15:00:50 #startmeeting Fedora QA meeting 15:00:50 Meeting started Mon Jun 3 15:00:50 2013 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:50 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:53 #meetingname fedora-qa 15:00:53 The meeting name has been set to 'fedora-qa' 15:00:56 #topic Roll call 15:01:07 morning folks, who's ready for fun? 15:01:21 * jreznik is ready for another meeting in a row 15:01:27 * jskladan tips his hat 15:01:38 * j_dulaney lurks 15:01:39 * satellit listening in Bellingham motel 15:01:49 jreznik: /me is here 15:02:15 huh, that didn't work 15:02:30 * kparal here 15:04:11 * brunowolff is here 15:04:11 * nirik is lurking as well 15:05:20 * adamw really needs to get that anti-lurking kit installed 15:05:45 * jreznik is working on an announcement for the devel-announce list, needs anti-lurking kit as he's more lurking now :) 15:06:17 #topic previous meeting follow-up 15:07:00 #info "adamw and anyone else to work on updating Common_F19_bugs" - I got that one done, it's pretty much up to date 15:07:24 #info "adamw to talk to rbergeron and make sure packagekit/policykit issue makes it into the beta release notes" - yup, got a mention of it into the release announcement, and no crazy slashdot conspiracy theories yet 15:07:51 #info "adamw to complete 'new model' criteria revision by converting the final criteria" - I started working on that Friday, I'll aim to complete it today or tomorrow 15:08:02 "tflink to update the iscsi test case for newUI" - tflink? 15:08:13 not done yet, on my todo list for today 15:09:31 .fire tflink first of the week 15:09:34 adamw fires tflink first of the week 15:09:34 aw, no bot in here? boo. 15:09:37 yay! 15:09:58 #info "tflink to update the iscsi test case for newUI" - not done yet but should be soon 15:10:12 #info "tflink to schedule first blocker review meeting for this week" - that was done, we had two review meetings last week 15:10:28 #info "adamw to look at extracting some test cases from Test_Day:2013-05-21_AnacondaNewUI_Followup for use as optional test cases (or even enforceable partitioning test cases) for the validation matrices" - not done yet but it's kind of a long-term plan, it's on my todo list 15:10:57 #info "martix to ensure spice test day is ready on time" - that seemed to be sorted out fine 15:11:58 #info "satellit_e to work on some test cases for SoaS" - satellit drafted up some ideas at http://wiki.sugarlabs.org/go/Fedora_19-sugar_test_cases 15:12:01 SoaS test cases? list of ideas http://wiki.sugarlabs.org/go/Fedora_19-sugar_test_cases 15:12:07 : ) 15:12:13 :) 15:12:20 sorry i didn't get a sec to review them yet, i'll do it today 15:12:27 any other follow-up items I missed? 15:13:49 I retested llvmpipe on a particular non-SSE2 CPU and it's still broken. 15:16:05 ah yeah I saw your comment 15:16:25 that's a bit of a pain. and you might be right about advising people with old machines to grab something other than GNOME 15:16:34 we could probably mark the bug as requiring a release note for that 15:17:03 #action adamw to mark LLVMpipe-without-SSE2 bug as needing a release note and explain why to docs team 15:19:56 #topic Fedora 19 Final planning 15:20:08 #info TC1 is due tomorrow, 2013-06-04, I will request the compose then 15:20:23 do we have any major worries at this point, or things we need to get in place soon? 15:20:38 * adamw can't think of anything urgent beyond the revised criteria 15:21:05 the sooner we have TC1 the better. let's request today :) 15:21:30 have we done much upgrade testing? I've seen several comments and bugs, but wasn't sure how much poking has been done so far 15:21:37 * jreznik would not be against requesting it sooner, if there's nothing we really want in tc1 15:21:47 I did some upgrade testing on UEFI 15:22:10 poor kparal touching UEFI 15:23:20 * kparal now has a master's degree in bios re-flashing 15:24:08 tflink: quite a few people have run fedup I think, and we did fill the matrix for beta 15:24:10 kparal: :) 15:24:38 adamw: yeah, I was just thinking about the less-straightforward configs that ended up causing trouble after F18 release 15:25:55 adamw: We should finally have a patch that resolves most of the UEFI variable issues 15:26:03 right, that's another area we could do with more test cases 15:26:11 mjg59: yay! so there _were_ still issues on the kenrel side? 15:26:18 what's the timeframe on landing that? 15:26:29 adamw: Well, no. Issues on the firmware side. 15:26:44 mjg59: what are the changes? 15:26:53 But I have finally come up with an elegant mechanism for making the damn things do what we want 15:26:55 mjg59: and how much you can do with fw side? 15:27:08 mjg59: nice to hear! 15:27:17 kparal: Change the threshold from 50% to 5KB, do something magic which will (in theory) force the firmware to trigger garbage collection 15:27:37 mjg59: sounds good. can we expect a blogpost about the details? 15:27:49 kparal: http://mjg59.dreamwidth.org/25091.html 15:27:54 mjg59: what do you achieve by changing 50% to 5 KB? 15:28:02 mjg59: thanks! 15:29:08 jreznik: It's less than the maximum variable size, so we can create a new variable that's larger than the remaining space 15:29:13 Or, rather, attempt to 15:29:42 mjg59: does it mean that kernel panics will never again be stored in uefi nvram, too risky? 15:29:43 * jreznik is reading the blog post 15:29:56 kparal: No, we should be safe in doing that 15:29:59 mjg59: any news on the timeframe? 15:30:13 adamw: It's on LKML, jwb has a scratch build 15:30:46 kparal: That's currently disabled because it turns out the code is shit (which may be, uh, my fault) 15:31:08 mjg59: well, I mean, getting it in an official f19 kernel build 15:31:55 adamw: Once we've got confirmation from some testing of the scratch build 15:32:15 kparal: hint hint :) 15:32:16 kparal: hint, hint ^^^ 15:32:24 thanks guys 15:32:27 adamw: hey, you were faster! 15:32:30 .fire jreznik there can be only one 15:32:30 adamw fires jreznik there can be only one 15:32:50 mjg59: can you give us a link to that scratch build? 15:33:00 Linked from https://bugzilla.redhat.com/show_bug.cgi?id=964335 15:33:06 #info mjg59 informs us a kernel patch is coming which ought to resolve most of the UEFI variable space issues. scratch build for testing linked from https://bugzilla.redhat.com/show_bug.cgi?id=964335 15:33:30 thanks matt! 15:33:31 I'll try to understand the blogpost and then test the build 15:34:47 * jreznik thinks he understands what mjg59 wants to achieve, hopes firmwares will be cooperative :) 15:34:54 #topic Test Days 15:35:47 #info https://fedoraproject.org/wiki/Test_Day:2013-05-28_Virtualization and https://fedoraproject.org/wiki/Test_Day:2013-05-30_Spice both went off pretty successfully last week, i saw lots of active testing on both 15:36:11 although - kparal, next time check in with the virt folks on the details of the live image; we kinda needed it to be installable for virt testing 15:36:29 they didn't say 15:36:34 why we need anaconda? 15:36:42 they weren't aware we dropped anaconda from test day lives 15:36:48 well, because you can't really do virt hosting from a live session 15:36:54 unless you have like 32GB of RAM 15:37:09 we just told people to use the beta lives and that worked out fine, so it wasn't a big deal 15:37:17 I see 15:37:43 in that case maybe a tailored Live wasn't needed at all 15:37:51 yeah, maybe 15:37:59 like i said, it didn't turn out to be a big problem, just a note 15:38:21 I'm waiting for dcbw to give me a list of packages for tomorrow's test day so that I can build a Live image for them 15:38:34 FreeIPA test day should be good, they're creating the image themselves 15:38:46 cool; 15:39:13 #info https://fedoraproject.org/wiki/Test_Day:2013-06-04_Network_Manager page looks ready, kparal is working on the live image 15:39:29 #info https://fedoraproject.org/wiki/Test_Day:2013-06-06_FreeIPA_Two_Factor_Authentication looks ready to go 15:39:46 any other test day notes from anyone? 15:42:11 #topic Open floor 15:42:19 welp, here we are folks, any other business/ 15:42:52 nothing I can think of, no 15:43:27 tflink: we have rather a lot of proposed blockers; should we do blocker review next? 15:44:02 probably, yes 15:44:11 oh, I hoped you wouldn't ask :) 15:44:11 that looks like at least 3 hours of fun 15:44:49 * tflink gets started on prep 15:45:14 kparal: c'mon, you know I never miss blocker review fun 15:46:42 well, I don't miss it either 15:46:43 * j_dulaney is off 15:46:51 Apologies for not participating 15:47:25 so that's blocker review fun in bat-channel #fedora-blocker-review at the top of the clock, folks 15:47:28 strap yourselves in 15:48:09 Note that if you are using rawhide kernels on f19, the grub(2) conf files don't (and won't) get updated. 15:48:42 oh, yeah, that bug's worth a call-out 15:48:47 do you have the number handy? 15:49:29 I can find it. I think it is closed though. 15:54:15 welp, sounds like that's everything 15:54:24 blocker review in 5 minutes folks, thanks for coming! 15:54:24 While looking, it looks like they might have pushed the fixed systemd to f19 today. Yeah! 15:54:28 yay! 15:54:31 #endmeeting