14:59:56 #startmeeting Fedora QA Meeting 14:59:56 Meeting started Mon Jun 28 14:59:56 2010 UTC. The chair is jlaska. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:59:56 Useful Commands: #action #agreed #halp #info #idea #link #topic. 14:59:59 #meetingname fedora-qa 14:59:59 The meeting name has been set to 'fedora-qa' 15:00:02 #chair kparal 15:00:02 Current chairs: jlaska kparal 15:00:11 #topic Gathering critical mass 15:00:21 * jskladan steps out of the shadows :) 15:00:38 * kparal looks around 15:00:55 * wwoods puts down his beer^Wcoffee 15:00:59 jskladan: kparal: greetings 15:01:03 wwoods: it's noon somewhere :) 15:01:05 i d highly suggest to add the eds mess 15:01:12 to the agenda 15:01:13 * Viking-Ice stands since there are no chairs left! 15:01:20 hicham: why, exactly, is that a QA problem? 15:01:46 wwoods: i think it is qa problem, since updates are now broken 15:01:47 hicham: you're welcome to bring up any QA suggestions related to that build during open discussion 15:02:14 hicham: yes, but we already have plans for tests in place to keep that sort of thing from happening 15:02:35 wwoods: ah ok, thanks 15:02:37 in the meantime it's still up to developers to *not* mess up the deps in the repos 15:02:49 and they keep doing it 15:02:54 this does not change our plans any 15:03:02 nor is it our fault that it keeps happening 15:03:03 morningalydoodly 15:03:06 hicham: for offline reading enjoyment ... https://fedoraproject.org/wiki/QA:Package_Update_Acceptance_Test_Plan 15:03:09 adamw: welcome 15:03:14 Viking-Ice: lurking? 15:03:23 Yup 15:03:25 hey there 15:03:48 alrighty, let's get movin' 15:03:58 I completely understand the frustration though 15:04:10 I'm going to breeze through the follow-up items ... I've got details on most of them ... 15:04:19 #topic Previous meeting follow-up 15:04:24 #info adamw to forward wwoods a good explanation of the nss-softokn problem scenario to ensure autoqa catches it in future 15:04:30 #link https://fedorahosted.org/pipermail/autoqa-devel/2010-June/000703.html 15:04:44 next up ... 15:05:05 wwoods: you were out last week, so surprise, I added an agenda item for you! 15:05:13 haa 15:05:14 queue the balloons and streamers 15:05:26 hooray! lucky winner! 15:05:28 wwoods: dunno if you have any updates, since this might be the first you saw of it ... 15:05:34 #info wwoods to evaluate nss-softokn dependency problem for proper 'depcheck' coverage 15:05:58 I don't recall the explanation of the exact dep problem 15:06:29 wwoods: we can follow-up on autoqa-devel list ... checkout the description adamw sent (see previous #link) 15:06:40 * skvidal recalls 15:06:47 * adamw re-assigned all his work to wwoods and went golfing 15:06:50 the nss-softokn thing was just a multilib issue iirc 15:06:51 check your to-do list 15:07:05 x86_64 version was older than i686 one 15:07:07 ie: the lib didn't get auto pulled in as multlib by mash for the update 15:07:14 but it did get pulled in for the GA 15:07:19 hicham: right 15:07:19 I think the only action here is to review whether this is something depcheck would have captured 15:07:21 skvidal: as jlaska says i already sent a description to autoqa-devel list 15:07:27 adamw: sorry 15:07:30 npnp :) 15:07:48 wwoods: okay taking this to the list? 15:07:54 yeah, that's fine 15:07:55 * jlaska has Michael McDonald in his head now 15:08:01 wwoods: thx 15:08:08 next up ... 15:08:11 #info Viking-Ice will start 'how to debug systemd problems' guide and send to list for review 15:08:16 quick summary: we'll want to write test case for that situation - and all the other common "your repos asplode" problems 15:08:29 yup have started to work on https://fedoraproject.org/wiki/User:Johannbg/QA/Systemd#Quick_Debugging_Tips 15:08:33 Viking-Ice: I see you've already got quite a bit of content 15:08:41 #link https://fedoraproject.org/wiki/User:Johannbg/QA/Systemd#Quick_Debugging_Tips 15:08:47 Crap, was in the head too long 15:09:08 jskladan: welcome :) 15:09:40 Viking-Ice: very nice ... anything you want to follow-up with next week ... or should we stay tuned to test@ mailing list 15:09:49 We hit a bit of snag at work this last week so the work got a bit halted + I still have not contacted Lennart about several things. 15:10:03 I will post the page for the list for review once I'm done 15:10:15 Viking-Ice: awesome, thanks again for tackling this 15:10:26 has releng got any plans on when we will switch to use systemd by default 15:10:29 up next ... 15:10:32 #info robatino to ask infrastructure team to archive previous RCs/TCs 15:10:37 #link http://fedorahosted.org/fedora-infrastructure/ticket/2241 15:11:23 I haven't seen any updates on this yet, we may wish to reach out to infrastructure@ list also 15:11:31 next ... 15:11:34 #info jlaska to review F14 test day wiki content 15:11:52 Nothing earth shattering here, just priming the wiki content for F14 test days 15:11:55 #link https://fedoraproject.org/wiki/QA/Fedora_14_test_days 15:12:11 If anyone has ideas or suggestions, feel free to write them down on the talk page -- https://fedoraproject.org/wiki/Talk:QA/Fedora_14_test_days 15:12:23 next ... 15:12:25 #info jlaska to create a ticket about possible enhancement/new test case for ABI stability check 15:12:30 #link http://fedorahosted.org/autoqa/ticket/190 15:12:45 that's been queued up for the copious spare time we all have lurking just around the corner! 15:12:56 and last ... 15:12:59 #info jlaska to cleanup (or remove) the Critical Path Packages#Background section so that it provides _some_ value 15:13:15 I looked at this briefly, but don't have any new content to offer yet 15:13:28 this should only take a minute or two though ... will knock this out for next week 15:13:34 #action jlaska to cleanup (or remove) the Critical Path Packages#Background section so that it provides _some_ value 15:13:57 okay ... time for the main event 15:14:02 #topic Fedora 14 QA Recommendations 15:14:19 You're probably getting tire of me saying it by now P) 15:14:34 * Oxf13 peeks in 15:14:45 Oxf13: howdy 15:14:59 so I'm just about done with filing tickets for the recommendations wiki 15:15:06 #link https://fedoraproject.org/wiki/Fedora_13_QA_Retrospective#Recommendations 15:15:16 An idea: get various small technical and comunity colleges to help with testing 15:15:20 there are a few more that I've not yet filed ... but the bulk is complete 15:15:24 i guess we should make sure all the tickets are assigned appropriately? 15:15:43 adamw: yup, I'm going to take this off the list ... and the next step is we get to work the queue 15:15:51 j_dulaney: there's an open floor section at the end of the meeting, can you raise it there? thanks. just keeps things on track =) 15:16:06 #info Next step, start to work the ticket queue ... https://fedorahosted.org/fedora-qa/milestone/Fedora%2014 15:16:17 A little late, but due to some internal communication, I want to ask this group if they thought a Fedora Knowledge Base would be useful to your efforts? 15:16:26 as a recommendation for future Fedora releases. 15:16:40 * jlaska adds to open-floor ideas 15:17:23 last point here ... if there is something you were itching to work on for Fedora 14 QA ... and there isn't a ticket 15:17:48 not to worry :) Just file a ticket and start workin' 15:18:01 Oxf13: btw ... I think that was down as a wishlist topic ... checking ... 15:18:07 ok 15:18:38 Oxf13: same general idea ... "jlaska - Pony - some web tool to help collaborative problem debugging. Similar to http://answers.yahoo.com/ perhaps? " 15:18:51 k 15:18:52 definitely filed under my pony category :) 15:19:08 but "knowledge base" might be a more appropriate title 15:19:24 okay ... any other comments on the retrospective or recommendations? 15:19:36 I'm not planning on including this topic in future meetings 15:19:39 nup, just good job, thanks :) 15:20:12 thanks all for the feedback, I hope I captured a manageable set of tasks for us on F14 15:20:37 #topic Proventester update 15:20:53 adamw: I have this as a running meeting topic, but I don't know if there are any updates this week 15:21:10 nothing much has happened. really we're waiting on bodhi now. 15:21:29 the only thing i considered is adding some instructions for mentors as well as instructions for testers. 15:21:45 adamw: nice, so that means the proventester instructions dafrito and you worked on are complete? 15:21:46 It seems the Wiki entry could be cleaned up a bit 15:21:48 i suppose it's worth nothing that the eds debacle gives a very clear indication of the need for this testing =) 15:21:51 #link https://fedoraproject.org/wiki/Proven_tester 15:21:54 Or I should say, made clearer 15:21:59 jlaska: well, i can't see any way to improve them 15:22:10 j_dulaney: which wiki entry, and what do you suggest could be improved? 15:22:12 adamw: nice, I call that complete! :D 15:22:23 jlaska: either that or i suck, yep =) 15:22:27 adamw: preoventester 15:22:40 the page jlaska just linked to? 15:23:10 Indeed 15:23:36 what do you suggest changing? 15:24:34 I'm thinking a little more on the mentoring and approval process 15:25:27 j_dulaney: ah, I think something adamw mentioned he might like to see. We'd welcome any content to get things started 15:25:31 note that there is also the https://fedoraproject.org/wiki/QA/JoinProvenTesters page 15:25:45 we should probably clean up the naming of both pages to be more co-operative 15:25:52 Indeed. I seem to have missed that 15:26:02 Maybe combine them? 15:26:33 possibly 15:26:51 The joinproventesters seems rather short to have its own entry 15:26:57 we can basically choose to have one big 'everything about proven testers' page, or probably three separate pages for process, joining, mentoring 15:27:27 one page is a nice simple concept, i just would want to make sure it doesn't get too long and hard to follow 15:27:38 * adamw 's sworn enemy is the eye-glaze effect ;) 15:27:45 I know that enemy 15:27:51 otherwise known as tl;dr 15:28:00 That's why I tend to print :) 15:28:07 (took me years to figure out what that meant) 15:28:49 Read through long bits of code, much? 15:28:52 this is where you really wish for "collapsable" sections in mediawiki 15:29:28 collapse all the sections by default except for the common top section, then expand the sub-sections as desired, for each of joining / mentoring 15:29:39 yeah, that would be ideal 15:29:39 * ianweller peeks in 15:29:44 Oxf13: I think rhe found an option that might be available to anyone interested 15:29:53 jlaska: i thought that was tables? 15:29:53 in fact, I think ianweller helped with adding the collapse css 15:30:08 adamw: yeah, but I believe there is similar support for
's too 15:30:09 yeah i added css/javascript for collapsing tables 15:30:15 but I could definitely be *wrong* 15:30:18 rhe would know best 15:30:24 or of course our wiki master ... ianweller 15:30:39 j_dulaney: it's something i try to avoid ;) 15:30:56 but anyway ... adamw j_dulaney: anything you guys want to try on the wiki for next week? 15:31:12 jlaska: Sure 15:31:35 j_dulaney: if you could make a separate draft of a combined page in your personal space on the wiki and send it to the list for comments that'd be awesome 15:31:47 adamw: Indeed 15:31:48 i'd like to know what aaron thinks too since he did such an awesome job on the proventesters page 15:31:58 good point! 15:32:02 (aaron = dafrito, not sure if he's around) 15:32:35 grab and #action items you want ... otherwise, I'll keep this #topic on the agenda for our next meeting 15:33:31 #action j_dulaney to draft a combined proventesters / joinproventesters wiki page for list review 15:33:43 ianweller: http://en.wikipedia.org/wiki/Wikipedia:NavFrame 15:34:09 #grab draft 15:34:18 Got it, will do 15:34:40 jlaska: i saw that, but it's not really making collapsible sections automaticaly 15:34:44 i can't spell today 15:34:54 ianweller: yeah, you're right 15:35:15 ianweller: hey, at least you didn't say automagically 15:35:19 there's a word that makes me want to punch things 15:35:26 * jlaska notes :) 15:35:37 okay, thanks adamw and j_dulaney 15:35:40 (and of course dafrito) 15:35:56 next up ... 15:35:56 the dude that taught me to program did things magically 15:36:17 unless his name was otto, you're good 15:36:22 kparal: I just joined this other meeting, do you mind taking the #chair from here? 15:36:24 (thank you, thank you, i'm here all week) 15:36:30 tip your waiter 15:36:31 jlaska: sure 15:36:49 Wish I could afford to go somewhere with a waiter 15:36:54 * kparal looking for agenda 15:37:02 kparal: I asked wwoods to drive the autoqa update for today 15:37:05 hey, denny's has waiters. technically. 15:37:31 so, we can hop on the AutoQA, right? 15:37:40 kparal: yeah, that's next 15:37:46 #topic AutoQA update 15:37:54 wwoods: you have the mic! 15:38:08 HELLO FEDORA QA 15:38:11 ARE YOU READY TO ROOOOOOCK 15:38:16 ahem sorry. 15:38:20 *applaud 15:38:23 I've already got some Godsmack going 15:38:25 haha 15:38:28 WOOOOOOOOOOOOOOOOOOOOO 15:38:33 so last week we (finally) got the post-bodhi-update hook merged 15:38:34 gogogogo 15:38:42 * adamw jumps on someone's head 15:39:04 yay 15:39:05 welcome chenrano2002! 15:39:08 we're working on a new 'helloworld' test to make sure the bodhi-watcher and autoqa hook are working properly 15:39:15 yes, I'm here:) 15:39:35 in what language? 15:39:41 #info post-bodhi-update hook is merged 15:39:46 the target is, of course, the 'depcheck' test 15:39:58 which should check all new proposed updates for dependency breakage 15:40:02 #info 'helloworld' test is a new big task now ;) 15:40:15 and refuse to let rel-eng move packages that might break the repos 15:40:27 helloworld is really needed for newbie, I hope to see it 15:40:30 *fanfare* 15:40:32 but that's a complex test that will require help from infrastructure/releng to set up properly 15:41:02 Where can I go to read up on autoqa? 15:41:03 so the helloworld test (written in python, like all the autoqa stuff) is what we'll use to test out the new hook. 15:41:25 https://fedorahosted.org/autoqa/ is where the code lives 15:41:30 Roger 15:41:32 https://fedoraproject.org/wiki/AutoQA is where the wiki docs are 15:42:01 ok, thanks wwoods 15:42:13 anything else from your end? 15:42:16 wwoods: what can I help with to complete that helloworld? 15:42:28 I think you and kparal had some run_once() suggestions ... just combine those and test? 15:42:35 there's more autoqa stuff going on, but I think kparal & jskladan will have more to say about that 15:42:38 jlaska: I believe so 15:43:02 jlaska: yeah - go for the quick-n-dirty-and-simple test and just dump locals() 15:43:11 kparal: wwoods: (helloworld) ... okay thanks, I'll update the ticket 15:43:22 * Viking-Ice starts Moshing. . . . 15:43:33 in the next week or two we might have a discussion about standardizing hook argument names, or tweaking the convention for passing hook args 15:43:46 but don't let that stop you from getting helloworld working now. 15:43:50 #topic AutoQA update - resultdb 15:43:51 at first, helloworld will just work with post-bodhi-update? 15:43:56 er 15:44:06 #topic AutoQA update 15:44:07 kparal: sorry ... keep going, I'll follow-up with you guys on fedora-qa after 15:44:15 * kparal misclicked 15:44:50 #info standardization of hook argument names will be discussed soon 15:44:55 kaparal can't wait to share his idea:) 15:45:10 jlaska: I believe helloworld should work for any hook 15:45:27 it will be just a generic test that will print out incoming arguments 15:45:31 that's all 15:45:37 and reports success :) 15:45:42 kparal: right, just didn't know how far you guys wanted me to go with this 15:46:01 but I'll hold off the standardization of hook args for later 15:46:02 No pretty stuff on screen? 15:46:25 maybe it could draw a pony 15:46:32 YAY 15:46:33 alright, ready for jskladan's update? 15:46:53 #topic AutoQA update - resultdb 15:46:57 well, updates on the ResultsDB front - i have a testing instance up'n'running on one of our machines (kudos to jlaska & infrastracture team), i already rewrote initscripts & rpmlint tests (see jskladan branch in autoqa git) so they store the data inside. I'll finish the other tests during this week and will set up the testing autotest-server instance so it runs these tests :) 15:47:18 sounds nice! 15:47:29 #info resultdb running on our staging server 15:47:32 so stay tuned for the magical tool being filled with data :) 15:47:56 and once there is some reasonable specimen 15:48:06 we can start working on some frontends 15:48:23 is it magical and revolutionary? 15:48:39 adamw: you bet 15:48:59 Can it make a cheesesteak appear in front of me? 15:49:01 so that's about it on the resultsDB for now :) 15:49:08 ok, thanks jskladan 15:49:21 #topic AutoQA update - autotest labels 15:49:34 and now for something different 15:49:39 j_dulaney: only if you can pronounce the mighty words of power :)) 15:50:05 I have created 'labels' branch that contains patched autoqa that can use autotest labels properly 15:50:09 at least I hope so 15:50:14 jskladan: I know not this heavy wizardry, yet 15:50:15 announcement here 15:50:18 #link https://fedorahosted.org/pipermail/autoqa-devel/2010-June/000742.html 15:50:39 we should now be able to specify "this test must run on Fedora 12", even dynamically 15:50:44 hooray! 15:50:48 shiny! 15:51:00 that was a big missing piece for a while, this is totally awesome 15:51:18 and we're now discussing best names for our labels: 15:51:21 #link https://fedorahosted.org/pipermail/autoqa-devel/2010-June/000740.html 15:51:36 does it just look for an existing machine that matches the labels, or can we actually spin up VMs as needed? 15:52:04 wwoods: it doesn't check for anything, it just schedules the job and atest takes care of the rest 15:52:20 if there is no machine matching that label, I guess it stays scheduled indefinitely 15:52:43 there are certainly areas for improvement :) 15:52:55 I still have to test is properly 15:53:08 So, how would I go about getting my machines talking to it all? 15:53:48 #info autotest labels can now be used when patched autoqa from 'labels' branch is used 15:53:51 j_dulaney: at this point, you don't. Although you can certainly run the tests at home if you like 15:54:31 I'm also working on a script that will ease VM installation for us, that's a second thing. I want to use it for my 'labels' testing 15:54:37 alright, that's from me 15:54:39 and you can set up your own autoqa server also. since, y'know, we're showing it's so damn easy ;) 15:54:45 woods: right 15:55:02 #topic Open discussion 15:55:10 now type your ideas 15:55:27 we've had two mentioned earlier already 15:55:31 i believe j_dulaney was first 15:55:41 ok, what was the name of the idea? 15:55:46 * kparal scrolls back 15:56:21 j_dulaney: An idea: get various small technical and comunity colleges to help with testing 15:56:39 #topic Open discussion - colleges to help testing 15:56:41 I was thinking of getting IT type classes at said colleges to do some of the testing for us 15:56:50 we've actually had one case of that already 15:57:10 j_dulaney: so what do you suggest should be done and how can you help? :) 15:57:48 vedran miletic teaches classes at a university, and he integrates Bugzappers into that; we've helped him out with having his students do some triage work a couple of times 15:57:52 it is a neat system 15:57:55 is there a school you're in touch with where you could start such a program? 15:58:17 oh cool so there's some documented prior work on how to do this stuff? 15:58:46 at the local community college, I know the instructer that teaches the appropriate class 15:59:13 I can't think of what said class is called at the moment, massive brain fart here 15:59:41 Anyway, I'm sure he'd be happy to have his students try to break applications once in a while. 15:59:55 well for sure ask if the instructor's interested, and point to the QA/Join page for ideas - https://fedoraproject.org/wiki/QA/Join 16:00:04 Roger that 16:00:04 if they'd like to try any of that stuff we can certainly arrange to help them out 16:00:53 ok, if you have some contacts, try to lure them to us, sounds good :) 16:00:53 I could also do the same thing at the local high school when class starts back up this fall 16:01:04 yup 16:01:16 ok, thanks 16:01:20 what was the other idea? 16:01:53 #topic Open discussion 16:01:59 * jraber wonders if anyone has heard an update regarding bug 552423 It is a COMMONBUG, but the last time I checked it was the only COMMONBUG with no workaround or planned fix 16:02:19 .bug 552423 16:02:21 kparal: Bug 552423 [abrt] crash in gnome-panel (/usr/libexec/wnck-applet @ wncklet_connect_while_alive / wncklet.c:183) - https://bugzilla.redhat.com/show_bug.cgi?id=552423 16:02:37 kparal: it was from oxf13 regarding a knowledge base 16:03:27 #topic Open discussion - Knowledge base 16:04:05 personally i hate the damn things, like i hate any large aggregation of likely-to-change-rapidly information without good mechanisms for staying up to date... 16:04:12 ok, so tell me what's the difference between knowledge base and a wiki? :) 16:04:31 Microsoft's sucks 16:04:39 Indeed, it is often useless 16:04:54 a well organized wiki is more granular and hence tends to have a better 'ownership' stake 16:05:10 newers need docs, so is it possible to convert/public useful wiki to PDF format for reading locally? 16:05:23 in a good wiki, some person/people or group is clearly the owner of a given page and should update it 16:05:23 what would be the benefit of a knowledge base for us then? 16:05:30 yeah, i'm struggling 16:05:33 Oxf13: is there any more detail on the idea? 16:05:48 what's the proposed format/content of the knowledge base? how is it differentiated from the wiki? how would it be written/maintained? 16:06:15 thanks for formulating those questions, adamw 16:06:51 * jlaska back 16:07:15 * kparal gives jlaska the throne back 16:07:24 I myself rather enjoy the idea of PDFs for basic stuff, but then one can just save the Wiki page 16:07:25 kparal: no sir ... take us to the end :) 16:07:43 don't want to distrupt the flow as much as I already have 16:08:09 we're just struggling with finding differences between knowledge base and wiki 16:08:33 I don't know if there was anything actionable with this topic 16:08:36 alright, seems Oxf13 is away, we can discuss it next time 16:08:49 #topic Open discussion 16:08:49 sure 16:09:07 anything else anyone? 16:09:09 sorry I was on another call. 16:09:12 ah 16:09:51 adamw: basically the idea of having a Fedora KB has been kicked around by a couple people, and before we put any resources into it, we have to make sure there will be somebody to consume/contribute to it, so that it'll be useful use of our resources. 16:10:06 #topic Open discussion - Knowledge base 16:10:12 are there answers to the above question? 16:10:16 questions* 16:10:47 adamw: nothing here ... let's come back to this when there is something more tangible 16:11:04 I think it could be a compelling discussion ... if there is a proposal or existing solution etc.. ?? 16:11:04 #topic Open discussion 16:11:13 my $0.02 at least 16:11:22 regarding bug 552423, if nobody has an update, should I ping the maintainer? (should I save this for the bugzappers meeting tomorrow?) 16:11:32 if the idea is just 'something called a knowledge base' then i'd say, no, on the surface it seems to be just a duplication of the wiki and i can't see any added value. but we need more detail to provide sensible feedback. 16:11:39 jraber: please raise this question in #fedora-qa channel 16:11:45 adamw: yup 16:12:27 alright, let's wait a minute and then close this meeting if noone objects 16:13:12 I can think of nothing else 16:13:18 thanks all for attending 16:13:22 #endmeeting