From Fedora Project Wiki

< QA‎ | Meetings

Attendees

People present (lines said):

  1. jlaska (92)
  2. kparal (59)
  3. maxamillion (30)
  4. wwoods (29)
  5. adamw (27)
  6. jskladan (15)
  7. skvidal (6)
  8. zodbot (4)
  9. Southern_Gentlem (1)
  10. jeff_hann (1)

Regrets:

Agenda

Previous meeting follow-up

  • current members of the 'qa' group ... apply to 'proventesters'
  • jlaska to follow-up w/ maxamillion after meeting

Fedora 13 RC test status

Upcoming test milestones:

  • 2010-04-29 - Test 'Final' Test Compose -- INPROGRESS
  • 2010-04-30 - Final Blocker Meeting (f13blocker) #3 (recap)
  • 2010-05-05 - Final Blocker Meeting (f13blocker) #4
  • 2010-05-06 - Test 'Final' RC

No upcoming test days:

Fedora 13 QA Retrospective
Ideas have already been coming in since F-13-Alpha. It's not too early to note QA issues you felt where handled well, or could have been improved. Your ideas will be used during F-14 QA goal planning. Voice your comments at Fedora_13_QA_Retrospective.

Proventesters check-in

  • The team discussed possible next steps, including ...
    • Mass subscribing current qa group members to proventesters
    • Updating bodhi to rely on proventesters group for critpath feedback
  • In order to move QA/JoinProvenTesters:Draft out of draft, what else is needed?
    • need test guidelines for proventesters (something to explain why they are needed), including how to prioritize update testing
    • need guidance for mentor responsibilities (manage proventesters requests, keeping documentation up to date)
      Kparal suggesting that proventesters may mentor others, within some reasonable limits, no need for special FAS group for mentors

Package Acceptance Test Plan check-in

AutoQA check-in

Several recent AutoQA updates, including:

  • jskladan made progress on the autotest-id to the autoqa test front
  • jskladan is adding more initscript LSB compliance tests to autoqa
  • skvidal provided a patch to allow maintainers a way to subscribe to AutoQA test results
  • wwoods finalizing post-bodhi-update watcher, expects a testable watcher by end of week

Open discussion - <Your topic here>

Please help test MODIFIED and ON_QA bugs

To contribute test feedback, please help test a MODIFIED or ON_QA blocker bug (see list).

Upcoming QA events

Action items

  1. jlaska to migrate approved FAS 'qa' members into 'proventesters'
  2. jlaska to request bodhi change to require 'proventesters' feedback for critpath

IRC Transcript

jlaska #startmeeting Fedora QA Meeting 14:59
zodbot Meeting started Mon May 3 14:59:44 2010 UTC. The chair is jlaska. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:59
zodbot Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:59
jlaska #meetingname fedora-qa 14:59
zodbot The meeting name has been set to 'fedora-qa' 14:59
jlaska #topic Gathering critmass 14:59
* adamw gets critically massy 15:00
* kparal gets massively critical 15:00
jlaska adamw: you're moving at a high rate of speed too, and as we all know f=m*a 15:00
jlaska kparal: hah! 15:00
* jeff_hann here 15:00
jlaska jeff_hann: welcome 15:01
adamw heh 15:01
* adamw send a call out to the critical massive 15:01
* jskladan Zug Zug. 15:01
jlaska jskladan: howdy 15:01
jlaska wwoods: lurking? 15:01
jlaska let's get started ... we've all got a lot on our plates 15:02
jlaska #topic Previous meeting follow-up 15:02
adamw om nom nom 15:02
jlaska I only captured 2 items from last week ... and we'll touch on them later in the agenda 15:02
jlaska so skipping those two items 15:02
jlaska any other follow-up from last week? 15:03
jlaska alrighty ... moving along ... 15:03
jlaska #topic Fedora 13 RC test status 15:04
jlaska I'm sure folks are already tuned into this topic, but just a reminder on where we are in the F13 schedule 15:04
jlaska The F-13 Test compose was provided last Thursday, and is in test as we type 15:04
jlaska #info F-13-Final-TC1 test results available at 15:05
jlaska many thanks to robatino for handling the announcement and wiki magic while rhe is out 15:05
kparal the results seem very good so far 15:05
jlaska kparal: that's great! 15:06
Southern_Gentlem ? 15:06
jlaska preupgrade seems to still be sensative to properly recovering from low disk space 15:06
jlaska According to the blocker bug numbers ... 15:07
jlaska #info 32 MODIFIED ON_QA F13Blocker bugs -- http://tinyurl.com/3ae37qy 15:07
* maxamillion is here 15:07
maxamillion sorry I'm late 15:08
jlaska #info 21 NEW ASSIGNED F13Blocker bugs -- http://tinyurl.com/3567tqk 15:08
jlaska maxamillion: no worries, you're double booked too! Welcome :) 15:08
maxamillion thankies :) 15:08
jlaska I know I have a large plate of bugs I need to provide test feedback on this afternoon 15:08
jlaska and we still have 21 NEW ASSIGNED bugs preventing the F-13-Final release candidate compose 15:08
adamw *trumpets* 15:08
* wwoods lurking 15:09
jlaska my "at risk" spidey sense is activating 15:09
jlaska wwoods: lurk away my good man! 15:09
jlaska #info there are no more scheduled test days for Fedora 13 15:10
* adamw is walking the blocker list atm 15:10
jlaska so at this point, it's all about verifying your bugs, and carefully reviewing incoming issues 15:10
adamw obviously the new/assigned bugs are the most worrying but it helps focus if we can close off as many of the ON_QA / MODIFIED ones as possible 15:10
jlaska I'd like to do a mass update of the MODIFIED ON_QA bugs ... just asking for updated test feedback. Do folks have any concerns about that? 15:11
jlaska any other thoughts related to F-13-Final testing before we move on? 15:12
jlaska #info for folks reading the recap, the F-13 QA schedule is available at http://poelstra.fedorapeople.org/schedules/f-13/f-13-quality-tasks.html 15:12
jlaska okay ... moving on to next topic ... 15:13
jlaska #topic Proventesters check-in 15:13
jlaska We discussed this briefly on the list last week 15:14
adamw jlaska: please don't do a mass update 15:14
adamw jlaska: i am doing more tailored notes on each bug as I walk 15:14
jlaska adamw: oh no? 15:14
adamw so it'd just be a duplication essentially 15:14
jlaska adamw: ah excellent, I'll refrain from the mass annoy 15:14
jlaska adamw: thx :) 15:14
jlaska Okay, with regards to /topic 15:14
maxamillion yeah .... 15:15
jlaska I tried to capture what needs to happen in the short-term and then long-term 15:15
jlaska so short-term, is it correct to say that we need to migrate 'qa' users into 'proventesters 15:15
maxamillion jlaska: did the proventesters group become authoritative in bodhi? 15:15
jlaska so that critpath bodhi feedback can continue? 15:15
jlaska maxamillion: not yet, we need to have testers migrated first I believe 15:15
maxamillion jlaska: yes, I think we need to move all curent 'qa' members into 'proventesters' so that we don't put any kind of road bumps in the way of forward QA progress in critpath 15:16
maxamillion jlaska: ah ok 15:16
jlaska maxamillion: dgilmore also noted that we'll need to migrate users by hand, there isn't a convenient script to mass migrating FAS users 15:16
maxamillion jlaska: *sweet* 15:16
jlaska maxamillion: I can take an action item to migrate them over today, and file a ticket against bodhi to request using 'proventesters' for critpath feedback instead 15:16
adamw remind me to be washing my hair that day 15:16
kparal how many members are in qa group? 15:17
maxamillion jlaska: that would be *awesome* 15:17
maxamillion kparal: quite a bit more than I had originally thought 15:17
maxamillion kparal: I'm logging in to check now 15:17
kparal 10 members 15:17
jlaska 10 approved right? 15:18
jlaska that's not too bad 15:18
maxamillion oh 15:18
maxamillion nvm 15:18
maxamillion I thought it was a lot more than that 15:18
jlaska maxamillion: there are a lot of unapproved requests iirc 15:18
maxamillion ah, ok 15:18
jlaska #action jlaska to migrate approved FAS 'qa' members into 'proventesters' 15:18
kparal dozens of unapproved 15:18
jlaska #action jlaska to request bodhi change to require 'proventesters' feedback for critpath 15:18
adamw because people kept applying to the group as they thought it was important 15:19
jlaska okay ... so once those 2 items are complete ... what's next? 15:19
kparal guidelines for mentors? 15:19
jlaska basically ... before we can remove QA/JoinProvenTesters:Draft from Draft 15:19
maxamillion jlaska: that is *the* question ... :/ 15:19
jlaska adamw: maxamillion you both were recommending providing guidance, and being less strict about membership 15:19
kparal guidelines for proventesters.... ) 15:19
jlaska does that mean we don't have a mentor program? 15:20
jlaska kparal: yes! 15:20
maxamillion kparal: +1 15:20
jlaska #info long-term, need test guidelines for proventesters (something to explain why they are needed) 15:20
adamw no, having mentors is fine, they do the 'providing guidance' bit. i think. 15:20
jlaska are mentors also the group that acts on the 'proventesters' group requests? 15:21
maxamillion jlaska: no, I like the idea of a mentor but I don't want it to be a formal checklist style mentoring, it should be more hands on 15:21
maxamillion jlaska: yes, I think they would be 15:21
kparal we need a document specifying how to do proventesters' work properly. because that will also help mentors do their guidance 15:21
maxamillion kparal: agreed 15:21
jlaska kparal: what types of things would be in that document? 15:22
jlaska like your package update acceptance plan? 15:22
maxamillion I like the package update acceptance plan 15:22
kparal jlaska: basically it would tell me which package update should be tested with highest priority and what should I look at when testing them 15:23
kparal to tell the truth, I don't even know how to test package updates myself 15:23
jlaska ah, so how to prioritize them, and what to do with them 15:23
kparal so some introduction into it 15:23
jlaska #info long-term, provide guidance on how to prioritize package update testing 15:23
adamw jlaska: i think the idea is mentors answer the group requests, yeah 15:24
jlaska nice, anything else we need to consider? 15:24
jlaska adamw: maxamillion: so do we need a wiki page about Mentor responsibilities? 15:24
maxamillion I think so, yes 15:24
jlaska #info long-term, need to outline mentor responsibilities 15:25
maxamillion a formal "this is how to test" I think might be hard to capture but I agree with kparal that it should be done 15:25
jlaska btw ... unless folks are chomping at the bit to take this on right now ... I'm just collecting ideas 15:25
jlaska once the bulk of testing F-13 is behind us, we can start to divide & conquer 15:25
jlaska yeah, I have a hard time capturing some of that stuff, but perhaps if we start with what we know and what we do now ... that'll get things moving in the right direction? 15:26
maxamillion jlaska: agreed, and since we're going to grandfather in the current qa members who weild the critpath karma sword I think we have plenty of time 15:26
jlaska anything else we need to think about or consider? 15:27
maxamillion I'm sure there are, but I can't think of any right now 15:28
jlaska do mentors need a FAS group? 15:28
jlaska or are they just approved 'proventesters' ? 15:28
kparal I would go the easy way 15:28
jlaska yes please! 15:29
jlaska :) 15:29
maxamillion jlaska: either one I imagine would work and people who are mentors can just be kept track of in a wiki doc or something 15:29
* jlaska has a conflict starting shortly ... 15:29
kparal proventesters may mentor others. of course supposing people are reasonable and don't start mentoring right after receiving membership 15:29
* adamw is on a very laggy line 15:29
adamw what else is on the agenda? 15:30
kparal adamw: QA/Meetings/20100503 15:30
jlaska adamw: an update from kparal on the package update acceptance test plan, AutoQA update, and open-discussion 15:30
jlaska #chair adamw kparal 15:30
zodbot Current chairs: adamw jlaska kparal 15:30
maxamillion kparal: +1 15:30
maxamillion (about the mentors bit) 15:31
jlaska maxamillion: kparal: that seems reasonable 15:31
jlaska okay, I think we've captured enough for now 15:31
maxamillion +1 15:31
jlaska #topic AutoQA check-in 15:32
kparal #info proventesters may mentor others, within some reasonable limits, no need for special FAS group for mentors 15:32
jlaska doh, wrong topic 15:32
jlaska #topic Package Acceptance Test Plan check-in 15:32
jlaska kparal: can you walk us through how things are looking on the PATP? 15:33
kparal ok, that's gonna be short: Package Update Acceptance Test Plan has been finalized 15:33
kparal draft status removed 15:33
jlaska YAY!!! 15:33
kparal it is now available here: QA:Package_Update_Acceptance_Test_Plan 15:33
kparal #link QA:Package_Update_Acceptance_Test_Plan 15:33
adamw yaaaaaay 15:33
kparal so now we know how to test. the thing missing is the implementation within AutoQA :) 15:34
jlaska details shetails! 15:34
jlaska :) 15:34
jlaska I linked to the 3 different autoqa milestones that implement the proposed plan in a previous meeting 15:34
adamw yeah, that'll just take a few minutes hehe 15:34
kparal well most of the tests are already available, we just don't have the infrastructure around 15:35
jlaska kparal: in keeping with the previous topic, do we want to also document the tests as wiki test cases/ 15:36
kparal yes, we have a ticket for it somewhere 15:37
wwoods oh man exciting stuff. 15:37
jlaska okay, cool! 15:37
kparal I think I mentioned it last meeting 15:37
* kparal wonders if he should take the chair? 15:39
kparal alright, jlaska is probably at the second meeting, let's move to another topic 15:40
jlaska kparal: can you move us through the AUtoQA update? 15:40
kparal so, who wants the mic? 15:41
jskladan OK, probably on me to talk (even thought there is not much to tell) 15:41
kparal #topic AutoQA check-in 15:41
* kparal corrects topic command :) 15:41
jskladan last week, i've been working on the autotest-id propagation to the autotest-client 15:41
jskladan (and subsequentely to the autoqa test) 15:42
jskladan so we can benefit from the direct links to stored results on the Autotest-server 15:42
jskladan hopefully the patch will be in upstream until the end of the week 15:42
wwoods so when a test finishes we actually have a link back to all the logs and everything, so we don't need to include so much junk in the email? 15:43
jskladan yes 15:43
wwoods just making sure 15:43
wwoods heh 15:43
kparal for example we can send just a summary/highlights in the email and link to the full results 15:43
jskladan hehe 15:43
jskladan other that that, i'm adding more LSB-compliancy check for initscripts 15:44
wwoods thanks for taking the lead on that one - I tried and failed to get that to work a while back 15:44
* jskladan steals all the stuff from the RHEL guys :) 15:44
jskladan other than that, jlaska told me, that he'll be packaging autoqa next week 15:45
jskladan so hopefully the functionality will be on the production server soon :) 15:45
jskladan any other updates from you guys? 15:45
wwoods jlaska has been working on visualizing the dependency graph needed for autoqa 15:45
wwoods it's terrifying 15:45
wwoods we're probably going to need to have a FAD with some Java packagers if we're expecting to get the entire thing accepted into Fedora 15:46
kparal well it's java, it is terrifying :) 15:46
wwoods indeed. 15:46
wwoods I've got a quick update on autoqa 15:46
* kparal takes the mic 15:46
jlaska we had a FAD in planning, but that's going to be rescheduled and go through planning again after F-13 is out the door 15:46
kparal oh sorry 15:46
wwoods but kparal is the chair, so I'll wait for him to hand off the mic 15:46
wwoods heh 15:46
kparal alrighty 15:47
maxamillion there are still java packagers around? .... I thought when dbhole orphaned half the java stack we were in trouble :/ 15:47
kparal so last week I spent a little time on skvidal's patch which he posted into our ML 15:47
wwoods ah, that's what I was going to talk about too 15:47
jskladan #info jskladan made progress on the autotest-id to the autoqa test front 15:47
kparal it should allow us to sent emails directly to package maintainers that are subscribed to receive it 15:47
jskladan #info jskladan is adding more initscript LSB compliance tests to autoqa 15:48
kparal there were few bugs and I think we will need some library functions tweaks, but it should be done pretty quickly 15:48
wwoods it's a per-package thing - maintainers can opt-in to getting email from rpmguard whenever it runs for certain packages 15:48
kparal I would like to post tested patch this week 15:48
skvidal kparal: I sent a second patch 15:48
wwoods we will probably want to document the opt-in procedure for interested maintainers 15:49
kparal skvidal: I know, I haven't seen it yet unfortunately 15:49
skvidal kparal: ? 15:49
kparal skvidal: but many thanks for your work 15:49
skvidal kparal: you didn't receive it? 15:49
kparal skvidal: received, but not reviewed :) 15:49
skvidal oh okay 15:49
wwoods and note that this is a temporary hack, to be used until resultsdb is able to handle this kind of thing 15:49
kparal and then we can expect many complaints about the contents :) 15:50
skvidal which is sorta the point 15:50
skvidal it'll help field what can be filtered later 15:50
kparal #info skvidal's patch should allow AutoQA to send direct emails to subscribed package maintainers 15:51
kparal ok, anything else from AutoQA? 15:51
wwoods one other thing 15:51
kparal go ahead 15:51
wwoods I'm still working on the post-bodhi-update watcher/hook - there are some limitations of the current bodhi design that make it tricky for us to do what we want 15:52
wwoods but I think I've got the design all worked out now 15:52
adamw sorry, my connect died 15:52
wwoods hopefully I will have it testable by the end of this week (so long as administrative things don't get in the way) 15:52
wwoods I'm also trying to write a big blog post to explain the depcheck test and why it's so hard 15:53
wwoods heh 15:53
* kparal looking forward 15:53
wwoods I'm trying to keep notes on things that would make the post-bodhi-update watcher/hook easier to implement, so we can make sure that stuff gets put into the design of bodhi2 15:53
wwoods (without bugging lmacken too much) 15:54
wwoods but yeah, keep your fingers crossed for some visible progress on that stuff. 15:54
wwoods that's all from me. 15:54
kparal ok, thanks wwoods 15:55
kparal as far as I see, we have open floor now 15:55
kparal #topic Open discussion - <Your topic here> 15:55
kparal anything you want to discuss? 15:56
jlaska kparal: nothing here, thank you for driving while I'm distracted :) 15:56
* adamw just focussing on blockers 15:56
adamw please everyone test some on_qa/modified, there's several anyone could test 15:57
kparal #info please test on_qa/modified bugs, there's several anyone could test 15:57
adamw thanks 15:57
wwoods link to the list? 15:58
wwoods (for lazyboneses such as myself) 15:58
kparal is this link correct? https://bugzilla.redhat.com/showdependencytree.cgi?id=507681&hide_resolved=1 15:58
kparal I know jlaska had two different links while back, but I think it should be the same 15:59
kparal #link https://bugzilla.redhat.com/showdependencytree.cgi?id=507681&hide_resolved=1 15:59
adamw that's all blockers 16:00
adamw i'll try and send out a mail later flagging up bugs that anyone can test 16:00
adamw i'm making a list 16:00
kparal #info adamw will send a list of easily testable important bugs 16:00
adamw hey hey hey, you just VASTLY upgraded my 'try' ;) 16:01
kparal adamw: now you have an action item :) 16:01
adamw hehe 16:01
kparal alright, one hour precisely 16:01
kparal if you don't have anything else on mind... 16:02
kparal thanks all for attending! 16:02
kparal #endmeeting 16:02

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!

Retrieved from "https://fedoraproject.org/w/index.php?title=QA/Meetings/20100503&oldid=472369"