17:00:12 #startmeeting F15 blocker review #3 17:00:13 Meeting started Fri Apr 29 17:00:12 2011 UTC. The chair is jlaska. Information about MeetBot at http://wiki.debian.org/MeetBot. 17:00:13 Useful Commands: #action #agreed #halp #info #idea #link #topic. 17:00:15 #meetingname F15-blocker-review 17:00:15 The meeting name has been set to 'f15-blocker-review' 17:00:18 #topic Roll Call 17:00:26 both adamw and tflink noted they'll be a few minutes late 17:00:31 anyone else here for the blocker review? 17:00:46 iirc, rbergeron may not be joining today as well 17:01:05 hey hey 17:01:13 let's get this party started, clumens is here 17:01:33 dgilmore, jsmith, robatino, anyone else? 17:01:40 * robatino here 17:01:47 sweet, thanks for joining :) 17:01:56 * jlaska waits another minute, and I think we can get moving 17:01:58 I'll keep an eye out for a while, but have to leave in just under an hour. 17:02:15 though i am going to relocate (with laptop) to the car dealer in a few minutes. 17:02:17 brunowolff: oh good thanks. Actually, I like the 'just under an hour' as a goal for meeting length :) 17:02:26 yo 17:02:31 adamw: that was fast 17:02:32 my laptop's playing up... 17:02:49 #topic Why are we here (this meeting, not life)? 17:02:55 #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. 17:03:04 and a few friendly #link reminders ... 17:03:06 #link https://fedoraproject.org/wiki/Current_Release_Blockers 17:03:08 #link https://fedoraproject.org/wiki/Fedora_15_Final_Release_Criteria 17:03:11 #link https://fedoraproject.org/wiki/QA:SOP_Blocker_Bug_Meeting 17:03:21 adamw: playing up ... being a bad thing? 17:03:47 whew, got it. 17:04:14 * jlaska updating blocker wiki ... will get moving in 1 min 17:05:15 talk amongst yourselfs, I'll give you a topic ... 17:05:40 okay, updated ... let's roll 17:05:41 how the canucks are gonna win the stanley cup? 17:05:49 grumble grumble 17:05:52 hola 17:05:53 why grape nuts contain neither grapes nor nuts? 17:05:59 dgilmore: hey there 17:06:08 okay, quick sweep of the approved bugs ... 17:06:26 #topic https://bugzilla.redhat.com/show_bug.cgi?id=695389 17:06:27 Bug 695389: medium, unspecified, ---, akozumpl, ON_QA, Traceback when writing a kickstart for an unused lvm-on-raid (TypeError: 'NoneType' object is not subscriptable) 17:06:30 #info Traceback when writing a kickstart for an unused lvm-on-raid (TypeError: 'NoneType' object is not subscriptable) 17:06:44 ON_QA, I just pinged Clyde for feedback on this bug 17:06:56 the fixed anaconda is already in stable (thanks to helpful proventesters) 17:07:02 yay 17:07:03 I think we can likely close this out next week 17:07:07 outstanding. 17:07:09 hopefully with confirmation from Clyde 17:07:14 thanks clumens for the updated build 17:07:20 * jlaska about to move on 17:07:48 #topic https://bugzilla.redhat.com/show_bug.cgi?id=696320 17:07:49 Bug 696320: unspecified, unspecified, ---, mgracik, NEW, After text-mode iSCSI install and boot, firstboot-text and getty are both running - unable to login on console 17:07:53 #info After text-mode iSCSI install and boot, firstboot-text and getty are both running - unable to login on console 17:08:24 looks like Lennart pointed out a possible problem with this issue 17:08:51 I think we just need to get in touch with mgracik ... if I have time, I was planning to retest with lennart's proposed change 17:09:11 yeah 17:09:16 nothing else needed here ... moving on ... 17:09:17 i remember we fixed this for the console case last release 17:09:18 is it even expeted that firstboot run on serial installs? 17:09:21 and lennart's explanation makes sense 17:09:33 clumens: expected, well not sure ... it always has 17:09:37 but dunno if that's expected 17:09:44 then i guess that's a "yes" 17:09:49 heh 17:09:57 clumens: yes it does run 17:09:58 * adamw expects the unexpected 17:10:11 "A whole new world ..." 17:10:20 anything else on this one 17:10:27 a new fantastic point of view? 17:10:32 * adamw hangs head in shame 17:10:33 :)) 17:10:39 stand tall young man! 17:10:42 adamw: no chocolate for you 17:10:51 the timing of the TC is a little different for the Final 17:11:04 no one to tell us no 17:11:05 it's scheduled for Monday 17:11:06 or where to go 17:11:13 rbergeron++ 17:11:14 jlaska: its due next week right? 17:11:20 * rbergeron rubs her magic lamp 17:11:21 #link http://rbergero.fedorapeople.org/schedules/f-15/f-15-quality-tasks.html 17:11:27 dgilmore: you bet ... early next week (monday) 17:11:40 rbergeron: aww, i'm not the only one 17:11:47 ok 17:11:48 this firstboot change isn't a TC must have ...but obviously would be nice to have as many fixed in the TC as possible 17:12:01 so we want to get things pushed over the weekend 17:12:06 * rbergeron wishes her new hair job had turned out more ariel-flaming-red 17:12:12 pictures please 17:12:14 :) 17:12:26 i'll get you some 17:12:40 mgracik (firstboot) isn't online right now ... I can drop him a message post-meeting 17:12:42 * dgilmore will make sure to do a stable push on sunday 17:12:51 dgilmore: excellent, thank you 17:13:03 #action jlaska - reach out to mgracik for input on 696320 17:13:11 #topic https://bugzilla.redhat.com/show_bug.cgi?id=684846 17:13:12 Bug 684846: unspecified, unspecified, ---, tbzatek, NEW, selinux denial prevents dbus activation of gnome-keyring-daemon 17:13:15 #info selinux denial prevents dbus activation of gnome-keyring-daemon 17:13:23 this is for gnome apps in kde 17:13:25 more or less 17:13:33 so again nothing hideous for tc 17:13:55 looks a bit stale 17:13:58 debugging seems to have kinda stalled 17:14:01 i can ping rdieter 17:14:09 okay, thanks adamw 17:14:27 #action adamw reaching out to rdieter for status of 684846 17:14:32 #topic https://bugzilla.redhat.com/show_bug.cgi?id=697834 17:14:33 Bug 697834: unspecified, unspecified, ---, rstrode, NEW, Other menu appears in default installation 17:14:36 #info Other menu appears in default installation 17:14:50 this one we may want to adjust the criteria and downgrade 17:14:56 desktop team seems to be changing their mind on it 17:15:27 that's certainly a valid outcome 17:15:38 adamw: is this taking place on desktop@ or somewhere else? 17:15:56 "I know that we've put this in the release criteria at some point, but I don't 17:15:59 think fixing it for F15 is feasible at this time; too many individual desktop 17:16:02 files to change." 17:16:05 nm 17:16:11 mostly discussing in the bug - yeah 17:16:27 okay, don't think there is anything we need to resolve in this meeting 17:16:35 the criterion was requested by desktop team in the first place, so if they want to change their minds, i don't mind, does anyone else? it's not one i'm particularly wedded to. 17:16:36 ehh if they want to ignore the criteria they wrote i guess we can 17:16:36 seems like the right people are talking (adamw+mclasen) 17:17:07 dgilmore: yeah, I'd say refine/remove ... I'd definitely want that criteria removed before we demote this one 17:17:22 yeah 17:17:25 adamw: I've got no skin in that game 17:17:38 definately agreed on changing criteria before demoting 17:17:51 #info Discussing adjustments to the stated desktop menu criteria, which may lower the importance of this bug 17:17:59 hey tflink! 17:18:05 moving on ... 17:18:09 as far as fixing it goes, mclasen and I are having a constructive discussion :P 17:18:13 :) 17:18:14 so yeah, i'll look after this one either way 17:18:19 roger 17:18:25 #topic https://bugzilla.redhat.com/show_bug.cgi?id=693809 17:18:26 Bug 693809: unspecified, unspecified, ---, tagoh, ON_QA, Error message about missing input methods should be removed 17:18:29 #info Error message about missing input methods should be removed 17:18:34 is this something that would be a 1 time thing? would we want to look at giving it a pass for just F15? 17:18:35 This is one of caillons babies 17:18:44 but that would be a discussion for not here 17:18:49 tflink: they could ask to have the criteria reinstated for F16 17:18:58 good point 17:19:16 I reached out to caillon for confirmation whether this addresses the reported issue 17:19:29 I'm not 100% clear on how to reproduce this, but I'd love to get extra ideas 17:19:34 s/ideas/feedback/ 17:19:37 isn't this one that only manifests on upgrade? 17:19:45 hmm, you might be right 17:19:57 if helpful, I can reach out on i18n/l10n lists for feedback? 17:20:07 "Ok, I see no imsettings-gnome package installed when upgrading F14 GA to F15 17:20:10 Beta RC2. though it gets pulled in if do yum groupupdate gnome-desktop." 17:20:18 yeah its upgrades only 17:20:36 but IIRC, callion's issue wasn't so much with the packages issue as the error message itself 17:20:40 its pulled in via comps 17:20:43 that's been changed too 17:20:44 so if I test an upgrade from F14->F15 .. . I should now see imsettings-gnome included? 17:21:00 but thats not used on updates 17:21:06 right 17:21:13 you might need some different IME or language stuff installed 17:21:47 #help 693809 - needs test feedback for updated imsettings to land in F15 17:22:03 i think we can just wait for caillon's feedback 17:22:07 tflink: I've queued up an upgrade test now, will see what I uncover there 17:22:23 agreed 17:22:36 adamw: shall we add this to the list of updates to include in the TC1 ticket? 17:22:41 sure 17:23:00 and by 'we' i mean 'you' 17:23:15 heh, I'll grab it ... just coordinating to avoid duplicate posts 17:23:26 dgilmore: are those RC tickets created yet? 17:23:43 #action jlaska - add https://admin.fedoraproject.org/updates/imsettings-1.2.2-1.fc15 to F15-TC1 rel-eng ticket 17:23:58 #topic https://bugzilla.redhat.com/show_bug.cgi?id=683629 17:23:59 Bug 683629: low, unspecified, ---, than, NEW, KWin doesn't start up in firstboot, error message: Configuration file "/root/.kde/share/config/kdedrc" not writable 17:24:11 #info KWin doesn't start up in firstboot, error message: Configuration file "/root/.kde/share/config/kdedrc" not writable 17:24:46 the reporter indicated they didn't see this problem with the F-15-Beta live cd 17:24:46 jlaska: not yet 17:25:14 yeah, I wonder if this has been fixed indirectly 17:25:46 adamw: can you add this to the rdieter list? 17:26:09 i'll ping in the ticket 17:26:13 thanks 17:26:16 #topic https://bugzilla.redhat.com/show_bug.cgi?id=682001 17:26:18 Bug 682001: unspecified, unspecified, ---, nphilipp, ASSIGNED, s-c-services - all read and disabled, needs to cope with systemd 17:26:22 #info s-c-services - all read and disabled, needs to cope with systemd 17:26:51 no updates since last week, other than a status change 17:26:58 I'll update to note that time is running out 17:28:13 there is always a workaround for this: don't ship s-c-s 17:28:19 we could note that too 17:28:23 btw, is anyone secretarying? 17:28:42 adamw: I'll circle back on these bugs we already discussed post meeting 17:28:50 we can divide and conquer for the proposed list coming next 17:28:58 ok 17:29:01 okay ... ready for some proposed blockers? 17:29:07 huh? lemme hear if folks! 17:29:09 well, i'll do them, i don't mind, just didn't want to tstep on any toes 17:29:11 sure! 17:29:54 adamw: okay, you've got secretarying duties ... tell us if you want someone else to grab a bug 17:29:58 #topic https://bugzilla.redhat.com/show_bug.cgi?id=696278 17:29:59 Bug 696278: high, unspecified, ---, dcbw, NEW, The system network services are not compatible with this version. 17:30:04 #info The system network services are not compatible with this version. 17:30:38 let's see, adamw asked for more info on this issue last week 17:31:01 this one is looking like a guy with a seriously messed up system, but i'm not sure why. 17:31:10 note that he didn't have rsyslog service enabled, which just ain't right. 17:31:54 it's a bit of a mess, to be honest. i'm not sure what to do with it. 17:31:59 I'm still reading, but yeah ... ugh 17:32:05 adamw: yeah i think this is a case of issolated messing 17:32:06 is it out of line to ask them to reinstall? 17:32:11 I just don't know where they are starting from? 17:32:16 yeah i was thinking the same 17:32:25 we can ask them to grab beta, or final tc1, and reinstall, and see what happens 17:32:48 heck, even the last successful live nightly? 17:32:51 or the boot.iso I pushed out earlier 17:33:29 I'm sorry, it's too hard to really pinpoint any single failure that could be at fault 17:33:36 * jlaska works up a proposed ... 17:34:22 proposed #agreed 696278 - continue to monitor issue - Unable to pinpoint any specific failures, reporters system seems horribly broken. Requesting reinstall and updated test results. 17:34:28 I think this needs some patching, but ... 17:34:30 ack/nak/patch 17:34:52 If we don't have anything more concrete the day prior to RC1, I think we have to punt on this? 17:35:13 ack 17:35:16 jlaska: agreed 17:35:17 yep 17:35:45 as much as I don't like the solution of 'reinstall' ... ack 17:35:59 yeah, me neither ... it feels lame 17:36:00 tflink: it's more a diagnostic technique in this case 17:36:09 heck, maybe even just booting a live image 17:36:16 they claim to have hit this just by installing, so...ask them to do the same thing and see if it happens again 17:36:23 if it's not reproducible it ain't a bug. or science. :P 17:36:41 I thought it was an upgrade 17:37:23 proposed #agreed 696278 - NEEDINFO? - Unable to pinpoint any specific failures, requesting retest/reinstall from reporter. 17:37:30 one reporter claims to have used preupgrade 17:37:42 while the initial report has "Since installing Fedora 15... 17:37:42 tflink: the initial reporter says after install 17:37:59 i'm not sure the two are even hitting the same thing, whatever it is they're hitting, but it's too difficult to tell. 17:38:07 I *think* the preupgrade issue should be fixed now with that other Beta issue we resolved (can't recall #) 17:38:24 any other votes, ideas? 17:38:27 adamw: yeah, but I'm not sure that means clean install 17:38:36 unless I'm missing a comment 17:39:12 tflink: recommendations? 17:39:29 jlaska: i think we need one of 2 things. more info on how the environment was and what they did. or them to redo it all 17:39:51 and preferablly knowing more about how they actually had things configured 17:39:54 I don't understand how all the stuff works together enough to offer anything new 17:40:25 it doesn't really ... I see it as adjusting our message to better communicate with the reporter 17:40:38 ... to attempt to further isolate the real problem 17:40:44 yeah 17:40:50 jlaska: yup 17:40:52 so howabout this ... 17:40:58 instead of just a blind "please reinstall" ... 17:41:09 something more in line with that dgilmore suggests 17:41:20 okay, will tweak 17:41:21 help us understand how this started 17:42:00 comments/concerns? 17:42:08 *1 17:42:10 +1 17:42:22 +1 17:42:36 #agreed 696278 - NEEDINFO? - Unable to pinpoint any specific failures, follow-up with reporter to further isolate failure 17:42:50 adamw: ping me if you need any preupgrade testing ... I've got a good setup for doing those reasonably quickly 17:43:03 #topic https://bugzilla.redhat.com/show_bug.cgi?id=698184 17:43:05 Bug 698184: high, unspecified, ---, rstrode, NEW, Enabling session saving with Gnome shell makes GUI login unusable 17:43:07 #info Enabling session saving with Gnome shell makes GUI login unusable 17:43:23 we voted on this midweek, right? i was +1 17:43:29 desktop team is working on fixing it 17:43:41 yeah, I've got 2 votes on this, well 3 if you count me 17:43:51 * jlaska works up proposed line ... 17:44:00 oh hey, good timing mclasen 17:44:10 run away! 17:44:11 :0 17:44:12 we're on the session saving bug 17:44:20 you have a proposed patch waiting on review from vuntz, right? 17:45:07 I think the gnome-session patch is not really necessary to fix the problem in the bug 17:45:28 proposed #agreed 698184 - AcceptedBlocker - impacted by the criteria that upgraded system must meet all criteria, and gnome-session saving was considered common enough to block 17:45:36 ack 17:45:41 ack 17:45:43 ack 17:45:58 adamw: I just need to actually fix the clone 17:45:58 mclasen: as far as the criteria are concerned i think we'd be happy if it didn't hang shell 17:46:01 +1 17:46:08 adamw: it doesn't hang the shell, actually 17:46:10 mclasen: okay, as long as you know what needs doing i think we're fine :) 17:46:14 it just runs mutter without any plugin 17:46:18 mclasen: yeah, it doesn't even need to honor the saved session :) 17:46:20 mclasen: well, that's the apparent effect 17:46:33 #agreed 698184 - AcceptedBlocker - impacted by the criteria that upgraded system must meet all criteria, and gnome-session saving was considered common enough to block 17:46:53 mclasen: fwiw tc compose is due monday so if we want the fix in that we need it over the weekend, but we don't _really_ need it to be fixed in tc, so don't worry too much about it 17:46:56 for rc i think you have a week or so 17:47:17 * jlaska moving on ... 17:47:32 #topic https://bugzilla.redhat.com/show_bug.cgi?id=696864 17:47:33 Bug 696864: unspecified, unspecified, ---, dchen, ON_QA, [abrt] ibus-chewing-1.3.9.2-2.fc15: Process /usr/libexec/ibus-engine-chewing was killed by signal 6 (SIGABRT) 17:47:37 #info [abrt] ibus-chewing-1.3.9.2-2.fc15: Process /usr/libexec/ibus-engine-chewing was killed by signal 6 (SIGABRT) 17:47:57 "Marking this as F15Blocker, since this bug is related to the unmet Final 17:48:00 Release Requirement: All elements of the default panel (or equivalent) 17:48:03 configuration must function correctly in common use." 17:48:06 adamw: I'm building it right now 17:48:11 mclasen: awesome, thanks 17:48:19 this is a language-dependent one so hits our wiggle clause 17:48:33 it sounds like the input method used by most mainland chinese is broken, which is indeed bad. 17:49:00 good news seems to be that this is already in ON_QA 17:49:00 i'm definitely +1 nth, probably +1 blocker. 17:49:04 right 17:49:18 already confirmed on the fix too, nice 17:49:40 so ... what about ibus-chewing is in the default panel (or overview) 17:49:54 it's an IME, right? 17:50:06 input method for traditional chinese? 17:50:10 tflink: yes. 17:50:26 jlaska: if you install with a language that needs an ime, then indeed it is part of the default panel 17:50:34 so it shows up on the panel for a default install if you have traditional chinese selected? 17:50:39 yeah. 17:51:03 okay, then I'd vote blocker for this 17:51:11 +1 blocker 17:51:16 for a while we all were seeing it, btw - it was that little 'keyboard' icon that showed up next to the accessibility one for a while 17:51:23 +1 blocker 17:51:25 until it was fixed not to show up unless you were using a language that needed it 17:52:06 the functional impact is really the worst thing, though - for chinese you _need_ an input method to type anything. if the input method is broken the system's almost unusable. so, yeah. 17:52:07 I need to leave now, will look at log later. 17:52:07 +1 as a blocker 17:52:16 just to be clear: all of this is placeholder for proper ibus/xkb integration that didn't land for 3.0 17:52:23 #agreed 696864 - AcceptedBlocker - Affects desktop default panel (or equivalent) criteria for all mainland chinese installs 17:52:36 brunowolff: as always, thank you for joining and lening your input 17:52:42 s/lening/lending/ too! 17:52:52 mclasen: ah, so there'll be proper shell-side integration with ibus in 3.2 / f16? 17:53:18 depends on finding some domain expert to work on this; it is complicated stuff... 17:53:23 I bet 17:53:44 Anything else on this one, looks like we're all set 17:53:57 nope, move on 17:54:01 #topic https://bugzilla.redhat.com/show_bug.cgi?id=692230 17:54:02 Bug 692230: unspecified, unspecified, ---, mchristi, ASSIGNED, /etc/init.d/iscsi check for network presence needs to be systemd aware 17:54:06 #info /etc/init.d/iscsi check for network presence needs to be systemd aware 17:54:17 I think we have the missing information we needed on this one 17:54:37 im +1 to this being a blocker 17:54:43 the problem was the /etc/init.d/iscsi script was checking to see if NM was running by using /var/log/subsys/NetworkManager 17:54:48 which is no longer used in the systemd world 17:54:59 there was an initial fix, and an improved one expected shortly 17:55:11 yup, +1 too 17:55:22 seems to hit the iscsi criterion pretty clearly. 17:55:26 impact ... any installs with a non-root network volume (iSCSI) don't mount the volume on boot 17:55:32 yeah 17:55:34 +1 blocker from me 17:55:39 +1 17:55:39 +1 from me 17:56:22 #agreed 692230 - AcceptedBlocker - Impacts non-root network (_netdev) partitions failing to mount on boot (e.g. iSCSI) 17:56:37 hey jsmith :) 17:56:39 #topic https://bugzilla.redhat.com/show_bug.cgi?id=699896 17:56:40 Bug 699896: unspecified, unspecified, ---, michel+fdr, ON_QA, AVX code generation is broken in LLVM 2.8 17:56:51 jlaska: I'm sneaky like that :-) 17:56:57 unless I'm missing something, -1 blocker, -1 nth 17:57:14 and I am missing something 17:57:14 Not sure which criteria this is supposed to match, so -1 blocker 17:57:16 * jlaska reads 17:57:25 ajax around? 17:57:28 this isn't the compiler, part of LLVM is used in mesa? 17:57:52 Ah, mesa... 17:57:55 "it makes the Mesa software renderer explode when used on 17:57:55 Sandybridge CPUs. 17:57:55 " 17:57:59 * jsmith always forgets that it uses LLVM 17:58:11 yeah, I read this as a "broken compiler" at first 17:58:17 i think what this means is 'no graphics for you'. 17:58:30 if you're using sandybridge, anyways 17:58:35 we have any sense for how many folks this would impact 17:58:43 More and more, over time 17:58:44 im +1 blocker 17:58:46 I wonder if this is related to some of the sandybridge intel graphics driver issues 17:58:47 or how common sandybridge is? 17:58:53 * jsmith withdraws his -1 17:59:02 i'd like to check with ajax though 17:59:04 jlaska: its intels latest and greatest 17:59:07 jlaska: sandy bridge is intel's current gen 17:59:07 * tflink withdraws -1 too 17:59:10 jlaska: Most of the new Intel machines have a Sandy Bridge proc 17:59:12 so it will be more and more common 17:59:18 do we have criteria that mesa must work? 17:59:27 well i'd like to check with ajax on the exact impact 17:59:28 there have already been issues around sandybridge in #fedora-qa 17:59:30 jlaska: gnome shell needs it 17:59:31 adamw: okay 17:59:35 not sure if they were mesa, though 18:00:05 adamw: did you want to invite him here, or come back to this later/post-meeting? 18:00:06 i just pinged ajax 18:00:10 -ENEEDMOREINFO 18:00:10 k 18:00:12 give him a minute or two 18:00:13 adamw: As did I :-) 18:00:32 i _suspect_ the impact of this would be 'you boot the system, Shell tries to start up, explodes' but we should make sure... 18:00:32 i guess we dont have enough to know if users will just go to fall bacr they will just up desktops 18:00:37 gahh 18:00:47 adamw: right 18:00:56 and sewed 18:01:00 but if this is mesa, you can't use shell with mesa, right? 18:01:06 or is it going to go to fallbak mode 18:01:13 greetings ajax! 18:01:14 tflink: mesa isn't just software rendering... 18:01:24 ajax: heya - we just wanted to know the exact impact of https://bugzilla.redhat.com/show_bug.cgi?id=699896 18:01:24 adamw: OK, didn't know that 18:01:25 Bug 699896: unspecified, unspecified, ---, michel+fdr, ON_QA, AVX code generation is broken in LLVM 2.8 18:01:33 ajax: how screwed will users be 18:01:53 will it go to fall back mode? or just give a messed up screen? 18:02:00 what? 18:02:17 oh. i see what you're asking. 18:02:18 well we're not clearly on exactly what's affected and what the result of the bug is 18:02:35 avx is an instruction set extension in sandybridge cpus 18:02:58 are you going to hit this with any sandy bridge system, even using the intel driver with hardware acceleration? or is it only if you're on software rendering for some reason? or what? 18:03:10 we use llvm for code generation in several paths 18:03:23 one is the whole of the software 3d renderer 18:03:37 one is vertex shaders on hardware that doesn't have it 18:03:57 if you hit any of these paths, whatever has the dri driver loaded will exit() 18:04:07 (on an avx chip) 18:04:11 ok 18:04:19 that could be gnome-shell, or it could be your X server 18:04:20 so...this is going to hit systems with sandy bridge CPUs and unsupported graphics cards, or graphics hardware that's missing shaders? 18:04:26 i think i stick with my +1 to it being a blocker 18:04:46 Sounds like a lot of potential to cause disruption, then? 18:04:51 given the impact i may be only +1 nth on this, but it's a bit academic if there's a fix in already, so let's not waste too much time 18:04:54 * jlaska wants to see the criteria impacted before voting 18:05:02 it's pretty nasty, yeah. 18:05:19 jlaska: that the desktop starts and runs 18:05:26 the place i found it was on a nouveau nvc0 system (which is kinda supported, but we can't ship the firmware yet, so you get software 3d) 18:05:27 ajax: would shell be running on affected systems? i mean, it doesn't work with software rendering, does it work on any hardware that's going to be using software fallback for shaders? 18:05:54 so this isn't just causing fallback mode, this is shell *and* fallback failing entirely 18:05:59 i'm just not sure there's a path here which is actually going to cause shell to crash on anything but a really weird setup, i mean, afaics, anything that could get hit by this bug is probably going to be running fallback mode 18:06:16 yeah, that's where I'm going too 18:06:34 this is only going to crash 3D-using stuff, right? so if you were in fallback mode the desktop wouldn't fall over? 18:06:43 at the least i say +1 as nth 18:06:48 not until you ran a GL app anyway 18:06:59 cheese counts as a gl app, now. 18:07:03 hum 18:07:05 fun 18:07:08 curses! 18:07:16 let's just pick one or the other, approve it, and move on 18:07:17 anyway, it's built and i've tested it. 18:07:19 the fix is going to get in either way 18:07:21 Definitely a NTH, and I'm leaning towards blocker as well 18:07:26 so...i will +1 whatever someone wants to propose =) 18:07:32 I'm going with NTH for the notes 18:07:34 Proposal: Blocker 18:07:37 +1! 18:07:43 ajax: thanks for the info 18:07:46 np 18:08:06 +1 18:08:30 #agreed 699896 - AcceptedBlocker - tested fix available in bodhi 18:08:32 I'm definately +1 nth on this one and leaning towards +1 blocker since I remember seeing several people having issues with sandybridge and the intel drivers 18:08:40 and late :) 18:08:49 tflink: nah, I'm just impatient! :D 18:08:58 #topic https://bugzilla.redhat.com/show_bug.cgi?id=699905 18:09:00 Bug 699905: unspecified, unspecified, ---, ajax, NEW, Mesa rebuild for fixed LLVM 2.8 18:09:03 #info Mesa rebuild for fixed LLVM 2.8 18:09:07 So is this just more of the same? 18:09:15 if the last one was a blocker, would this one be, too? 18:09:24 added as blocker to make sure that it made it past freeze? 18:09:37 +1 as its needed to go with above 18:09:42 yeah, it was cloned from the previous issue 18:09:49 yeah. mesa has a funny build system that has to construct a private dso copy of llvm at build time from the static libs 18:10:08 ok 18:10:09 so +1 18:10:17 +1 18:10:18 so if i fix a bug in llvm, i have to rebuild mesa too 18:10:28 #agreed 699896 - AcceptedBlocker - mesa rebuild required to address previously accepted blocker 699896 18:10:30 (this'll be fixed soon, but it's how it is right now) 18:10:34 ajax: lucky you 18:10:43 #topic https://bugzilla.redhat.com/show_bug.cgi?id=700276 18:10:44 Bug 700276: high, unspecified, ---, pbrobinson, MODIFIED, Enabling session saving with Gnome shell makes GUI login unusable 18:10:53 #info Enabling session saving with Gnome shell makes GUI login unusable 18:10:57 didnt we do this already 18:11:00 This is the issue robatino discovered 18:11:03 different bug 18:11:10 it was cloned 18:11:10 bug number, anyways 18:11:11 i dunno why 18:11:16 yeah, I'm not clear on the clone 18:11:19 different component? 18:11:22 but mclasen is here if we need guidance 18:11:41 ah, this one has the later discussion, heh 18:11:41 robatino: ooh, good news with the latest comment ... new update coming 18:11:45 the gnome-session bug should probably just be closed 18:11:47 i did ask why it was cloned, but no-one answer 18:11:48 ed 18:11:53 if the previous bug was a blocker then this should be also 18:11:55 I really should have moved it instead of cloning 18:12:04 okay, i'll close that as a dupe of this 18:12:08 and we already +1ed it, so move on? 18:12:09 because it is a mutter issue, not a gnome-session issue 18:12:12 So close 698184 as a DUP of 700276 ? 18:12:21 I just wasn't sure initially, and there was another ptach for gnome-session... 18:12:35 lets wait until 700276 is confirmed fixed, maybe 18:12:37 but yeah 18:13:02 well we're not closing 700276 18:13:09 proposed #agreed 700276 - AcceptedBlocker - if updated package resolves issue, DUP 698184 against this bug 18:13:14 just marking 698194 as a dupe of it so we don't have two bugs for the same thing... 18:13:41 we don't really need one bug per component, we need one bug per issue 18:13:52 yeah, can we just DUP it now? 18:13:55 yup 18:14:02 sure 18:14:14 proposed #agreed 700276 - AcceptedBlocker - addresses previously accepted gnome-session saving issue 18:14:22 ack 18:14:25 how many more bugs do we have? 18:14:25 ack 18:14:34 proposed #agreed 698184 - CLOSED DUPLICATE of 700276 18:14:42 1 more blocker, 5 proposed NTH 18:14:45 i have to leave for physio in 45 mins 18:14:51 dgilmore: 1 more on the proposed blockers, and ... waht tflink said 18:14:57 :) 18:15:06 #agreed 700276 - AcceptedBlocker - addresses previously accepted gnome-session saving issue 18:15:09 #agreed 698184 - CLOSED DUPLICATE of 700276 18:15:18 #topic https://bugzilla.redhat.com/show_bug.cgi?id=679579 18:15:19 Bug 679579: medium, unspecified, ---, jglisse, ASSIGNED, [R2XX] GNOME Shell graphics corruption 18:15:23 and 700276 is back in MODIFIED, fwiw 18:15:29 comment #39 neatly captures this 18:15:40 mclasen: sweet, thank you ... robatino and I will test that shortly 18:16:04 we've been told r100 and r200 just aren't going to work for shell, but they don't fail the test for shell support, so at present they try to run shell and you get a bad result 18:16:21 what we need here is for the shell support test to have a blacklist so it can go to fallback mode for r100/r200 18:16:22 adamw: owen has a blacklist patch for gnome-session-is-accelerated 18:16:33 I'm trying to track down if that already landed in f15 or not 18:16:35 mclasen: cool, can you ask him to note it in this bug? 18:16:58 i'm +1 blocker on this bug as long as that's understood to mean 'r100 and r200 should fall back', not 'r100 and r200 should work with shell' 18:17:15 well, more properly, either of the two is an acceptable fix, but the first is what we'll get ;) 18:17:28 adamw: and this basically impacts any desktop criteria for folks with those adapters? 18:17:28 +1 on what adamw said 18:17:34 r100/r200 are the first two generations of Radeon chips, btw - anything up to the Radeon 9500, it's quite a bit of hardware 18:17:37 sine it doesn't properly do accel, or use fallback as intended? 18:17:41 s/sine/since/ 18:17:47 * jlaska needs new fingers or a new keyboard 18:17:48 jlaska: yes, because they'll wind up with a pretty useless shell rather than fallback mode 18:18:06 r100 and r200 is fairly common? 18:18:17 jlaska: it's old, but common at the time 18:18:25 this is ATI hardware of 2000-2002 vintage roughly 18:18:33 i still have boxes with radeon 18:18:33 quite a lot of them are still around having been transferred into newer systems 18:18:40 7000 cards 18:18:51 radeon 8500 and i think 9200 would be common examples affected by this 18:18:54 #link http://www.smolts.org/reports/view_devices?device=r200&search=Submit+Query 18:19:00 I still have one, but I gave up using it a long time ago due to problems with linux 18:19:09 it'd be quite hard to construct a good smolts query as the adapters could identify as a lot of things 18:19:10 at least 1 18:19:23 adamw: oh right, yes 18:19:36 r100 and r200 are the general names for these generations of chips 18:19:56 each chip would have its own codename (say rv230) and would have been marketed under yet another name, like Radeon 8500 or Radeon 7000 or whatever 18:20:12 proposed #agreed 679579 - AcceptedBlocker - impacts desktop criteria for making for an unusable desktop for users with r100 or r200 series graphics 18:20:16 ack 18:20:35 proposed #agreed 679579 - AcceptedBlocker - impacts desktop criteria making for an unusable desktop for users with r100 or r200 series graphics 18:20:37 ack, with the conditions adamw mentioned before 18:21:08 going once ... twice ... 18:21:21 #agreed 679579 - AcceptedBlocker - impacts desktop criteria making for an unusable desktop for users with r100 or r200 series graphics 18:21:35 okay gang, let's walk through the proposed NTH's ... 18:21:50 #topic https://bugzilla.redhat.com/show_bug.cgi?id=699786 18:21:51 Bug 699786: medium, unspecified, ---, dcbw, NEW, VPN connections broken in the compat interface 18:21:53 guess i'd better start paying attention again 18:21:53 #info VPN connections broken in the compat interface 18:22:01 clumens: you're next buster! 18:22:43 so, as described, this means - no VPNs with knetworkmanager 18:22:43 Is there a criterion for VPN connections? 18:22:44 I don't think we'd consider anything VPN as a NTH 18:22:53 jsmith: NTH doesn't have criteria =) 18:23:01 adamw: waah, we have _loose_ criteria :) 18:23:02 OH NOES 18:23:13 loosest criteria in town 18:23:14 jlaska: well, i can see a situation where you couldn't get online with the live image inside a corporation or something, but yeah, it doesn't worry me hugely 18:23:22 in most cases you should be able to install an update to fix this 18:23:22 bugs which constitute infringements of the desktop-related Fedora_Release_Criteria as applied to non-default desktops 18:23:25 bugs which result in a system being unable to reach a graphical environment 18:23:28 significant installer bugs which do not meet the criteria to be blocker bugs 18:23:34 jlaska: KDE is a supported desktop, so that doesn't come into play here 18:23:45 jlaska: if this hit a criterion it'd be a blocker. it doesn't, though. 18:24:07 +1 to NTH, -1 to blocker in my book 18:24:11 ditto 18:24:13 didn't we accept the enterprise WPA bug as NTH? 18:24:32 tflink: yeah, i'd say that's probably a more common case than requiring a VPN to get any network connectivity. 18:24:49 remember, the main test for NTH is 'do we really need to fix this bug in shipped images' 18:24:55 right 18:24:59 if we can fix it well with an update, it's quite hard for it to be NTH 18:24:59 is this needed to pull down updates? 18:25:06 yeah, and a livecd connecting to VPN would most likely be a custom job 18:25:29 so unless you could often get into a situation where you need a VPN to pull down updates...i'm probably -1 on this 18:25:55 let's ask that in the bz too 18:26:09 I can think of situations where it would be a bit of a pain without VPN but I can't really come up with one that would keep you from updating 18:26:26 well i think we can vote for now and they can re-propose if they don't like our vote... 18:26:30 proposed #agreed 699786 - RejectedNTH - doesn't appear to impact ability to download and install updates 18:26:33 ack/nak/patch 18:26:35 * adamw votes -1 nth -1 blocker 18:26:36 ack 18:26:39 ack 18:26:49 tflink: only one i know of is places that use vpn for wireless security 18:27:00 i know the uni i went to does something like that 18:27:07 dgilmore: yeah, thats what I was thinking too, but you could get wired 18:27:17 at least in the situations I've seen 18:27:19 you cant get anywhere on the wireless network until you vpn 18:27:28 tflink: true 18:27:40 my school is pretty much like that, too 18:27:51 if that turns out to be more common, I'd recommend patching the criteria to include VPN? 18:28:07 they also make you use a socks or proxy server to do the bandwidth accounting 18:28:45 i've got "If you believe there are significant cases where it would be difficult or impossible to pull down updates without VPN connectivity, please re-propose with examples - thanks!" in the text for my comment 18:28:50 if that makes people happy :) 18:28:52 thx ... was going to ask you ... 18:28:58 works for me 18:29:04 okay 18:29:05 move on? 18:29:23 #agreed 699786 - RejectedNTH - doesn't appear to impact ability to download and install updates. Will reconsider if VPN turns out to be required for downloading package updates 18:29:29 adamw: +1 :-) 18:29:31 #topic https://bugzilla.redhat.com/show_bug.cgi?id=696943 18:29:33 Bug 696943: medium, unspecified, ---, akozumpl, POST, logs got by virtio are different from the originals on guest 18:29:37 #info logs got by virtio are different from the originals on guest 18:29:47 NTH criteria - can't be fixed in an update, it's installer 18:30:15 and in terms of need, this is something hongqing has been exploring w/ ales and he's hoping to use this for monitoring virt installs in some automation 18:30:35 the fix is described as low-risk, so +1 18:30:53 yeah, seems good to me. 18:30:54 +1 from me, obviously 18:31:04 https://admin.fedoraproject.org/updates/gnome-session-3.0.1-2.fc15 is the update with the r100/r200 blacklist 18:31:08 +1 18:31:16 +1 18:31:21 #agreed 696943 - AcceptedNTH - low-risk allows for improved virtio logging during installation 18:31:30 mclasen: cool, thanks 18:31:36 #topic https://bugzilla.redhat.com/show_bug.cgi?id=699923 18:31:37 Bug 699923: low, unspecified, ---, bcl, NEW, livecd doesn't have /var/log/dmesg 18:31:45 i'm all for this one too 18:31:48 e-z patch 18:32:06 so this is to make sure proper logs show up in automated reports from livecd installer problems? 18:32:08 "This would be nice to have so that we don't need to request that users add 18:32:11 sounds good, simple fix, +1 18:32:12 /var/log/messages to bugs manually." 18:32:16 +1 NTH 18:32:19 +1 to NTH 18:32:21 +1 NTH 18:32:26 adamw: yep 18:32:38 +1 nth 18:32:48 #agreed 699923 - AcceptedNTH - ensure consistent log locations for bug reports from livecd environment 18:32:55 #topic https://bugzilla.redhat.com/show_bug.cgi?id=700085 18:32:56 Bug 700085: unspecified, unspecified, ---, akozumpl, POST, iSCSI login dialog allows unselecting all devices and continuing - unable to login to any iSCSI devices after 18:32:59 #info iSCSI login dialog allows unselecting all devices and continuing - unable to login to any iSCSI devices after 18:33:10 this was a minor UI nit that I requested 18:33:28 +1 NTH 18:33:30 if you aren't paying attention, you can prevent adding any iSCSI volumes during install 18:33:42 clumens can speak to the patch risk 18:33:52 looked good to me 18:34:06 since we're early and have time to fix any explosions, sure, +1 18:34:24 I'll test this in an updates.img before TC1 18:34:27 +1 18:34:34 if there are bugs, it shouldn't hit anything outside of iscsi 18:34:47 #action jlaska - test 700085 using an updates.img prior to Final TC1 18:35:20 #agreed 700085 - AcceptedNTH - low risk installer iSCSI user-interface change 18:35:26 one more! 18:35:27 #topic https://bugzilla.redhat.com/show_bug.cgi?id=694897 18:35:28 Bug 694897: unspecified, unspecified, ---, cwickert, ASSIGNED, Icons missing in desktop menu 18:35:32 #info Icons missing in desktop menu 18:35:38 this one had an issue in the whiteboard, fixed 18:35:51 oh, i screwed this up 18:35:53 so we're good, whee 18:36:03 oh, already accepted? 18:36:10 yeah 18:36:11 Appears to be 18:36:14 #info Already AcceptedNTH ... no need to revisit 18:36:18 we marked it as AcceptedBlocker not AcceptedNTH 18:36:23 yep, but it had acceptedblocker in the whiteboard, so the script picked it up 18:36:32 Ladies and gentlemen ... superb job, we are done with the list for today 18:36:37 #topic Open Discussion 18:36:38 w00t! 18:36:40 * tflink is slower than adamw, aparrantly. needs to type less 18:36:53 yay 18:36:57 Again, I can't thank you folks enough for taking the time to work through the list each week 18:37:00 Kudos to each of you! 18:37:03 tflink: he's got the fastest phallenges in the west 18:37:05 you're damn right you can't 18:37:06 SEND BEER 18:37:11 HAHA! 18:37:22 you can tell it's the end of the meeting from the tone 18:37:25 wow 18:37:28 adamw: Be careful what you wish for! 18:37:30 =) 18:37:46 okay, any additional business folks would like to raise? 18:37:51 or, bidness, if you will? 18:37:51 * jsmith has nothing else 18:37:56 alright, i did all the secretary-ing from the point where i raised it - the first few bugs need doing 18:37:57 adamw: no beer for you 18:37:59 * adamw has nothing else 18:38:02 dgilmore: awww. 18:38:03 adamw: I'm on it 18:38:37 * adamw outta here 18:38:41 have a great time @ LinuxFEST northwest, to any who are attending! 18:38:43 adamw: i think we can get you an espresso drip though 18:38:45 if anyone's gonna be at lfnw i'll see 'em there 18:38:56 thanks everyone, I'll send minutes to your favorite lists 18:38:59 * dgilmore wont be there 18:39:07 * jlaska will #endmeeting in 30 seconds 18:39:25 jl gracious 18:39:38 #endmeeting