <pjones> right 17:12:54 <pjones> or worse; they may boot and have subtle bugs that don't make any sense. 17:13:34 <tflink> crap, I was hoping for no proposed blockers today 17:13:40 <jwb> there were reports in the upstream binutils bug of syslinux being impacted as well. not sure if that is going to matter for 32-bit live media or not 17:14:00 <cpuobsessed> not just a fedora issue? 17:14:08 <l0ft1369> that's what I was wondering 17:14:39 <tflink> well, we seem to have enough people to get started 17:14:42 <jwb> no but that doesnt matter much 17:14:44 <tflink> on with the boilerplate! 17:15:05 <tflink> #topic Introduction 17:15:19 <tflink> #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. 17:15:43 <tflink> We'll be following the process outlined at: 17:15:44 <tflink> #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 17:15:51 <tflink> The bugs up for review today is available at: 17:15:52 <tflink> #link https://fedoraproject.org/wiki/Current_Release_Blockers 17:16:06 <tflink> The criteria for release blocking bugs can be found at: 17:16:06 <tflink> #link https://fedoraproject.org/wiki/Fedora_17_Final_Release_Criteria 17:16:06 <tflink> #link https://fedoraproject.org/wiki/Fedora_17_Beta_Release_Criteria 17:16:07 <tflink> #link https://fedoraproject.org/wiki/Fedora_17_Alpha_Release_Criteria 17:16:29 <tflink> any objections to starting with our lone proposed blocker? 17:16:48 <l0ft1369> 822981? 17:16:55 <tflink> #info 1 proposed blocker 17:16:55 <tflink> #info 4 accepted blockers 17:16:55 <tflink> #info 7 proposed NTH 17:17:03 <tflink> #topic (822981) binutils 2.22.52.0.1 miscompiles the kernel 17:17:03 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=822981 17:17:03 <tflink> #info Proposed Blocker, NEW 17:17:04 <buggbot> Bug 822981: unspecified, unspecified, ---, nickc, NEW, binutils 2.22.52.0.1 miscompiles the kernel 17:17:25 <tflink> jwb, pjones: how likely is it that this is going to cause problems? 17:17:31 <tflink> I assume that it affects the kernels in stable? 17:18:17 <tflink> is there anything we can do to poke at the issue in order to get a better feel for how much of a problem it is? 17:18:29 <tflink> wait, this isn't going to require a mass rebuild, is it? 17:19:02 <pjones> no, just kernel 17:19:08 <tflink> good 17:19:08 <drago01> syslinux? 17:19:25 <jwb> my understanding of the problem is fairly limited. it causes some relocated kernels on 32-bit to crash because those are post-processed 17:19:50 <pjones> drago01: no; syslinux may be /telling/ the kernel it's relocated, but the bug is fundamentally that the kernel is built wrong 17:20:12 <drago01> pjones: ah ok 17:20:53 * tflink is reading the lkml thread 17:21:08 * l0ft1369 wonders if anyone is running a 32 bit kernel right now? 17:21:22 * tflink does on some machines 17:21:40 <l0ft1369> yes I have a virtual machine I am trying to throw together right now 17:22:32 <cpuobsessed> i'm running 32bit now 17:22:40 <jwb> i run it on a local machine here. the kernel is clearly misbuilt, but it has been booting fine 17:22:41 <cpuobsessed> in vbox 17:22:54 <jwb> i'm guessing this is due to it being BIOS, which pjones already suggested will probably be ok 17:22:55 <pjones> tflink: if userland has an unrelocatable section, that's fine, it just gets mapped at whatever address it needs to be mapped at. 17:23:32 <pjones> tflink: when we relocate the kernel, though, we *have* to relocate all of it, because, you know, virtual memory isn't a thing. 17:24:38 <cpuobsessed> pjones, how would i cause a crash? just random chance? 17:24:51 <tflink> one would think that we'd have seen reports of problems by now since binutils-2.22.52.0.1-10 was submitted almost 2 months ago 17:25:07 <tflink> but like you said, it may be causing problems that just aren't attributed 17:25:10 <pjones> cpuobsessed: well, if we don't relocate that section and we do need to be, we're probably overwriting /something/ else with it 17:25:39 <pjones> cpuobsessed: on machines where relocating is required, we're probably calling into your firmware randomly? 17:26:01 <pjones> that is, instead of calling the functions that should have been relocated and weren't because they're marked A 17:26:42 <pjones> though that does seem weird given which symbols it is; I can't completely explain it. 17:26:43 <cpuobsessed> i see, most likely if your machine has been running for while? 17:27:42 <pjones> Oh, hrm. 17:28:06 <l0ft1369> so is this a userland only thing? 17:28:44 <pjones> wait, lemme read some more code. I need to understand how these symbols are actually used. 17:29:01 <cpuobsessed> i think this is an NMI, has anyone reported this? 17:29:06 <tflink> this sounds like one of those bugs whose impact can only be determined after release :-/ 17:29:08 <cpuobsessed> as far as Fedora goes 17:29:27 <cpuobsessed> tflink: +1 17:29:40 <pjones> tflink: maybe. 17:29:44 <tflink> is there a fix available? 17:29:51 <pjones> the more I read it, the more it seems like it's just not going to free up initcode memory. 17:30:06 <l0ft1369> Arch rolled back whatever patch intorduced it 17:30:15 <l0ft1369> but that's the only "fix" I've read about 17:31:58 <nirik> theres some 'tip' commits that fix it, but also reorg stuff and bring in new features. hurray. 17:32:25 * cpuobsessed is waiting for 3.4 kernel for his CR-48 17:32:56 * nirik suggests we move on a circle back to this when more is known? 17:33:01 <jwb> l0ft1369, upstream binutils dropped the patches that changed the section symantics 17:33:07 <jwb> it's been fixed in upstream binutils 17:33:34 <tflink> yeah, it sounds like more info is needed before making a decision 17:33:34 <l0ft1369> ok 17:33:47 <cpuobsessed> so roll back binutils before building kernels 17:33:54 <l0ft1369> ew 17:33:54 <cpuobsessed> simple 17:33:56 <cpuobsessed> ? 17:34:06 <l0ft1369> but what if you are trying to do it through yum? 17:34:06 <cpuobsessed> tflink: next! 17:34:07 <tflink> #info this bug was just filed. impact and details are not clear yet, will revisit when more info is available 17:34:22 <cpuobsessed> ack 17:34:28 <l0ft1369> yea knew that was coming 17:34:29 <tflink> cpuobsessed: I'd really rather not roll back something like binutils if we don't absolutely have to 17:35:08 <cpuobsessed> tflink: agreed, only the real real power users build their own kernels 17:35:11 <tflink> ok, on to the accepted blockers - this should be quick since we went over them yesterday 17:35:11 <jwb> we could, you know, fix it too 17:35:13 <jwb> but hey 17:35:18 <cpuobsessed> i was more thinking on the fedora hosted side of things 17:35:26 <l0ft1369> ouch 17:35:30 <tflink> jwb: fix it? That sounds like crazy talk :-D 17:35:39 <jwb> anyway, gotta run 17:36:01 <cpuobsessed> fix it, absolutely but how long would that take? 17:36:06 * tflink is kidding, of course 17:36:08 <cpuobsessed> until then 17:36:17 <tflink> anyways ... 17:36:24 <tflink> #topic (794927) PackageKit cannot import GPG keys - Fatal error: GDBus.Error:org.gtk.GDBus.UnmappedGError.Quark._pk_5ftransaction_5ferror.Code14: Invalid input passed to daemon: char '<' in text! 17:36:28 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=794927 17:36:29 <buggbot> Bug 794927: urgent, unspecified, ---, hughsient, MODIFIED, PackageKit cannot import GPG keys - Fatal error: GDBus.Error:org.gtk.GDBus.UnmappedGError.Quark._pk_5ftransaction_5ferror.Code14: Invalid input passed to daemon: char '<' in text! 17:36:31 <tflink> #info Accepted Blocker, MODIFIED 17:36:55 <cpuobsessed> i've run into that one myself 17:36:56 <tflink> not much to say here 17:37:01 <tflink> new builds are in RC2 17:37:05 <tflink> needs testing and verifivcation 17:37:23 <tflink> #info new PK builds are in RC2, this needs testing and verification so we can close it 17:37:27 <tflink> #undo 17:37:27 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x2d0e1b10> 17:37:38 <tflink> #info new PK builds are in RC2, this needs testing and verification so we can determine whether or not it is fixed 17:38:10 * cpuobsessed hasn't had a chance to test Rc1 17:38:35 <tflink> #topic (821077) Firstboot window does not fit the screen after install from F17 TC4 Final KDE Live 17:38:38 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=821077 17:38:39 <buggbot> Bug 821077: unspecified, unspecified, ---, mgracik, ON_QA, Firstboot window does not fit the screen after install from F17 TC4 Final KDE Live 17:38:40 <tflink> #info Accepted Blocker, ON_QA 17:38:43 <tflink> this feels ... pointless 17:38:52 <tflink> all of the accepted blockers are in the same boat 17:39:02 <tflink> #info new PK builds are in RC2, this needs testing and verification so we can determine whether or not it is fixed 17:39:12 <tflink> #topic (822123) Graphical grub is unusable slow 17:39:12 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=822123 17:39:12 <tflink> #info Accepted Blocker, ON_QA 17:39:13 <buggbot> Bug 822123: unspecified, unspecified, ---, pjones, ON_QA, Graphical grub is unusable slow 17:39:14 <adamw> try that again 17:39:17 <adamw> or...don't 17:39:22 <tflink> #undo 17:39:22 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x2d0e1910> 17:39:24 <tflink> #undo 17:39:24 <zodbot> Removing item from minutes: <MeetBot.items.Link object at 0x2d0e1a50> 17:39:26 <tflink> #undo 17:39:26 <zodbot> Removing item from minutes: <MeetBot.items.Topic object at 0x2c620650> 17:39:27 <adamw> the 'new PK builds' do not fix firstboot. =) 17:39:34 <tflink> #undo 17:39:34 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x3035f7d0> 17:39:38 * l0ft1369 is lost 17:39:58 <tflink> #info the new firstboot build is in RC2, this needs testing and verification so we can determine whether or not it is fixed 17:40:10 <cpuobsessed> twiddled the notes 17:40:12 <tflink> details ... 17:40:23 <tflink> #topic (822123) Graphical grub is unusable slow 17:40:23 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=822123 17:40:23 <tflink> #info Accepted Blocker, ON_QA 17:40:24 <buggbot> Bug 822123: unspecified, unspecified, ---, pjones, ON_QA, Graphical grub is unusable slow 17:40:50 <tflink> #info new anaconda build is part of RC2 - this needs more testing and verification to determine whether or not it can be closed 17:41:16 <tflink> #topic (822495) Fedora 17 DVD build does not contain device-mapper-multipath 17:41:19 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=822495 17:41:21 <buggbot> Bug 822495: urgent, urgent, ---, kanarip, MODIFIED, Fedora 17 DVD build does not contain device-mapper-multipath 17:41:22 <tflink> #info Accepted Blocker, MODIFIED 17:41:48 <cpuobsessed> is it there in RC2? 17:41:50 <tflink> #info spin-kickstarts and comps were changed for RC2 - this needs to be tested to determine whether or not it can be fixed 17:42:19 <tflink> #info a new spin-kickstarts build will be needed before release if the changes are accepted 17:43:19 <tflink> ok, I think that's it for the accepted blockers 17:43:54 <cpuobsessed> brb 17:44:24 * tflink is going to skip any of the proposed NTH w/o associated builds 17:44:56 <tflink> #topic (821957) Missing Requires(post) on grep and coreutils 17:44:56 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=821957 17:44:57 <tflink> #info Proposed NTH, MODIFIED 17:44:58 <buggbot> Bug 821957: unspecified, unspecified, ---, tagoh, MODIFIED, Missing Requires(post) on grep and coreutils 17:45:46 <tflink> this doesn't sound very NTH-worthy 17:46:03 <cpuobsessed> what does it mean? 17:46:18 * l0ft1369 is glad someone else asked that 17:46:28 <tflink> extra error messages, I think 17:47:03 <cpuobsessed> if someone submitted this to me, i'd reject it (but then again I can't see the bz) 17:47:23 <adamw> if the errors are harmless then why the hell is it running the commands at all? 17:47:26 <pjones> since you're very likely to already have those installed by the time fontutil gets there, I think this is a "do it zero-day". 17:48:37 <adamw> the scenario cited in the upstream report is a live compose 17:48:46 <adamw> where you can never quite rely on what ordering yum's going to pick 17:48:51 <pjones> adamw: it's building a cache to improve usability 17:48:54 <adamw> i've seen this kind of thing zoom by in my yum compose logs before 17:49:07 <adamw> right. so if you happen to be building a live image and fontconfig gets installed before mkdir, you won't have the cache. 17:49:12 <adamw> not critical, but not 'harmless', i guess. 17:49:24 <adamw> unless it just does it again later. 17:49:55 <pjones> yeah, dunno. 17:50:00 <pjones> I mean, I would hope so, but... 17:50:07 <adamw> i suppose if we're going to take it that far, then what's it going to build a cache _of_ if you haven't got your fonts installed yet? ah, foo. 17:50:37 <adamw> it's late and we don't know for sure that this breaks anything...so -1. 17:51:06 <tflink> yeah, this doesn't seem to be a huge problem right now 17:51:18 <l0ft1369> what could it break? 17:51:38 <adamw> font packages all seem to run fc-cache in their own %post, but i don't see the mkdir anywhere else, immediately. 17:51:51 <adamw> l0ft1369: theoretically, you could not have a fontconfig cache, i guess. which probably makes font rendering slower or something. 17:52:19 <pjones> gotta wonder why the dir isn't owned by a package 17:52:22 <l0ft1369> adamw: so basically at worst people's terminals would be crawling instead of running through scripts? 17:52:24 <tflink> proposed #agreed - 821957 - RejectedNTH - This claims to be harmless error messages and likely would result in issues building a font cache. While unfortunate, does not seem critical and thus rejected as NTH 17:52:33 <adamw> yeah, that seems odd. why mkdir it in post instead of just owning it. oh, who knows. 17:52:39 <adamw> l0ft1369: probably not even that drastic. 17:52:41 <pjones> ... and it is! 17:52:50 <adamw> so...it's entirely redundant. well, that's dumb. 17:52:56 <pjones> eddie:~/devel/fedora/kernel$ rpm -q fontconfig -l | grep var 17:52:56 <pjones> /var/cache/fontconfig 17:52:56 <pjones> /var/cache/fontconfig 17:52:58 <pjones> a couple of times. 17:53:07 <adamw> belt, braces _and_ suspenders. 17:53:18 <adamw> ack 17:53:20 <l0ft1369> well then I think we should advise people to enjoy the slow walk 17:53:39 <tflink> other ack/nak/patch? 17:53:41 <adamw> in fact, i'd make it less strong, tflink 17:53:51 <adamw> 'could possibly result in temporary issues when generating live images' or something 17:54:04 <adamw> since it seems pretty strongly like the cache is going to get built one way or another anyhow 17:54:15 * cpuobsessed has next to no experience building images 17:54:26 <tflink> proposed #agreed - 821957 - RejectedNTH - This claims to be harmless error messages and could possibly result in temporary issues with building a font cache. While unfortunate, does not seem critical and thus rejected as NTH 17:54:26 <l0ft1369> neither do I 17:54:39 <adamw> ack 17:54:47 <tflink> proposed #agreed - 821957 - RejectedNTH - This claims to be harmless error messages and could possibly result in temporary issues with building a font cache and slowing down font rendering. While unfortunate, does not seem critical and thus rejected as NTH 17:55:26 <l0ft1369> yes 17:55:38 <cpuobsessed> ack 17:55:42 <l0ft1369> oh right 17:55:43 <l0ft1369> ack 17:55:47 <tflink> #agreed - 821957 - RejectedNTH - This claims to be harmless error messages and could possibly result in temporary issues with building a font cache and slowing down font rendering. While unfortunate, does not seem critical and thus rejected as NTH 17:55:56 <tflink> #topic (820751) F17 gcc miscompiles lilypond 17:55:57 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=820751 17:55:57 <tflink> #info Proposed NTH, MODIFIED 17:55:58 <buggbot> Bug 820751: unspecified, unspecified, ---, jakub, MODIFIED, F17 gcc miscompiles lilypond 17:56:45 <adamw> dgilmore already pulled this in, so i guess we should +1 it. =)( 17:56:50 <cpuobsessed> is lilypond installed by default? 17:56:54 <adamw> executive decisions ftw! 17:57:02 <tflink> yeah, I was about to check that - i was pretty sure it got pulled in to TC6 17:57:09 <adamw> no, but i think the point is that if there's a problem in gcc, it should be fixed on the images 17:57:24 <cpuobsessed> i see 17:57:38 <cpuobsessed> i could test 17:58:34 <tflink> proposed #agreed - 820751 - AcceptedNTH - This could cause problems with c++ compilation with the gcc as released on DVDs 17:59:08 <adamw> ack 17:59:57 <tflink> any other ack/nak/patch? 18:00:08 <bcl> ack 18:00:22 <tflink> #agreed - 820751 - AcceptedNTH - This could cause problems with c++ compilation with the gcc as released on DVDs 18:01:15 <tflink> #topic (816554) Keyboard layout does not correspond with default layout selected during installation (F17 TC1) 18:01:18 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=816554 18:01:19 <buggbot> Bug 816554: unspecified, unspecified, ---, rstrode, NEW, Keyboard layout does not correspond with default layout selected during installation (F17 TC1) 18:01:20 <tflink> #info Proposed NTH, NEW 18:02:30 <tflink> what is going on with my scripts, isn't this accepted? 18:02:55 <adamw> doesn't have the whiteboard field 18:03:03 <adamw> and i see no note of 'acceptance' in the bug 18:03:04 <adamw> so...no? 18:03:18 <tflink> sorry, looking at the wrong bug 18:03:19 <bcl> blocks: F17-accepted 18:03:19 <l0ft1369> yea I don't see it in the wiki 18:03:24 <l0ft1369> lol 18:04:06 <drago01> oh god not again 18:04:13 * drago01 hates keyboard layout bugs 18:05:41 <drago01> anyway +1 nth (I'd even even +1 it as a blocker given the "can't login" impact) 18:06:05 <tflink> yeah, I was wondering about blocker material 18:07:15 <drago01> yeah thinking about it ... this should be a blocker 18:07:25 <tflink> adamw: do you remember how we handled these in F16? 18:07:37 <drago01> it is not obvious that the layout is US so users just can't login 18:08:28 <adamw> oh, man, i don't remember. i don't know if this has changed. 18:08:38 <drago01> "There may be times where a requirement is unmet only in a particular configuration, such as with some keyboard layouts but not others, or if a particular character is used in a username, password or passphrase. In such cases, the release team should use their judgement and refer to precedent to determine whether or not the issue should be considered to block the release." 18:08:39 <adamw> i thought we'd decided to enforce that your selected keyboard layout is used all the way through. 18:08:41 <tflink> but this is different from the F16 bug 18:08:58 <tflink> this isn't unlocking the screen, this is logging in 18:09:04 * adamw fucking hates keyboard layout bugs. 18:09:38 * l0ft1369 seconds that notion 18:10:08 * tflink is +1 blocker if this affects all non-US keyboards 18:10:45 <adamw> i think we need to talk to halfline to figure out what's going on here, if it's changed, and what the intent is. but provisionally at least +1 nth/ 18:11:13 <kparal> the meeting is still in progress? 18:11:17 <drago01> yes 18:11:17 <tflink> yep 18:11:18 <adamw> why has this never shown up in testing before? 18:11:21 * kparal joins in 18:11:25 <adamw> we have a test that explicitly asks you to check keyboard layout 18:11:26 <tflink> good question 18:11:33 <adamw> and people have been claiming to have run the desktop tests all the way since alpha 18:11:38 <tflink> it was reported 2012-04-26 18:11:47 <tflink> the better question is why wasn't it flagged as NTH or blocker 18:11:48 <adamw> i'm fairly sure i tested it myself once or twice, with UK English 18:11:58 <adamw> at least for Beta 18:12:01 <drago01> the more explicit criterion that I have proposed a while ago got shoot down 18:12:11 * nirik wonders what media/version they were using there. 18:12:30 <tflink> I assume that petr was using RC1 18:12:33 <adamw> both testers are rh qa so they'll be on the test composes. 18:13:28 <nirik> ok 18:14:10 <tflink> proposed #agreed - 816554 - AcceptedNTH - This is very annoying for users with non-us keyboards, at the very least. Accetped as NTH and proposed as a blocker - will re-visit after determining the number of affected users and exactly what is going on 18:14:28 <nirik> sure, ack. 18:14:32 <kparal> ack 18:14:33 <drago01> -1 18:14:39 <tflink> drago01: patch? 18:14:49 <drago01> err 18:14:54 <drago01> didn't read the end 18:15:00 <drago01> so +1 18:15:02 <drago01> sorry 18:15:18 <drago01> but it is more then just "annoying" 18:15:22 <tflink> #agreed - 816554 - AcceptedNTH - This is very annoying for users with non-us keyboards, at the very least. Accepted as NTH and proposed as a blocker - will re-visit after determining the number of affected users and exactly what is going on 18:15:38 <tflink> drago01: like I said, "very annoying at the very least" 18:15:54 <tflink> it just seems odd that we haven't had more reports if its as bad as it sounds 18:16:07 * tflink has a hard time believing that we have only 3 users testing with non-us keyboards 18:16:20 <drago01> seems like most testers either have us keyboards or silently switch the layout after install 18:16:22 <pjones> tflink: testing might be the operative part there 18:16:57 <tflink> pjones: true but one would think that we had at least one more tester in eastern europe or asia 18:17:11 * tflink is thinking non-roman chars on keyboard 18:17:39 <tflink> #topic (822241) filesystem flushing notifications show up when ejecting read-only optical media 18:17:42 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=822241 18:17:44 <buggbot> Bug 822241: medium, unspecified, ---, tbzatek, ON_QA, filesystem flushing notifications show up when ejecting read-only optical media 18:17:45 <tflink> #info Proposed NTH, ON_QA 18:18:05 <tflink> from the bug, we have -3 and +1 18:19:00 <kparal> but we have a week more now. I don't insist on my -1 too much 18:19:11 <adamw> anyone changing votes with the extra week? it is a nice bit of cleanup. 18:19:21 <kparal> but that means we should have RC3 soon 18:19:26 <drago01> it has +3 karma 18:19:41 <tflink> proposed #agreed - 822241 - RejectedNTH - While this looks bad, it is mostly an annoyance and users are unlikely to hit this using livemedia - could easily fix this with an update. 18:19:50 <tflink> before you yell, I wanted to finish typing - can change 18:20:05 * tflink isn't strongly -1 18:20:18 <tflink> would have been nice to get this into RC2 for testing, though 18:20:25 <tflink> if we do take it 18:20:52 * adamw is finding it hard to care 18:20:55 <brunowolff> RC2 has already been annouced. 18:20:58 <adamw> so that should be ack, really. 18:21:03 <adamw> we should default to 'no'. 18:21:35 <tflink> so, ack/nak/patch? 18:21:38 <kparal> ack 18:21:53 <drago01> +1 18:21:59 <drago01> we have a 0 day update anyway 18:22:02 <brunowolff> ack 18:22:14 <tflink> drago01: I assume that's an ack instead of a +1 NTH? 18:22:20 <drago01> tflink: yes 18:22:27 <tflink> #agreed - 822241 - RejectedNTH - While this looks bad, it is mostly an annoyance and users are unlikely to hit this using livemedia - could easily fix this with an update. 18:22:37 <tflink> #topic (820748) system-config-kickstart is broken 18:22:37 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=820748 18:22:37 <tflink> #info Proposed NTH, MODIFIED 18:22:38 <buggbot> Bug 820748: high, unspecified, ---, clumens, MODIFIED, system-config-kickstart is broken 18:23:12 <tflink> -1 NTH as this can be fixed with an update 18:23:30 <drago01> do we ship it on the dvd? 18:24:13 <kparal> drago01: it seems we don't 18:24:43 <kparal> I don't see it there 18:25:00 <adamw> -1 18:25:04 <drago01> then -1 18:25:05 <kparal> therefore there's nothing to discuss 18:25:15 <kparal> -1 18:25:17 <tflink> proposed #agreed - 820748 - RejectedNTH - The affected package is not included in the DVD and thus can be easily fixed by an update 18:25:21 <brunowolff> ack 18:25:22 <kparal> ack 18:25:36 <adamw> ack 18:25:38 <tflink> #agreed - 820748 - RejectedNTH - The affected package is not included on the DVD and thus can be easily fixed by an update 18:26:22 <tflink> #topic (815413) Preupgrade from F16 to F17 Beta does not regenerate pam.d/system-auth*, causing systemd-loginctl not to know about user sessions, breaking PulseAudio 18:26:25 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=815413 18:26:27 <buggbot> Bug 815413: unspecified, unspecified, ---, systemd-maint, MODIFIED, Preupgrade from F16 to F17 Beta does not regenerate pam.d/system-auth*, causing systemd-loginctl not to know about user sessions, breaking PulseAudio 18:26:28 <tflink> #info Proposed NTH, MODIFIED 18:29:21 * tflink isn't sure about this one 18:29:27 <kparal> I don't see much into it. but my opinion is in comment 35 18:30:43 <tflink> if this only affects a few upgrades and could be fixed with an update, I'm slightly -1 NTH 18:30:58 <drago01> what means "a few" 18:31:10 <drago01> i.e do we know which ones are affected and why? 18:31:19 <adamw> really old systems, basically 18:31:21 <tflink> systems upgraded through many releases 18:31:44 <drago01> many = ? 18:31:49 <adamw> i'm not sure we're entirely convinced the fix is really going to do much, as that file is unlikely to be unchanged on any system where it'd help... 18:31:52 <adamw> i think at least back to 13. 18:31:58 <drago01> not everyone reinstalls every second release 18:33:07 <adamw> no, but given the sparsity of people reporting this bug, not a lot of people have dragged their systems along that far. and it's trivial to fix for those who have. 18:33:16 <adamw> and i don't think the update actually is sure to fix them anyhow. 18:33:37 <adamw> it only updates the stock content of that file, i'm pretty sure it's marked noreplace and will be modified on any running system...imbw though. 18:35:42 <tflink> it sounds like we're slightly -1 NTH on this 18:36:42 <kparal> no objections here 18:36:56 <adamw> i'm -1 given the minor nature of the problem, the importance of the package and the question of whether in practice it fixes anything anyhow. 18:37:59 <tflink> proposed #agreed - 815413 - RejectedNTH - The bug in question should affect few users and there are questions about whether or not the proposed fix would actually fix upgrades, anyways. 18:38:46 <kparal> ack 18:38:50 <brunowolff> ack 18:39:29 <adamw> ack 18:39:33 <tflink> #agreed - 815413 - RejectedNTH - The bug in question should affect few users and there are questions about whether or not the proposed fix would actually fix upgrades, anyways. 18:39:39 <tflink> ok, that's all of the proposed NTH 18:40:03 <tflink> we have one more proposed blocker that didn't show up becase I messed up the proposal process 18:40:14 <tflink> #topic (822430) PK is stuck after installing a single update 18:40:14 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=822430 18:40:15 <tflink> #info Proposed Blocker, NEW 18:40:16 <buggbot> Bug 822430: unspecified, unspecified, ---, hughsient, NEW, PK is stuck after installing a single update 18:40:47 <kparal> I have tried to reproduce comment 6 18:40:57 <kparal> but I didn't succeed, PK didn't get stuck 18:41:00 <tflink> I haven't hit this again, so I'm not sure it's worthy of being a blocker but I did see PK freeze on me when all updates were selected 18:41:07 <kparal> OTOH it was not a clean RC1 installation 18:41:45 <kparal> there are definitely some issues there. but they seem erratic and therefore hard to fix in a short timeframe 18:42:25 <tflink> and it also seems as if the worst-case scenario is that a reboot is required 18:42:28 <adamw> has anyone talked to hughsie? 18:42:40 <adamw> i think at the least you might have to try and repro it with pkmon running 18:42:42 <adamw> or repo with pkcon 18:43:02 <adamw> not sure there's much we can with a report of 'it froze once while running it graphically, i have no information' :) 18:43:14 <tflink> very true 18:43:23 <kparal> I don't know where to dig for more information 18:43:28 <kparal> nothing in messages 18:43:54 <kparal> pkmon doesn't output much 18:44:24 <adamw> i think you get more from pkmon if you leave it running the whole time of the operation, but yeah, hughsie might know how to get better info 18:45:26 <kparal> I don't think it currently qualifies as a blocker 18:45:40 <tflink> yeah, not unless more people start hitting it 18:45:52 <tflink> I might have done something wrong 18:45:53 <adamw> ok, -1 18:46:40 <tflink> proposed #agreed - 822430 - RejectedBlocker - This sounds like it could be a problem, but there is only 1 reporter and no debug information. For now, at least, it is not a release blocking issue. 18:46:54 <kparal> ack 18:47:03 <brunowolff> ack 18:47:22 <adamw> ack 18:47:29 <tflink> #agreed - 822430 - RejectedBlocker - This sounds like it could be a problem, but there is only 1 reporter and no debug information. For now, at least, it is not a release blocking issue. 18:47:35 <tflink> ok, I think that's everything 18:47:42 <tflink> so much for a quick meeting :-/ 18:47:49 <tflink> #topic open floor 18:48:05 <tflink> anything I missed or other stuff to discuss? 18:48:25 <kparal> that bug about old kernels... 18:48:42 <kparal> 820351 18:48:49 <adamw> we can go back to the binutils bug, if we didn't make a determination on that 18:48:53 <adamw> seems pretty clearly -1 now 18:50:06 <tflink> #topic (822981) binutils 2.22.52.0.1 miscompiles the kernel 18:50:06 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=822981 18:50:06 <tflink> #info Proposed Blocker, NEW 18:50:07 <buggbot> Bug 822981: unspecified, unspecified, ---, nickc, NEW, binutils 2.22.52.0.1 miscompiles the kernel 18:50:35 <adamw> so with the new info, i'm solid -1 18:50:36 <tflink> it sounds like the impact of this is going to be much less than initially feared 18:50:49 <tflink> NTH? 18:51:09 * tflink is also -1 blocker, slightly -1 NTH unless there is a fix in the next day or so 18:51:12 <adamw> not even worth that 18:51:31 <adamw> so for anyone not following in #fedora-kernel, the new info is we can't come up with any practical case where fedora users will actually hit his 18:51:32 <adamw> this 18:52:06 <adamw> the only scenarios we could come up with are 32-bit EFI (which fedora does not support) and using kdump on a 32-bit kernel (which is somewhat esoteric and fine to fix with an update) 18:52:33 <adamw> for various reasons, 32-bit BIOS installs and all 64-bit installs don't hit this. 18:52:39 <adamw> that's why we haven't had major panic in the streets. 18:52:42 <tflink> and I imagine that almost all users who are doing kdump on kernels will be updating 18:53:26 <pjones> and again - 32-bit bios will if the machine has a very uncommon memory map 18:53:48 <adamw> oh, right. but we're not sure if there's any of those out there any more... 18:54:13 <tflink> proposed #agreed - 822981 - RejectedBlocker - The only way that a user could hit this on a configuration supported by Fedora is using kdump with a 32-bit kernel or using odd hardware. As such, it can be fixed with an update and not a blocker 18:55:10 <pjones> adamw: more likely they're not used as 32-bit often 18:55:48 <adamw> pjones: oh, i see. 18:55:55 <adamw> ack 18:56:33 <tflink> other ack/nak/patch? 18:58:01 <brunowolff> ack 18:58:23 <pjones> I acked in the bug 18:58:35 <tflink> #agreed - 822981 - RejectedBlocker - The only way that a user could hit this on a configuration supported by Fedora is using kdump with a 32-bit kernel or using odd hardware. As such, it can be fixed with an update and not a blocker 18:58:41 <tflink> close enough :) 18:58:53 <tflink> #topic (820351) It's possible not to get a kernel from the new release when doing an upgrade 18:58:56 <tflink> #link https://bugzilla.redhat.com/show_bug.cgi?id=820351 18:58:58 <buggbot> Bug 820351: high, unspecified, ---, anaconda-maint-list, NEW, It's possible not to get a kernel from the new release when doing an upgrade 18:58:59 <tflink> #info Accepted NTH, NEW 18:59:09 <tflink> kparal: was there anything in particular that you wanted to mention? 18:59:10 <kparal> I have just a question about this 18:59:18 <kparal> a lot of people hit this ATM 18:59:29 <kparal> but after we unfreeze stable updates, it should not happen, right? 19:00:00 <tflink> as long as F17 kernels go to stable before the F16 ones do, I imagine so 19:00:28 <adamw> it...could still happen more than it used to 19:00:36 <kparal> currently main repo is frozen, stable updates are empty and that's the reason why people get fc16 kernels 19:00:42 <adamw> but it shouldn't be _as_ much of a problem 19:00:43 <adamw> right. 19:00:47 <tflink> you're guaranteed to hit this with preupgrade, though 19:01:03 <kparal> tflink: even after f17 gold? 19:01:03 <adamw> tflink: why guaranteed? only if you have updates-testing enabled in f16, that i can think of. 19:01:19 <tflink> adamw: preupgrade only grabs from the release repo - not updates or updates-testing 19:01:22 <adamw> i did ask the kernel team if they can try and wiggle things to keep the f17 kernel ahead of f16 where possible and they said they would. 19:01:30 <adamw> tflink: not updates? well that's insanely stupid. are you _sure_? 19:01:32 <kparal> tflink: are you sure about that? 19:01:33 <tflink> unless I'm completely wrong 19:01:39 <tflink> not 100% sure, no 19:01:58 <tflink> I can test to figure it out 19:02:04 <kparal> that's something to check then 19:02:15 <adamw> the changelog sure references 'updates' a lot for something which doesn't use updates. ;) 19:02:18 <kparal> otherwise the problem would affect everyone 19:02:24 <adamw> ChangeLog: Throttle updates in download/depsolve callbacks to avoid excess CPU use 19:02:24 <adamw> ChangeLog: some GUI updates based on mizmo's mockups; add "Show beta" checkbox 19:02:24 <adamw> NEWS: - Try to get new packages for *all* repos, including updates (bug 473966) (Will Woods) 19:02:25 <buggbot> Bug https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=473966 medium, low, ---, skvidal, CLOSED CURRENTRELEASE, RfE: Preupgrade should downloading new distributation with updates 19:03:22 <tflink> that does sounds like I'm wrong 19:03:30 <adamw> though it's pretty old. 19:03:35 <kparal> the reality might be different. let's check 19:03:38 <tflink> which would be a good thing, to be honest :) 19:03:54 <kparal> hmm, how do we check? 19:04:23 <tflink> I have a couple ideas 19:04:29 <adamw> # Attempt to set up any other enabled repo on the system 19:04:30 <adamw> self.setup_new_repos(download_progressbar) 19:04:30 <adamw> # Complete repo setup 19:04:30 <adamw> self.complete_repo_setup() 19:04:33 <tflink> but it might be easiest to check the code first :) 19:04:38 <adamw> why not just test a preupgrade from f15 to f16? 19:04:42 <adamw> that should make it pretty clear. 19:05:03 <tflink> true 19:05:10 <tflink> easy enough, too 19:05:13 <adamw> anyhow 19:05:28 <adamw> the bug kinda sucks, but there's not a whole helluva lot we can do about it. except maybe the anaconda dodge jkl proposed. 19:05:49 <kparal> I just wanted to check. this answers my question 19:05:51 <adamw> or, y'know, put an epoch in the kernel and then clean up all the messy bodies left over after the kernel team commits suicide. 19:06:00 <kparal> :)) 19:06:21 <tflink> adamw: somehow, that sounds less than optimal 19:07:01 <adamw> okay, so, i just installed from 17 RC1 DVD, picked 'french' as the keyboard layout, set username and password as 'azerty' - i.e. hit the qwerty keys on my US keyboard - and logged in just fine. 19:07:42 <kparal> adamw: I did the same, posted that into that bz 19:08:00 <tflink> either way, sounds like we're just about done 19:08:09 <kparal> it clearly depends on language 19:08:15 <kparal> re-testing with czech 19:08:31 <tflink> #action tflink to test F15 to F16 preupgrade to verify that preupgrade uses updates repo 19:08:51 <tflink> #topic open floor 19:08:55 <tflink> anything else? 19:08:56 <adamw> <halfline> well 19:08:56 <adamw> certain layouts come paired with us 19:09:00 <adamw> that explains a lot. 19:09:21 <kparal> there's not reason to pair czech with us 19:09:27 <kparal> we have all the ascii characters 19:10:38 <adamw> cz-us-qwertz is paired 19:10:43 <adamw> cz-lat2 isn't 19:10:48 <kparal> which channel do you discuss this on? 19:10:50 <adamw> if i'm reading /usr/share/systemd/kbd-model-map correctly 19:10:55 <adamw> #fedora-desktop on gimp 19:11:02 <kparal> ugh 19:11:35 <kparal> reproduced with cz-us-qwertz 19:11:46 <adamw> which is default? 19:11:54 <kparal> us in gdm 19:12:06 <adamw> i mean, which is the one a normal czech person would pick from the list 19:12:19 <adamw> lat2 or us-qwertz 19:12:53 <kparal> normal IT guy uses czech qwerty == lat2. layman user uses czech qwertz == cz-us-qwertz 19:13:08 <adamw> i see 'Czech' and 'Czech (qwerty)', I guess the obvious thing is to go 'Czech'...which is cz-us-qwertz? 19:13:12 <kparal> yes 19:13:17 <adamw> erf. 19:13:35 <kparal> you have no idea how we hate those guys who invented this layout in the time of typewriters 19:13:41 <adamw> heh. 19:13:46 <adamw> well, anyway, we kinda have a stalemate 19:13:47 <kparal> but it's the default everywhere now 19:13:56 <adamw> there's a lot of layouts paired with US, so this is gonna affect quite a few people 19:14:02 <adamw> but halfline is dead against changing the code quickly 19:14:05 <adamw> so we're kinda screwed. yay! 19:14:16 <adamw> when you hit this, can you at least change the layout in gdm? there's a switcher? 19:14:48 <tflink> we could always block release for another month 19:15:02 <adamw> excuse me while i join the post-epoch kernel team 19:15:13 <adamw> halfline is sticking to the line that we shipped f16 this way 19:15:25 <kparal> adamw: yes, there is a switcher. you can't switch it in "unlock locked desktop" screen though 19:15:46 <tflink> kparal: wouldn't you figure this out before you got to a locked screen, though? 19:15:56 <adamw> there's not a lot you can _do_ about it. 19:16:19 <adamw> so the impact isn't really 'can't login' but 'can't login until you figure out the problem', which still sucks but sucks less. 19:16:47 <tflink> and there's always the question of how you'd figure it out if you have only 1 machine 19:16:57 <tflink> livecd, I suppose - hoping you had one around 19:17:01 <kparal> http://imgur.com/LZxJ9 19:17:18 <kparal> you have to figure out to switch keyboard, then it's fine. you have to do it for every login 19:17:27 <adamw> well, i mean, just look at the screen. 19:17:33 <adamw> kparal: or ditch the US layout in control center, of course. 19:17:47 <adamw> we can fix it post-release with an update, halfline suggests change the code upstream then put it in updates-testing for a while 19:18:09 <adamw> i think, if we shipped f16 this way, we can probably hold our noses and pass it 19:18:22 <adamw> we need to do a quick f16 install with an affected layout and check 19:19:15 <kparal> I think it's fine, the problem exists since f15 I believe 19:19:22 <kparal> another problem is here: http://imgur.com/AUjzR 19:19:41 <kparal> the layout can be changed, but you have to click on the label, which is totally unintuitive 19:19:50 <kparal> and no matter which layout I choose, I can't unlock it 19:19:54 <kparal> weird 19:20:03 <adamw> that's the unlock screen bug that was already reported? 19:20:19 <kparal> yes, a different bz 19:20:22 <adamw> what keys are different in a czech layout? is 'qwerty' enough? 19:20:36 <kparal> qwerty is "qwertz" under czech layout 19:20:49 <adamw> i'm just trying to repro on f16 19:20:49 <kparal> y and z are swapped 19:20:57 <kparal> otherwise all the letters are same 19:21:13 <adamw> with the cz-us-qwertz layout 19:21:53 <kparal> yes 19:22:08 <adamw> kparal: "It doesn't matter if it shows 'cz' or 'en' the 19:22:08 <adamw> english layout is still active (and password is written with it)" 19:22:20 <adamw> kparal: so it seems like even if you try and use the switcher it's broken, according to pschindl 19:22:23 <adamw> that's from the bug report 19:24:44 <adamw> so if we assume that these problems have been around since f16...i'd say we make the login one not a blocker (which means it won't get fixed), and take the lock screen one as nth or even blocker since it cannot be worked around, and backport the fix 19:25:03 <kparal> adamw: the reason we hit this so infrequently is that we use "czech qwerty" layout most of the time 19:25:18 <adamw> kparal: i see. but there are several other affected languages 19:25:42 <kparal> adamw: about the locker screen - it can be worked around by using Switch User and using gdm to log back in 19:25:44 <adamw> i believe it's everything in /usr/share/systemd/kbd-model-map where the second column is (something),us 19:25:50 <adamw> kparal: oh ew. 19:27:04 <adamw> like, a ton of indian layouts, all the arabic layouts, a few others 19:27:10 <adamw> tamil 19:28:10 * adamw goes to update the bug 19:30:44 <kparal> tflink: end of meeting? 19:31:01 <kparal> we can probably close it 19:31:03 <tflink> I was waiting for you guys 19:31:59 <tflink> #info next blocker review meeting would be 2012-05-25 @ 17:00 UTC (hopefully not needed) 19:32:09 <tflink> Thanks for coming everyone! 19:32:17 * tflink will send out minutes shortly 19:32:17 <adamw> are we making a call on the keyboard bugs? 19:32:19 <adamw> or waiting for more info? 19:32:56 <tflink> either way, it sounds rather blockery to me 19:34:00 <kparal> do you want to see more fun? 19:34:07 <kparal> try to select your keyboard layout here: http://imgur.com/1hkO1 19:34:26 <tflink> can't you cancel, though? 19:34:34 <kparal> sure, you can cancel 19:35:00 <kparal> yeah, that works, true 19:35:02 <tflink> do we have enough info to vote here? 19:35:44 <adamw> we're talking promoting https://bugzilla.redhat.com/show_bug.cgi?id=741446 to blocker at this point, right? 19:35:45 <buggbot> Bug 741446: unspecified, unspecified, ---, rstrode, NEW, Can't log in after Lock screen 19:35:48 <adamw> remember there's two separate bugs. 19:36:12 <tflink> I thought we were talking about 816554 19:36:47 <adamw> well as far as 816554 goes, i have to vote -1 blocker, since we can't plausibly fix it in any reasonable time frame, and you _can_ pick the correct layout at the login screen. 19:37:14 <adamw> but for 741446 i might be +1 blocker, as we can possibly fix it (need mclasen's advice on whether the fix is safe to backport), and the workaround is much less 'correct' (use switch user) 19:37:29 <kparal> I agree, the fix is not reasonable now for 816554 19:38:43 <kparal> 741446 is already nth, and it's nastier than the gdm one 19:39:13 <kparal> let's wait for desktop team opinion 19:39:17 <tflink> then I guess we could leave it for now 19:40:09 <adamw> mm, okay 19:40:20 <adamw> shall we #topic 816554 and vote it down so it's not blocking up the list? 19:40:55 <kparal> we can easily make it nth, but it won't get fixed anyways 19:41:08 <tflink> #topic (816554) Keyboard layout does not correspond with default layout selected during installation (F17 TC1) 19:41:11 <tflink> #link http://bugzilla.redhat.com/show_bug.cgi?id=816554 19:41:12 <buggbot> Bug 816554: unspecified, unspecified, ---, rstrode, NEW, Keyboard layout does not correspond with default layout selected during installation (F17 TC1) 19:41:13 <tflink> #info Proposed NTH, NEW 19:41:15 <tflink> #undo 19:41:15 <zodbot> Removing item from minutes: <MeetBot.items.Info object at 0x315d9810> 19:41:21 <tflink> #info Proposed Blocker, NEW 19:42:03 <adamw> so, -1 on grounds stated above (implausibility of fix, presence of workaround, fact that f16 had this too, fact that only a subset of keyboard layouts are affected) 19:42:03 <tflink> proposed #agreed - 816554 - RejectedBlocker - While nasty, there is no feasable fix within the time left before F17 GA and there is a reasonable workaround (select correct keyboard layout from GDM menu) 19:42:08 <adamw> feasible 19:42:32 <tflink> proposed #agreed - 816554 - RejectedBlocker - While nasty, there is no feasible fix within the time left before F17 GA and there is a reasonable workaround (select correct keyboard layout from GDM menu). F16 was released with this bug and the sky hasn't fallen yet 19:42:45 <adamw> ack 19:42:56 * tflink wonders how this was released w/ F16 and hasn't been reported yet 19:43:17 <kparal> petr and josef said it had been reported several times 19:43:25 <kparal> I don't remember, can't say 19:43:28 <kparal> ack 19:43:43 <tflink> #agreed - 816554 - RejectedBlocker - While nasty, there is no feasible fix within the time left before F17 GA and there is a reasonable workaround (select correct keyboard layout from GDM menu). F16 was released with this bug and the sky hasn't fallen yet 19:44:12 <tflink> #info should be added as a commonbug 19:44:28 <adamw> i think i added that field 19:45:02 <tflink> you did, indeed 19:45:04 <adamw> yeah, i confirm the bug is in f16. just tested. 19:45:17 <adamw> and the fallback gdm in f16 doesn't even have a keyboard layout switcher :/ 19:45:19 <adamw> man, we release some crap. 19:45:25 <adamw> .fire entire QA team 19:45:25 <zodbot> adamw fires entire QA team 19:45:57 <tflink> how many i18n test days have we had with this bug? 19:46:03 <tflink> oh well 19:46:18 <tflink> #topic open floor (take 3) 19:46:27 * tflink assumes that there are no other topics 19:47:50 <tflink> #endmeeting