15:00:02 #startmeeting Fedora QA Meeting 2010-09-13 15:00:02 Meeting started Mon Sep 13 15:00:02 2010 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 15:00:02 Useful Commands: #action #agreed #halp #info #idea #link #topic. 15:00:14 #meetingname fedora-qa 15:00:14 The meeting name has been set to 'fedora-qa' 15:00:19 #topic gathering 15:00:29 $TIMEZONE_APPROPRIATE_GREETING, everyone 15:01:12 who's around? 15:02:10 * fenris02 waves 15:02:14 * adamw watches dust ball roll by 15:02:18 hiya fenris02 15:02:33 robatino: hi 15:02:40 hi 15:03:33 hey, quiet group today.. 15:03:45 wwoods: ping? 15:04:20 oh well, let's get started 15:04:26 * Viking-Ice sneaks inn 15:04:27 #topic previous meeting follow-up 15:04:29 heya viking 15:04:44 so the last meeting was 2010-08-30, we skipped last week 15:04:44 http://fedoraproject.org/wiki/QA/Meetings/20100830 15:05:10 "adamw contact lmacken about proventester metrics https://fedorahosted.org/bodhi/ticket/456" 15:05:20 well, I poked the ticket, and luke's just in the last day or so updated it 15:05:51 he writes "As for proventester metrics, that is something that takes longer to generate at the moment. We'll tackle that in Bodhi v2.0. "; I may ask him for some more info on that 15:06:00 lmacken: not around at present, are you? 15:07:00 spose not! 15:07:09 "jlaska build and deploy new autoqa release " 15:07:22 er, anyone know anything about this? I'm probably guilty of not keeping as up to date with autoqa as i should 15:08:44 maybe we'd better leave it on for next time, then... 15:09:00 adamw: I am! 15:09:12 ooh, yay. /me wonders if he's lagging 15:09:16 * jskladan lurks 15:09:21 * kparal joins late 15:09:27 and here's kparal and jskladan, yay... 15:09:29 so, going back in order 15:09:45 lmacken: so you mean we'll get proventester metric reports when bodhi 2 is in place? 15:10:13 adamw: the latest bodhi code that I'm in the process of deploying to staging right now will list unapproved critpath updates at the top of the updates-testing digest mails. 15:10:34 the other part of the ticket you created was to list the top proventesters, which will have to wait due to current database constraints 15:10:49 awesome. thanks. 15:11:09 what's the timeframe on bodhi 2 (revenge of bodhi) looking like? 15:11:19 shall I have bodhi spam proventester-members@fp.o with other events? 15:11:33 adamw: hopefully by F15? 15:12:09 bodhi currently has a nagmail job that spams maintainers when their critpath update is unapproved for 2 weeks... however, that's not useful to the maintainer at all -- so shall I send it to proventester-members@ instead? 15:12:24 I could also have it send a mail for each new critpath update upon submission? 15:12:27 hum, possibly...maybe do it as a trial and we'll see who complains 15:12:35 sounds good :) 15:12:48 i'll contact you outside of the meeting and we can set it up, be best to send an announcement email before starting it 15:12:58 ok 15:13:05 anyone have an opinion whether that's a good idea? 15:13:24 anyone can get the new critpath updates via RSS, but some may prefer email :\ 15:13:33 definitely for the 2-week mark ones (maybe even 1 week w/o approval?) 15:13:44 * Viking-Ice throws in ... https://fedoraproject.org/wiki/How_to_debug_Systemd_problems 15:13:54 * wwoods is here, whoops 15:14:05 Viking-Ice, thanks for putting that together. nice page. 15:14:19 Viking-Ice: can we keep it for open discussion? 15:14:25 trying to stay on one topic at a time 15:14:32 yup 15:14:46 lmacken: okay, i'll email you after the meeting 15:14:48 moving on... 15:14:54 I just threw that one so back to topic I myself am a rss junkie 15:15:12 kparal: we had an action item for jlaska last time: "jlaska build and deploy new autoqa release " 15:15:17 kparal: do you know how he did with that? 15:15:25 adamw: that's done 15:15:33 * kparal looking for release number 15:15:34 coolbeans 15:15:42 we can keep details for the autoqa update later i guess 15:15:47 okay, on to this week's agenda... 15:16:04 #topic F-14-Beta TC1 testing -- feedback needed! 15:16:14 so, just like it says on the tin - TC1 is out, we should be testing it 15:16:47 the install test matrix is pretty full, mostly green with some red dotted in there: https://fedoraproject.org/wiki/Test_Results:Current_Installation_Test 15:17:12 desktop matrix is empty, so far, doesn't help that we don't have official live images: https://fedoraproject.org/wiki/Test_Results:Current_Desktop_Test 15:17:16 i'll try and do something about that 15:18:12 anyone have any comments on tc1? 15:18:18 the RC is scheduled for the 16th - will there be a TC2? 15:19:05 i don't know the answer to that for sure, i suspect we probably don't need one...oxf13, around? thoughts? 15:21:12 guess he isn't 15:22:15 #info TC1 is out and in testing, looking quite good, RC1 scheduled for 09-16 15:22:25 #topic F-14 Test Days 15:22:47 so, we had the systemd test day on 2010-09-07: https://fedoraproject.org/wiki/Test_Day:2010-09-07_Systemd 15:22:56 the results were pretty useful and mostly encouraging 15:23:15 i sent out a recap - http://lists.fedoraproject.org/pipermail/test/2010-September/093463.html 15:24:07 any notes on that? 15:25:00 ok 15:25:07 #info systemd test day completed and recap sent out 15:25:36 I personally am on the opinion that systemd is ready for F14 15:26:06 that's mostly a decision for fesco; they've asked for our input and i said about the same thing, but it's in their hands now 15:26:27 jlaska and i also said that we thought it wouldn't be impossible to revert to upstart between tc1 and rc1, but it would be somewhat risky 15:26:34 yup ( going through the fesco ticket now ) 15:26:47 we said we thought rc1 was the deadline for switching: whatever we put in beta rc1 should be what goes in final 15:28:34 ok, so coming up this week we have the installer i18n / keymap test day - basically checking support for non-english languages and keyboards in the installer 15:29:00 that's https://fedoraproject.org/wiki/Test_Day:2010-09-16_AnacondaTranslationKeyboard 15:29:42 igor suares and hurry have been working hard on this one and it looks pretty well set up, so that's great 15:29:49 anyone have questions/concerns/comments on that event? 15:31:27 ookay, sounds good then =) 15:31:33 :) 15:31:33 just that a Test Day for this thing is an Really Great Idea 15:31:36 and long overdue 15:31:37 heh 15:31:45 yup it really is 15:31:57 we did some half-assed testing last release but this is much better and looks like they've done great work on it 15:32:11 so kudos and many thanks to igor and hurry for getting it set up 15:32:26 if it works out well, we might consider making it a per-release thing 15:32:27 #info installer keymap / translation test day is looking prepped and ready to go, group vote of thanks to igor and hurry for preparation 15:32:31 definitelu 15:32:32 y 15:32:44 it's one of those things that's always going wrong :) 15:33:13 so the following week is the virtualization test day: no page for that yet, but we have a ticket - https://fedorahosted.org/fedora-qa/ticket/114 15:33:35 I'm around. No current plans for a TC2, however that could just be a ticket away 15:33:56 jforbes filed the ticket and volunteered to run the test day also, so if you're around jforbes, here's a quick reminder that it's coming up in 10 days now :) if you need help prepping, do let us know 15:34:18 Oxf13: ok, thanks. i'm not aware of a pressing reason for one, we had the firstboot issue with tc1 but the parameters of that are well known and the fix looks fine from testing 15:34:30 adamw: yup, aware of it :) 15:34:35 I will get the pages up this week 15:34:40 jforbes: awesome 15:34:41 most of it is recycled content 15:35:24 always the easiest 15:36:43 #info jforbes is aware of the pending virtualization test day and is confident he'll have it prepped in time 15:37:06 okay, that's test days... 15:37:17 #topic AutoQA 15:37:24 time to hand over to wwoods / kparal! 15:37:30 #admin wwoods 15:37:36 d'oh, is that the wrong one? 15:37:49 uh 15:37:50 depends what you want to do? 15:37:51 do what now? 15:37:56 make you a mod for the meeting... 15:38:06 * adamw forgets the command 15:38:12 chair? 15:38:19 that's it 15:38:21 let kparal do it - I think you've done this before? 15:38:25 #chair kparal 15:38:25 Current chairs: adamw kparal 15:38:35 wwoods: it's just so you can do action items and stuff while doing the autoqa update if you like 15:38:40 #chair wwoods 15:38:40 Current chairs: adamw kparal wwoods 15:38:45 if you don't want to use the powahs you can ignore them =) 15:38:52 alright :) 15:39:05 jlaska mentions package acceptance stuff in the agenda, and of course anything else you've been working on, let us know 15:39:23 well let's see - I'll give a quick depcheck update 15:39:30 wwoods: go ahead 15:40:06 depcheck has a bunch of new unittests handling various multilib situations 15:40:20 one of the newer unittests proved that depcheck alone cannot prevent problems like the nss-softokn error 15:40:36 so we'll need another test to guard against that 15:41:02 another test showed that depcheck doesn't currently prevent *file* conflicts from happening 15:41:02 cool 15:41:18 it only checks the RPM headers, at the moment. I think that's probably wrong, though 15:41:24 #info wwoods has been creating unittests for depcheck to investigate various multilib cases 15:41:50 and so I'll be adding an RPM test transaction to depcheck - which is what yum does before it installs packages 15:42:11 #info wwoods will also add an rpm test transaction to depcheck to catch file conflicts 15:42:47 we're also getting very close to being able to use mash in depcheck, which will ensure that we're properly testing the correct set of packages as input 15:43:07 jskladan informs me that he's just gotten mash to run and produce some output, so yay! 15:43:44 I'm hoping depcheck will be giving us useful (informational only) data very soon - before the end of the month, at the latest. 15:43:48 jskladan: heh, but you have not seen the output yet :) 15:43:51 and that's all for depcheck. 15:43:54 but let's stay tuned :) 15:44:00 jskladan: heh true! I'm still hopeful, though 15:44:09 #wwoods hoping that depcheck will be able to provide informational data by the end of the month 15:44:38 ok, some more updates: 15:44:50 jlaska released autoqa-0.4.0-1 recently 15:45:09 some big changes are inside, see changelog 15:45:09 #info jlaska has pushed the new autoqa 0.4.0 15:45:13 (producing *correct* data is really the reason this is taking so long - we want to be nearly 100% sure that it's doing the right thing before we start making it bother people.) 15:45:37 and I have posted some patches fixing problems in multihook tests 15:46:01 oh, and yes - autoqa will be presented on FUDCon Zurich, this week, by me 15:46:11 everyone who's anyone will be there 15:46:21 so we suppose some new people that will come and have a look at the project 15:46:22 #info kparal presenting on autoqa at FUDCon Zurich 15:46:29 adamw: :D 15:46:48 did you check in with wwoods on presentation ideas? he already presented autoqa at the last FUDCon NA so he may have good ideas for you 15:47:18 adamw: I sent out slide drafts, I'll gladly hear any recommendations 15:48:07 ok, and those last 3 things together make one question, mainly for wwoods, I guess 15:48:18 kparal: if you'd like a look at my slides.. hang on for a link 15:48:24 will we want to make another release including those multihook tests soon, before fudcon happens? 15:48:40 s/tests/patches 15:48:49 http://tinyurl.com/autoqa-slides 15:49:03 kparal: yes, if at all possible 15:49:22 I'll be reviewing your patchset directly after the meeting 15:49:29 wwoods: if you can review the patches, jlaska told me he should be able to make another release 15:49:46 and if anything goes wrong, he will deploy back the current release 15:50:24 #info team would like to make a new release with multihook tests before fudcon if possible 15:51:14 ok, that's it I guess, no more updates 15:51:24 come to FUDCon for more info about AutoQA! :) 15:51:31 will do :) 15:52:00 thanks to autoqa team 15:52:05 #topic open discussion 15:52:08 soo, open discussion 15:52:12 i think viking had something for us 15:52:26 Well I finished the debug page for systemd 15:52:54 and I need to write another one for udev then I think we got the whole boot up covered 15:53:34 coolbeans 15:53:36 what was that link again? 15:53:49 https://fedoraproject.org/wiki/How_to_debug_Systemd_problems 15:55:03 #info viking-ice has completed his page on debugging systemd 15:55:08 thanks for that! 15:55:31 yeah I should have done that long time ago just things been crazy at $dayjob 15:56:24 don't worry 15:57:15 okay, so anyone have anything else for open floor? 15:58:15 * wwoods shakes head, sits on hands 15:59:00 much easier than the other way around 15:59:17 the 'inverse jazz hands', feared by every broadway actor 15:59:54 welp, looks like we're done... 16:00:32 thanks for coming, everyone! 16:00:34 #endmeeting