QA/Meetings/20130225

From FedoraProject

< QA | Meetings
Revision as of 23:06, 26 February 2013 by Adamwill (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Contents

Attendees

  • adamw (106)
  • tflink (39)
  • Viking-Ice (33)
  • j_dulaney (22)
  • Martix (17)
  • nirik (10)
  • zodbot (4)
  • robatino (4)
  • jreznik (3)
  • misc (2)
  • nb (1)
  • abadger1999 (1)
  • mkrizek (1)
  • satellit_e (1)
  • pschindl (1)
  • viking-ice (0)

Agenda

  • Previous meeting follow-up
  • Call for Test Days
  • Trac tickets CCed to list
  • Open floor

Previous meeting follow-up

  • adamw to write a second draft (of the automatic blocker proposal) with andre's proposed changes and stronger explanation not to put 'grey area' bugs in the automatic blocker list - this was done
    • We agreed that second draft is ready to go into production
  • adamw to draft up changes to the blocker bug meeting SOP for 3-hour hard limit, no-reviews-during-qa-meetings, and a dedicated channel for meetings, send to list for further discussion - this was done
    • j_dulaney, tflink, viking-ice, jreznik all vote +1 on changes: adamw would like more releng/devel feedback before going to production
  • viking-ice to discuss the 'smoke test for spins' idea further with nirik and cwickert - not yet done

Call for Test Days

  • tflink suggests an upgrade test day, but notes issues with timing - we can try to co-ordinate with wwoods to handle that

Trac tickets CCed to list

  • What do we do about overly development-y trac tickets being CCed to test@?
  • viking-ice notes the qa trac was originally intended solely as a 'qa task management' thing, not for devel
  • tflink is provisionally +1 to at least a separate mailing list for qa-devel
  • Everyone agrees in general that having the bugs in QA trac and the discussion spammed to test@ is a bad idea
  • tflink will weigh various possible responses and make a detailed proposal to the list

Open floor

N/A

Action items

  • adamw to push 'automatic blocker' proposal to production
  • adamw to try and gather a bit more feedback on blocker process changes this week
  • viking-ice or adamw to file a trac ticket for the smoke-test-for-spins idea
  • tflink to take a look at the question of tracking qa tool discussion and bugs/tickets and make a broad proposal about what to do

IRC Log

adamw #startmeeting Fedora QA meeting 16:01
zodbot Meeting started Mon Feb 25 16:01:27 2013 UTC. The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot. 16:01
zodbot Useful Commands: #action #agreed #halp #info #idea #link #topic. 16:01
adamw #meetingname fedora-qa 16:01
zodbot The meeting name has been set to 'fedora-qa' 16:01
adamw #topic roll call 16:01
* adamw is here, also an idiot. 16:01
* tflink is here ... in both channels :) 16:01
* satellit_e listening 16:01
* mkrizek is here 16:01
* j_dulaney sends more fail to adamw 16:02
* Martix smells dead shark 16:02
Martix meat 16:02
nb hi 16:02
* jreznik is around, idiot as always :) 16:03
* adamw is also on a bus to whistler and phoning this one in 16:03
adamw alrighty! 16:03
* pschindl is here 16:03
adamw #topic Previous meeting follow-up 16:04
adamw note on this one - we may want to go a little more in depth on each of these, as they're kinda topics in their own right 16:05
adamw "adamw to write a second draft (of the automatic blocker proposal) with andre's proposed changes and stronger explanation not to put 'grey area' bugs in the automatic blocker list" 16:05
adamw so I did that, and sent it to the list; not much further feedback, does that mean everyone's OK with it? 16:05
tflink yeah 16:06
adamw https://lists.fedoraproject.org/pipermail/test/2013-February/113909.html 16:06
adamw cos if no-one yells, I'm gonna go ahead and put it live 16:07
robatino i was wondering if the boot criteria is hardware-specific and if that will cause problems 16:07
robatino since it may fail to boot only on some platforms 16:07
* j_dulaney is +1 16:08
* Viking-Ice joins in 16:08
adamw robatino: i tried to word it quite specifically 16:08
Viking-Ice ship igt 16:09
Viking-Ice mean ship it 16:09
Viking-Ice  ;) 16:09
adamw robatino: 'conditional failure is not an automatic blocker' basically means 'if it boots for anyone, it's not an automatic blocker' 16:09
adamw i could try and make that wording less legalistic :) 16:09
robatino ok, but if it fails to boot for one person they'll have to check with others before making it an automatic blocker 16:09
robatino which seems to make it similar to the situation with regular blockers 16:10
j_dulaney Comment to that affect? 16:10
adamw robatino: we can see how it shakes out in practice; what i'm thinking is that, usually, we get a pretty good handle on the actual cause of major bugs quite quickly 16:10
Viking-Ice yup 16:10
tflink yeah, that sounds like a plan to me 16:10
adamw it should be pretty clear if we know the actual cause of a bug whether it's a 'total DOA' or not 16:10
adamw you know, if the cause is 'we left vmlinuz off the image', then...:) 16:10
adamw #info "adamw to write a second draft (of the automatic blocker proposal) with andre's proposed changes and stronger explanation not to put 'grey area' bugs in the automatic blocker list" - this was done: https://lists.fedoraproject.org/pipermail/test/2013-February/113909.html 16:11
adamw #agreed second draft is ready to go 16:11
adamw #action adamw to push 'automatic blocker' proposal to production 16:11
adamw okay, on to: 16:12
adamw "adamw to draft up changes to the blocker bug meeting SOP for 3-hour hard limit, no-reviews-during-qa-meetings, and a dedicated channel for meetings, send to list for further discussion" 16:12
adamw I also did that: https://lists.fedoraproject.org/pipermail/test/2013-February/113910.html 16:12
adamw only really got one reply so far, from jaro: I was expecting more discussion 16:12
* j_dulaney votes to put it into effect 16:13
* tflink should have replied on-list but is +1 on the changes 16:14
adamw i'd feel more confident with a bit more list feedback, but hey 16:14
Viking-Ice sorry I've been to busy here in brno to catchup in what's been happening on all the mailing list but then again I'm kinda obvious +1 to those changes ;) 16:15
tflink I think that the only changes we haven't already been doing is the channel for meetings and the no-blocker-stuff-during-qa-meetings 16:15
* nirik thinks all those make sense. 16:16
jreznik adamw: consider it as my +1, I don't really see a need for further discussion 16:16
tflink but we can wait another week for comments, it's not like we have a blocker meeting this week 16:16
jreznik and we can always revisit... 16:16
adamw tflink: true 16:16
adamw #info "adamw to draft up changes to the blocker bug meeting SOP for 3-hour hard limit, no-reviews-during-qa-meetings, and a dedicated channel for meetings, send to list for further discussion" - this was also done, https://lists.fedoraproject.org/pipermail/test/2013-February/113910.html 16:16
adamw #info j_dulaney, tflink, viking-ice, jreznik all vote +1 on blocker process changes 16:17
adamw #action adamw to try and gather a bit more feedback on blocker process changes this week 16:17
adamw "viking-ice to discuss the 'smoke test for spins' idea further with nirik and cwickert" - viking, nirik, did you guys get anywhere with this? 16:17
nirik nope. 16:18
adamw concise! 16:18
adamw maybe we should have a trac ticket so we don't lose the idea, or something 16:18
Viking-Ice Well I actually met with cwickert here in brno but this topic eluded our discussion 16:19
Viking-Ice yeah we should add it the trac so it wont get lost 16:19
adamw #action viking-ice or adamw to file a trac ticket for the smoke-test-for-spins idea 16:19
adamw an action item to file a trac ticket...mmm, I can smell the bureaucracy 16:20
adamw #topic Call for Test Days 16:20
adamw so, many thanks to martix for taking charge of test days for this cycle 16:20
adamw #chair tflink viking-ice 16:21
zodbot Current chairs: adamw tflink viking-ice 16:21
adamw (forgot) 16:21
Martix your welcome :-) 16:21
adamw martix sent out the call for test days: https://lists.fedoraproject.org/pipermail/test/2013-February/113900.html 16:21
adamw we have quite a few submitted and planned already, it looks like, but did anyone have any ideas lying around to add to the list? 16:21
Martix I just went through proposals and trying to fit them in schedule right now 16:22
adamw #info martix is working proposals into the schedule at present 16:22
tflink upgrade might be interesting - a bit difficult to do with timing, though 16:22
adamw #info tflink suggests an upgrade test day, but notes issues with timing 16:23
adamw we could see if will has a timetable for fedup changes for f19 and try to co-ordinate 16:23
Martix 4/04 Printing 16:23
Martix 4/11 l10n 16:23
Martix 5/02 i18n 16:23
adamw the package set is usually stable enough for testing upgrades at least by beta 16:23
Martix 5/23 FreeIPA 16:23
Martix 5/30 Virtualization 16:23
Martix 6/06 SSSDImproveADIntegration 16:23
j_dulaney In theory 16:24
Martix that list of new proposals 16:24
Martix *thats 16:24
tflink adamw: true, the only variable then becomes which repos are being used 16:24
adamw we can fiddle with that for a test day 16:24
adamw sort of thing a test day lets you do, in fact 16:24
adamw any other ideas? 16:24
Viking-Ice I'm still of the notion we should get rid of that schedule 16:24
j_dulaney Is networking on the list 16:25
j_dulaney  ? 16:25
adamw j_dulaney: i believe it's already arranged 16:25
j_dulaney Okay 16:25
adamw Viking-Ice: how do you mean? 16:25
Viking-Ice adamw, instead of fixed schedule with explicit dates available we simply note down the time people want to host test day 16:26
adamw Viking-Ice: personally i still kinda like the idea of going mainly with thursdays just to help people fit it in to their schedules 16:27
Martix j_dulaney: I can extend "Network Manager Test Day" to "Networking Test Week" 16:27
adamw Martix: only if the networking folks feel it's needed, 16:27
Martix adamw: right, if they will come with this 16:27
j_dulaney Martix: It shouldn't be necessary 16:28
Viking-Ice adamw, the down side of that is that people look at a schedule and see the "thursday" they are free is occupied by some other component 16:29
j_dulaney The biggest thing I can think of off the top of my head with NM is the arrival (finally!) of a cli 16:29
adamw Viking-Ice: yeah, it's a cost/benefit thing indeed 16:29
Viking-Ice adamw, hence the schedule is a bad thing and hosting it only on thursday is even worse 16:29
adamw Viking-Ice: we should probably make that a separate topic for another meeting though, still two to get through here 16:29
Viking-Ice from my pov 16:29
Martix j_dulaney: nmcli testing is alredy planned for NM Test Day 16:29
j_dulaney Indeed 16:30
adamw welp, seems like that's all the ideas... 16:30
adamw as a heads-up, i may pop off for a few minutes in 5 mins or so, switching internet connections. anyhow 16:30
adamw #topic Trac tickets CCed to list 16:31
adamw so there's been some discussion lately about how it may not be good to have lots of development-related tickets CCed to test@ 16:31
adamw this is happening because we're using the QA trac instance for tool development 16:32
adamw i'm packaging a trac plugin which would allow us to direct the mails for different components to different places, which is one way of addressing the problem 16:32
* nirik can build the epel version and get it installed later today 16:32
j_dulaney Maybe have a qa-devel list? 16:32
adamw another suggestion is to set up another trac instance for tooling, or turn the autoqa trac instance into a more general qa-dev one 16:32
adamw nirik: i did a build, didn't submit an update though 16:32
Viking-Ice <shrug> plugs +1 to separate qa-devel trac instance for qa related development work 16:32
Martix my apologize, I just closed bunch of previous Test Day tickets :-) 16:33
j_dulaney No, that's not devel 16:33
adamw i think either approach would work, i don't really mind which - i figure tflink and martix and kparal maybe get the biggest say in what fix to pick, as they're the ones doing most of that work 16:33
j_dulaney Martix: That happens 16:33
adamw Martix: that's fine, those tickets are appropriate for the list 16:33
Viking-Ice fedora-qa was never supposed to be used for anything else but request from the community 16:33
* tflink doesn't care a whole lot either way about where the tickets live 16:33
adamw #info viking-ice says the qa trac was originally intended solely as a 'qa task management' thing, not for devel 16:33
nirik which doesn't mean it can't be used for other things now. 16:34
Viking-Ice yes an request tracker not bug tracker 16:34
nirik anyhow, whatever works. 16:34
adamw i don't really see a huge difference between the two approaches in the end, they'd achieve the goal, and either is pretty easy to do. 16:34
tflink but I'm +1 to at least discussing a qa-devel@ list - it's been on my list of things to propose 16:34
adamw #info tflink is provisionally +1 to at least a separate mailing list for qa-devel 16:35
* j_dulaney is also +1 16:35
Viking-Ice +1 to seperated mailing list and a trac instance 16:35
adamw i can see that a line between 'community tasks' and 'tool development' is a reasonable line to draw between two separate tracs, and it's not like trac instances cost money, so maybe we can just go with that 16:35
tflink if we move the blocker tracking app's tickets, I'd rather move to a separate instance instead of to autoqa trac, though - more granularity in ticket assignment 16:36
adamw nirik: is there a process you can point to for setting up a new trac instnace? just file a ticket with releng? 16:36
adamw er, admin 16:36
nirik adamw: file a ticket with infras 16:36
tflink yeah, the biggest cost would be my time in configuring stuff and moving tickets 16:36
* j_dulaney can do that 16:36
nirik a new trac will cost eleventy million dollars! (well, ok, not really, just file a ticket. ;) 16:37
adamw tflink: best do it early when there isn't a lot of work to do then i guess 16:37
Viking-Ice tflink, well arent you the one that's causing this mess in the first place with all your development ;) 16:37
adamw #info to get a new trac instance we just file a ticket with websites 16:37
adamw tflink: yes, damnit, stop making awesome tools ;) 16:37
tflink j_dulaney: if you're talking about the ticket moving and configuration, I'd rather have myself or mkrizek do that since we'll be the ones using it most for now 16:37
j_dulaney No, I meant file the ticket 16:37
j_dulaney And sit on nirik to do it :) 16:37
adamw whoops, i've gotta drop out briefly, back in ~5 16:37
tflink Viking-Ice: I suppose that's one way to think of it :-P 16:37
adamw tflink and viking can drive 16:37
tflink less tool work means more time for other tasks :-D 16:38
tflink anywho, I'm fine with whichever approach as long as we decide sooner than later 16:38
tflink trac is trac for the most part - migrating will mess up a few minor things relating to ticket numbers but these are small issues 16:39
Viking-Ice Well I'm +1 to seperated mailing list and a trac instance 16:39
nirik you could try the cc thing and if it doesn't work do a new one? 16:39
* nirik doesn't have a horse in the race 16:39
tflink nirik: did agilo ever get back into fedorahosted or is there still a conflict with another plugin? 16:40
Viking-Ice So I propose that we create qa-devel mailing list and qa-devel trac instance 16:40
nirik tflink: it conflicts. ;( 16:40
tflink Viking-Ice: if we do that, I'd like to combine that mailing list with autoqa-devel 16:41
tflink but I also want to send that proposal out to the list (autoqa-devel@) before actually doing it 16:41
Viking-Ice tflink, what does the autoqa people think about that ? 16:42
Viking-Ice same thoughts 16:42
Viking-Ice so we should postpone until feedback from them? 16:42
tflink there have been some small discussions around it - the conclusion was mostly "let's see how many other things qa-devel related have much discussion" 16:42
j_dulaney Combine auto-qa list and qa-devel list, but keep seperate tracs for the two 16:42
tflink yeah, I'm strongly -1 on moving the blocker tracker app tickets to the autoqa trac 16:43
Viking-Ice these project are hosted on fedorahosted right 16:44
tflink either way, it might be better to float a proposal on test@ before making changes 16:44
Viking-Ice and there they do have their own bug trac right so why not use those then? 16:44
tflink Viking-Ice: depends on what's requested 16:44
tflink you don't have to request 1:1 trac:repo/project 16:45
* adamw back 16:45
Viking-Ice tflink, I see well perhaps that's the problem then 16:45
tflink Viking-Ice: how so? 16:45
adamw tflink: do you want to take an action item to look into the options and make a proposal on what new stuff to create? 16:45
Viking-Ice tflink, missing trac instance for those projects 16:46
Viking-Ice like upstream bugzillas 16:46
tflink eh, I specifically didn't request one 16:46
* tflink is not thrilled about another trac instance to admin 16:46
adamw yeah, i think one trac for all qa tools might be the best option 16:47
adamw trac is kind of a pita to admin 16:47
tflink adamw: if we were talking about another bug tracker, maybe 16:47
tflink trac isn't really set up to do multiple projects well, IMHO 16:47
adamw okay 16:47
* j_dulaney just thought he saw abadger1999 out of the corner of his eye, but it was someone that looked remarkably like him 16:47
abadger1999 heh 16:47
adamw i think we're at the point where it'd be best for someone to go look at the issue and come up with a broader proposal...i think all the stuff to consider has been raised 16:48
adamw okay if i give that to you tflink? 16:48
tflink yeah, I can do that 16:48
adamw #action tflink to take a look at the question of tracking qa tool discussion and bugs/tickets and make a broad proposal about what to do 16:48
adamw we can discuss tflink's proposal next week (or when it gets done) 16:49
adamw that'll give us more detail to chew on 16:49
adamw #agreed everyone agrees in general that having the bugs in QA trac and the discussion spammed to test@ is a bad idea 16:49
adamw #topic Open floor 16:49
adamw so, anything for open floor, folks? 16:49
Viking-Ice audio in desktop criteria 16:50
Martix updated https://fedoraproject.org/wiki/QA/Fedora_19_test_days with new proposals 16:50
Viking-Ice so how do people feel that we add audio to the desktop criteria you know press play and actually get sound of speaker ? 16:51
adamw i think we already have that 16:52
tflink it's in the test cases, not 100% sure about criteria 16:52
adamw Beta #19 16:52
adamw " In most cases, the installed system must be able to play back sound with gstreamer-based applications (see Blocker_Bug_FAQ) " 16:52
adamw if anything i think Beta is a bit early, but it's in there. 16:52
Viking-Ice alriiighhhty then ;) 16:53
tflink nvm, then 16:53
adamw 'gstreamer-based applications' is a bit desktop team-specific - that's one of the ones that came straight from desktop team back at FUDCon Whatever and never got 'abstracted' 16:53
adamw i can improve that as part of the criteria revision stuff. which i'm still working on. 16:53
Viking-Ice I just got asked here in BRNO about that but was unsure if we actual had that 16:53
adamw well, there ya go :) 16:53
tflink aren't most audio things using gstreamer, anyways? 16:54
tflink I didn't think that was gnome-specific 16:54
adamw tflink: oh, right, i think KDE defaults to gstreamer backend these days too 16:54
adamw they have an abstraction layer on top of gstreamer because, you know, yo dawg i heard you liked audio abstraction 16:55
tflink it's maintained outside of the gnome project, anyways 16:55
Viking-Ice yup anything else anyone? 16:56
* adamw sets fuse for 9am 16:56
adamw i have snow to abuse 16:56
* j_dulaney thinks anyone that actually *likes* snow is clinicaly insane 16:57
adamw i certainly am clinically insane, but i don't think the diagnosis was made on the basis of fondness for snow ;) 16:57
tflink j_dulaney: tell that to the huge ski/snowboard industry :) 16:57
tflink or ice fisherman 16:57
misc adamw: not only on that 16:58
adamw misc: they let me out of the institution on weekends! 16:58
misc adamw: that's not because everyone is crazy there that you should your employer "the institution" 16:59
* j_dulaney would much rather it be 100 F 16:59
adamw misc: haha. 16:59
tflink j_dulaney: and you call me crazy ... 16:59
adamw Red Hat: Keeping Crazy Engineers Off The Streets Since 1998 16:59
adamw alrighty, thanks for coming everyone 16:59
adamw same time next week 17:00
adamw #endmeeting 17:00

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!