16:00:41 #startmeeting Fedora QA Meeting 16:01:01 #topic gathering able minded persons 16:01:26 * Oxf13 16:01:35 * dpravec :) 16:01:36 Oxf13: howdy 16:01:44 * jlaska tips hat to dpravec 16:01:47 * stickster lurks 16:01:48 * skvidal wonders what this crazy QA is all about 16:01:48 * kparal is here 16:01:57 skvidal: it's all about you baby! 16:02:03 my baby? hell no 16:02:05 stickster: kparal: greetings :) 16:02:09 skvidal: hehe 16:02:11 * skvidal is safely childless 16:02:34 skvidal: speaking off ... err, well unrelated to, I need to ping you for some Durham guidance 16:02:35 * Viking-Ice verifies that it's the actual skvidal not one of the clones.. 16:02:46 Viking-Ice: I'd love to see that test case! 16:02:50 Viking-Ice: how can you tell? is there a bar code? 16:02:54 jlaska: durham is the bom-diggity - everyone loves it 16:03:18 or do you have to get all harrison ford? :) 16:03:29 jlaska: we tag all our clones.. 16:05:05 soooo... 16:05:08 *twiddles thumbs* 16:05:10 wwoods: is lurking as well 16:05:21 adamw: that's it ... just meeting to meet :) 16:05:26 nahh ... jk 16:05:30 alright, i'm off! 16:05:30 alright ... diving in 16:05:34 heh 16:05:34 * adamw has left #fedora-meeting 16:05:52 ...aaahhhhh, got cha 16:05:52 quick give him the action items! 16:06:01 #topic previous meeting follow-up 16:06:07 * wwoods lurk 16:06:13 adamw: not us which have the messages localized usually :) 16:06:28 I'm walking through the proposed agenda - https://www.redhat.com/archives/fedora-test-list/2009-August/msg00419.html 16:06:46 and in an effort to speed up our first topic ... we've assigned no action items from last week 16:07:07 we've got a few updates of course, but we can touch on those in the agenda and 'open discussion' 16:07:33 let's dive into the Alpha (or come up for air) 16:07:38 #topic F-12-Alpha update 16:08:12 Oxf13 composed a F-12-Alpha-RC1 this weekend 16:08:28 Liam and Rui are in training this week, so I'll be kicking off an updated test matrix shortly ... https://fedoraproject.org/wiki/QA:Fedora_12_Alpha_RC1_Install_Test_Results 16:08:40 yeah, it's nice and blank right now 16:08:47 you betcha 16:08:51 gotta start somewhere right 16:08:57 would be easier if it started full though :) 16:09:31 Oxf13: Liam was watching that ticket for updates on the alpha compose, so we might need to work out best practices for communicating drops 16:09:44 probably something we can tackle for Beta 16:09:46 how's the blocker list looking? are we pretty clear now? 16:10:06 #topic F-12-Alpha readiness 16:10:15 adamw: looking good ... 2 MODIFIED bugs right now 16:10:27 bug#517475 16:10:28 Bug https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=517475 medium, low, ---, pmatilai, MODIFIED, geode not i686 arch in rpmrc.in 16:10:40 Oxf13: there's some feedback in that bz that might make more sense to you 16:11:04 Oxf13: I haven't heard back from Warren on helping test that issue, however, the reporter seems to be very active ... which is a 'good thing' [tm] 16:11:08 what it means to me is that we'd have to change more than just that little rpm bit to get this working for geodes, and that's change I can't stand behind for ALpha 16:11:40 this was a nice to have for me, so I'm on board with that 16:12:01 the other MODIFIED issue ... bug#516941 16:12:02 Bug https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=516941 medium, low, ---, kernel-maint, MODIFIED, kms broken and can cause oops without git3 upstream PAT patches 16:12:03 note that the RC does not have this test rpmbuild with it 16:12:16 that one just needs somebody with the right hardware to verify 16:12:38 although I think Clyde did that 16:12:40 Clyde chymned in on that bug 16:12:54 I'm actually having problems booting rawhide today with RADEON 16:13:03 but I'll follow-up in the bz 16:13:18 oh... I just closed the bz... 16:13:23 well, I suppose you can re-open 16:14:21 yeah, I'll follow-up on that ... I had to boot with 'nomodeset radeon.modeset=0' to login today 16:14:45 so ... quick round the bases from folks in the meeting on how we stand 16:14:52 #topic F-12-Alpha readiness - installer 16:15:17 as of Friday (anaconda-12.14) we have good results in thanks to Liam and Rui - https://fedoraproject.org/wiki/QA:Fedora_12_Alpha_TCRegression_Install_Test_Results 16:15:28 will retest now with anaconda-12.15 - https://fedoraproject.org/wiki/QA:Fedora_12_Alpha_RC1_Install_Test_Results 16:15:43 folks always welcome to participate 16:15:55 I'm pulling down some bits to test. 16:15:57 there are a few new anaconda bugs and topics on fedora-test-list that need some resolution 16:16:03 (installer related) 16:16:05 I'm rather frightened of the liveCD state so I'm pulling those first 16:16:47 Based on the information from Thu/Fri test run ... we're in decent shape 16:16:54 should know more once we have more data in of course 16:17:05 any other comments on the installer readiness? 16:17:58 kparal: you recently bit the bullet and installed rawhide to laptop, any issues? 16:18:26 jlaska: it went surprisingly ok :) 16:18:28 we'll have to give the go/no-go pretty quickly 16:18:42 yeah 16:19:08 #topic F-12-Alpha readiness - xorg-x11 16:19:11 like, today I guess 16:19:27 adamw: any updates on X-related issues you've been tracking? 16:19:49 jlaska: i also installed rawhide again recently and it was much better than few weeks ago 16:19:49 well, there's some icky bits, but i'm not aware of anything show-stopping 16:19:56 i have one issue related to multi-monitor setup which i had no time to report yet. i don't know if that relates to xorg or not 16:19:58 dpravec: awesome! 16:20:04 f13 made a call not to include the x.org for fixing the display blanking bug, which is fine, we can fix that later 16:20:06 still kde is having big problems 16:20:27 kparal: perhaps more a fit'n'finish issue ... https://fedoraproject.org/wiki/Test_Day:2009-07-07_Fit_and_Finish:Display_Configuration 16:20:45 adamw: is there a list of known gotchas for KMS (or X) related stuff on the radar for F-12-Beta already? 16:20:46 there's a couple of nouveau issues - i caught someone on -devel-list or -test-list who's seeing a bug I saw with the version in rawhide, text changing colours - but that's not critical enough to break freeze for 16:20:53 I gather the blanking issue is near the top 16:21:13 and on my system it's got two icky failure modes which i need to talk to ben about, but aren't concrete enough to block anything 16:21:21 jlaska: does it make sense to add bug reports to old test days wiki pages? i don't know if anyone will really notice 16:21:25 yes, the blanking issue is definitely being worked on and fixed but just won't make the alpha images 16:21:40 kparal: it's a good thing to do, people do look at old pages (people like me :>) 16:21:46 ok 16:22:28 jlaska: known gotchas we will put on the Common Bugs page, i don't have a list myself at present, i don't know of any issues general enough to go on such a list, except maybe the nouveau text color change thingy 16:22:31 adamw: okay, I may need your expertise on this radeon issue I'm having w/ rawhide ... but I'll be sure to consult the Xorg/Debugging page first (https://fedoraproject.org/wiki/Xorg/Debugging) 16:22:53 yeah, first port of call would be the X and kernel logs after a failure to start X 16:22:59 right on 16:23:06 darn sailors :) 16:23:32 i think in general X is pretty solid right now, especially for an alpha 16:23:50 okay ... those are the 2 big component areas I think we're plugged in on 16:24:01 I have kernel and desktop-apps as general items 16:24:12 if anyone has thoughts or concerns around those component areas? 16:24:12 my feeling just from running rawhide is we're pretty solid for alpha, fwiw 16:24:50 none of the big-hitter applications are broken, we've got annoying things like the notification area bug fixed, it's looking pretty decent to me 16:25:21 #topic F-12-Alpha readiness - $other 16:26:01 we've got some test day coverage on live images recently (including installation) ... so that's comforting 16:26:07 adamw: yeah, I'm having that sense as well 16:26:23 * nirik is doing nightly composes now... feel free to test them and provide feedback. 16:26:42 I am running into some compose issues, but there should be images there still. 16:26:47 nirik: that reminds me ... how different (package-set-wise) are those from official rel-eng live images? 16:27:00 none. They are made from spins git head. 16:27:06 awesome 16:27:07 they should be the same thing. 16:27:21 nirik: let's give an official shout out to nirik for providing daily rawhide live images :) 16:27:29 >shout< 16:27:34 * kparal shouts 16:27:42 shout! 16:27:47 * dpravec claps 16:28:02 ha. ;) not if I can just get a newer kernel on the compose machine so I can avoid e2resize errors, everything should be golden. ;) 16:28:09 s/not/now/ 16:28:34 alrighty ... let's change gears over to what's on the horizon ... 16:28:40 #topic AutoQA update from wwoods 16:28:57 wwoods: last week was another busy week it seems 16:29:28 * tk009 comes in really late 16:29:33 heya 16:29:43 tk009: fashionably late! 16:29:45 yeah - the short summary is: I set up a way to pass the systemwide config to the tests 16:29:59 put some email config stuff in the system autoqa.conf 16:30:03 and taught all the tests to send email 16:30:17 so now they're all sending email to the autoqa-results list 16:30:25 https://fedorahosted.org/pipermail/autoqa-results/2009-August/thread.html 16:30:33 wwoods: I noticed those emails :) 16:30:38 holy conflicts batman 16:30:45 updated the test-writing notes a bit: https://fedoraproject.org/wiki/Talk:Autotest 16:31:09 fixed the watcher scripts and got them running on a cron job, so the tests run automatically 16:31:22 oh and I fixed the 'conflicts' test 16:31:33 what was the issue there? 16:31:39 so now we're getting emails to the autoqa-results list every time the repos update or there's new rawhide images 16:31:48 jlaska: the code didn't exist. heh! 16:31:54 had to ask around to find a copy of it 16:31:55 wwoods: doh! 16:32:10 this is great ... https://fedorahosted.org/pipermail/autoqa-results/2009-August/000155.html 16:32:13 wwoods: personally there is some interesting information on the Talk page but it never occurred to me to look for it there. maybe some more visible place would be better 16:32:33 and check it into the autoqa repo and teach the test wrapper to save the list of conflicts 16:32:47 kparal: I'll move it over when it's ready 16:32:55 that's why it says "to be integrated into main page" 16:33:09 ok :) 16:33:20 I'm going to work on a first draft of the "how to write a test" document this week 16:33:40 also gotta fix the rats_install test and make rats_sanity stop running on non-rawhide repos 16:34:03 the documentation would be great also from another reason. maybe i could write about it now, james? 16:34:16 oh, and I summarized the current progress for fedora planet: http://qa-rockstar.livejournal.com/8215.html 16:34:42 if you've got tests you want to write, I'd love to hear about it 16:34:48 kparal: sure what's up? 16:34:50 wwoods: there are other people who would love to plugin theirs efforts into autoqa 16:34:51 which might be what kparal is hinting at? 16:34:58 wwoods: nice post! 16:34:59 we need more documentation 16:35:06 dpravec: definitely - trying to wrote docs as fast as I figure these things out 16:35:13 if we can help somehow, wwoods, please ask us 16:35:19 we will do our best to help you 16:35:25 well there appeared a new effort about providing a test package sanity tool for fedora 16:35:29 dpravec: excellent, thanks very much 16:35:53 currently the effort is lead by Petr Splichal (nick psss on #fedora-qa) 16:36:19 there is a short wiki page for those in redhat: https://wiki.test.redhat.com/BaseOs/Projects/FedoraTps 16:36:20 doh! he just left ... kparal we could perhaps invite Petr to our next meeting if the time is right 16:36:45 jlaska: we will ask him to come, right 16:36:50 wwoods: you could be interested in that because petr expressed desire to collaborate, because his task is a subset of our task 16:36:56 kparal: looks like an internal wiki page ... we might want to get him to post his ideas to fedoraproject wiki? 16:37:38 jlaska: currently the page is only for tracking existing project - which caused that he found autoqa and came to us to ask about it 16:37:45 *projects 16:37:57 jlaska: sure, we will talk to him. 16:38:29 dpravec: kparal: thanks gents 16:38:40 therefore hopefully petr splichal could help us write some of the test cases for autoqa 16:38:46 one of the first tasks to get this in autoqa would be to talk to the bodhi/koji maintainers to figure out a good way to get notification of new package builds/update creation 16:38:57 because autoqa does not have a post-koji-build or post-bodhi-update hook implemented yet 16:39:25 you could also run package-level tests on each new package when the *repo* updates 16:39:36 i believe he can help about it too, because petr is in BaseOS QE team 16:39:59 wwoods: you got links to the 2 watcher scripts now? 16:40:28 wwoods: pretty sure that's going to require the message bus in the long run. 16:40:51 wwoods: in the short run, we can create an email address that all build notifications go do, and that email address has procmail to poke autoqa 16:40:53 wwoods: the best way to get notification is to poll koji for new builds 16:41:28 jlaska: heh, actually no - the git trac plugin only shows the master branch 16:41:31 (email/procmail has the bonus that when the bus comes up, we just replace the email/procmail, but the poking of autoqa remains the same) 16:41:35 and I haven't merged the autotest rewrite yet. doh! 16:42:01 oh they hide it ... 16:42:02 http://git.fedorahosted.org/git/?p=autoqa.git;a=tree;f=hooks;h=317a0d2ea7f670ea6445f9b530e3a9a746e1a44a;hb=refs/heads/wwoods-autotest 16:42:24 http://git.fedorahosted.org/git/?p=autoqa.git;a=tree;f=hooks;hb=refs/heads/wwoods-autotest works too 16:42:30 yeah, that's clearner 16:42:33 cleaner too 16:43:15 anyway I don't want to get too deep into implementation details here 16:43:23 right right, sorry 16:43:42 but yeah, TPS for fedora is definitely a great idea and we should sketch out some goals 16:43:58 maybe set up a milestone in the autoqa trac and file some tickets with specific work items 16:44:03 wwoods: i have told petr to contact you so you can talk about it 16:44:09 kparal: awesome, thanks 16:44:25 wwoods: should i put it in trac? 16:45:15 dpravec: mm, not yet - let's take some time to decide what exactly we want to accomplish 16:45:28 ok 16:46:08 that'll be a topic to track for future meetings ... but we will revisit 16:46:46 wwoods: thanks for the updates, looks like great progress! 16:47:00 I'll see what petr says and we'll try to get some specific goals written down 16:47:11 and then we'll work on a milestone/tickets and stuff 16:47:33 I know that going through all the documentation/planning, planning, planning/ticket writing is a bit.. unsexy 16:47:55 but it's reeeally helpful for longish-term stuff 16:47:58 yes 16:48:00 agreed 16:48:29 alright, changing topics ... 16:48:44 #topic Fedora Test Day updates 16:48:56 just a quick run through of previous and upcoming test days 16:49:18 since both the main track and the fit'n'finish track are rockin' and rollin' 16:49:35 thanks everyone for pitching in on last weeks test days 16:49:41 * Fit'n'Finish - Peripherals 16:50:02 I didn't see a summary from mclasen on that yet, so I'm not sure how successful that event was for him 16:50:13 for the logs 16:50:13 http://fedoraproject.org/wiki/Test_Day:2009-08-04_Fit_and_Finish:Peripherals 16:50:26 i did run my script on the peripherals test day page and send the result to mclasen with a suggestion he do a report 16:50:28 didn't hear back yet 16:50:38 adamw: great, thanks for being proactive there 16:50:50 adamw: I think you should package that script :) 16:50:55 speaking of ... 16:51:09 * 2009-08-13 - NetworkManager - see the recap adamw sent - https://www.redhat.com/archives/fedora-test-list/2009-August/msg00377.html 16:51:21 it's too hideous to package :) 16:51:26 haha ... never! 16:51:26 :D 16:51:27 adamw: i have seen the script on the mailing list i think. is it also somewhere on the wiki? 16:51:48 anyway, for the record, it shows 10 NEW, 2 ASSIGNED, one CLOSED for the peripherals day 16:51:55 kparal: not yet, but i will add it to the SOP page 16:52:01 great 16:52:16 Heads up gang ... 2 upcoming Test days for this week 16:52:32 #topic Fedora Test Day - 2009-08-18 - Printing 16:52:36 https://www.redhat.com/archives/fedora-test-list/2009-August/msg00368.html 16:52:44 See mclasen's announcement ^^^ 16:52:57 that's the fit 'n' finish day for the week 16:53:02 you got it 16:53:23 dpravec and kparal are running the show this week for the main track test day 16:53:29 #topic Fedora Test Day - 2009-08-20 - ABRT 16:53:41 when will the liveimages for fit and finish appear? 16:53:47 are you guys on top of everything? shout if you need any help with test cases, wiki page etc 16:54:11 http://fedoraproject.org/wiki/Test_Day:2009-08-20 16:54:18 adamw: i'm just finishing building the livecd image. that should be prepared soon 16:54:20 poelcat: I'll check-in w/ mclasen on that 16:54:27 poelcat: perhaps wecan use the F-12-Alpha-RC1 images 16:54:31 we need to put newer abrt packages to our live cd images 16:54:36 #action jlaska to ping mclasen on which live images to use for Printing test day 16:54:56 and i think we will upload them tomorrow before americans will get to work 16:55:21 abrt team will hopefully push new abrt to rawhide soon enough 16:55:26 (tomorrow) 16:55:35 * nirik notes if there are no special needs, you're welcome/encouraged to use the nightly live spin composes. 16:55:41 our wiki page is still undergoing some cleanup and refactoring 16:55:56 nirik: we have special needs, for example reboot after installing abrt 16:56:03 which is hard to achieve using live cd :) 16:56:17 dpravec: would a live image w/ persistant storage help? 16:56:21 yes, I know. I was mentioning it for other test day cases. 16:56:30 many of them shouldn't need special packages or the like. 16:56:34 jlaska: we are also including some more stuff for testers 16:56:59 we should add a programm which will not be from fedora (any tips?) 16:57:12 dpravec: what is the program? 16:57:14 also i'm putting some extra icons on the desktop. i will update the "howto create a test day livecd" guide on the wiki so it can be used also in the future 16:57:17 to test if checking signed packages will work 16:57:24 jlaska: whatever 16:57:40 a one that can be crashed using kill -11 16:57:42 :) 16:58:06 ie, compiled C app or python application 16:58:16 dpravec: I think it's okay to post some test code on your fedorapeople page and link to it any test cases that require it 16:58:39 alternatively, if the manual steps aren't too difficult ... that's fine too 16:58:44 it can be just hello world and sleep 200 seconds 16:59:19 back to important stuff: 16:59:30 we will need you on the test day 16:59:50 please come to test it 17:00:24 dpravec: thanks! Looking forward to another focused event on ABRT (formerly known as crash-catcher) 17:01:00 * jeff_hann joins in late, sorry 17:01:13 #topic Fedora test day - 2009-08-27 - Dracut 17:01:17 jeff_hann: welcome :) 17:01:35 jlaska: thanks and hello folks 17:01:46 Viking-Ice: did you want to take the lead on organizing QA for the dracut event? 17:02:36 or are you looking for someone to drive this event? 17:03:41 I'll take an action to circle back with you on that 17:03:57 #action who has the ball for dracut test day? 17:04:03 alrighty ... open mic time ... 17:04:07 #topic 17:04:32 any issues, concerns or ideas? 17:04:43 yes, i have one 17:04:54 dpravec: go for it, what's up? 17:04:56 http://www.vncrobot.com/ got recently opened using gpl2 license 17:05:05 what's the status of gnome-power-manager? 17:05:07 under name of t-plan robot 17:05:17 and it can be used for automated testing over VNC 17:05:22 (written in java) 17:05:24 jeff_hann: thx, we'll touch on your topic right after dpravec 17:05:35 #topic Open discussion - vncrobot 17:05:48 that means that we can use it to test virtual machines or even just gui applications 17:06:13 i am still not sure about this one, but this might be interesting 17:06:32 sounds like something wwoods might like... 17:06:41 (i would love to have pythonic one...) 17:07:00 the kvm guys do something now to automation manual graphical installs in KVM 17:07:10 I'm not clear what they use, but it's probably not this 17:07:28 i wanted to make sure something important will not slip away unnoticed 17:07:42 i think testing over vnc has its big advantages 17:07:54 imagine testing bootsplash 17:07:58 or root login in gdm 17:08:02 or xdm 17:08:05 right on 17:08:33 seems like something to evaluate should any future testing demand graphical verification 17:08:49 yes thats what i am talkign about 17:09:33 people in rh are often using dogtail, but that could be hard to use in some scenarios 17:09:57 (dogtail should work in gnome and kde well) 17:10:09 wwoods: what do you think? 17:10:26 I don't have objections to specific tools/technology ... I'd just want to understand what the high-level goal if we were to focus time+energy on it 17:10:42 jlaska: i absolutely agree 17:10:50 for now i am just mapping the space 17:11:04 you probably know more about dogtail than I do! :) 17:11:13 seems like it could be really useful once we have some idea how we'd set up for a GUI functional test 17:11:24 heh, i asked one developer to show me how he is using it :) 17:12:24 dpravec: thanks for the link ... if anyone has experience with this ... please do share your ideas w/ dpravec 17:12:41 #topic Open discussion - status of gnome-power-manager 17:12:48 ok... 17:12:57 jeff_hann: you asked about g-p-m, was there something specific you wanted to discuss? 17:13:13 I found the same bug I reported a while ago still present on my lappy 17:13:21 which bug is this? 17:13:28 HP Pavilion dv9750ed 17:13:49 which bug, not which laptop :) 17:13:49 it fails to acknowledge the fact that the power is unplugged 17:14:04 well, ok, and did you file a report on it? 17:14:11 only after I plug it back in and unplug again it works 17:14:33 well, not yet, wanted to check if anyone else klnows something about it 17:14:38 *knows 17:14:52 I don't see that issue reported on the previous fit'n'finish battery event - https://fedoraproject.org/wiki/Test_Day:2009-07-21_Fit_and_Finish:Batteries_and_Suspend 17:14:54 it's just a bug. systems have quirks in how they notify of ac status changes, it's not unusual. 17:14:56 jeff_hann: bug number? or bug url? 17:15:18 there was a similar bug on my laptop, it got fixed. if it's happening for you in current rawhide, just file a bug 17:15:18 one moment please 17:15:33 no, F11 17:15:38 with updates-testing 17:16:39 oh, yeah, sorry, f11 :) 17:17:05 is there a better answer than ... can you try a live image of rawhide? 17:17:18 * jlaska thinks that might be the best option? 17:17:40 no, not really, bugs like this should be fixed in stable releases 17:17:47 no reason to make people try rawhide 17:17:52 just file a bug with appropriate information 17:17:53 #512964 is related, still searching... 17:18:09 adamw: okay 17:18:15 I will have rawhide also installed on that machine, on a external HD 17:18:18 most importantly you need to kill devkit-power-daemon then run it from a console and paste the output you get when reproducing the problem 17:18:39 * jlaska checks https://fedoraproject.org/wiki/How_to_debug_DeviceKit_problems 17:18:59 #512995 17:19:02 doesn't have stuff for power atm :/ 17:19:42 there looks like more to the story than that. 17:19:53 otherwise it'd be broken for _all_ systems presently, and it isn't. 17:20:20 either that bug only affects certain systems, or it's only in rawhide. 17:20:26 so please just file a new one, on f11. 17:20:41 oh, and when the battery got empty, my lappy got shutdown in a second, no shutdown procedure 17:20:53 ok, I will do that 17:21:02 maybe it was poweroff, not shutdown? 17:21:14 gotta check 17:21:35 jeff_hann: thanks for raising the topic ... you can reply to the meeting minutes with the new bz (or an existing one that you find a match on) 17:21:36 also no lower brightness or anything else 17:21:54 alright folks, let's close it out for today 17:22:07 unfortunately, I have to run now, since I am late somewhere; I will report the bug ASAP 17:22:11 we can continue on #fedora-qa or as a follow-up to the minutes I'll send 17:22:13 jeff_hann: thanks! 17:22:16 jeff_hann: this works for me (brigtness, turning off) 17:22:26 it depends on hw in many cases 17:22:28 * adamw has to go for a reboot, nouveau has nerfed up again 17:22:28 well, we'll see 17:22:31 thanks for your time folks ... happy testing! 17:22:34 #endmeeting