From Fedora Project Wiki
Attendees
- Will Woods (wwoods)
- Adam Williamson (adamw)
- James Laska (jlaska)
- Jesse Keating (f13)
Previous meeting follow-up
- [jlaska+adamw] mediawiki semantic update (packaging and hosting)
- REVIEWED - 490001 - Review Request: mediawiki-semantic - The semantic extension to mediawiki
- UNDER REVIEW - 490171 - Review Request: mediawiki-semantic-forms - An extension to MediaWiki that adds support for web-based forms
- User:tibbs posted some additional concerns for mediawiki-semantic-forms around licensing which I haven't followed up on yet
- request infrastructure hosting
- [f13] - autoqa: work on monitor for post-tree-compose
- We can mark this as "done".
- we have a monitor script that watches for new rawhide trees and calls autoqa accordingly
- it's a bit hardcoded for a path, but that's fine for now
- [wwoods] - autoqa: work on better (clearer) reporting from existing tests
- at the moment we're just working on getting the reporting to actually happen
- but yeah, the emails from e.g. repoclosure are kind of verbose
- [adamw] - will file a F10 bug for the syslinux issue, initiate discussion with pjones or jeremy for building a updated F10 package
- bug was filed, marked as a dupe, we added an entry to the release notes
- i think last time i saw the bug a workaround was being worked on...
- https://bugzilla.redhat.com/show_bug.cgi?id=492370
- [jlaska] - check-in with Warren Togami on the rawhidewatch blog ... can Fedora QA help ... what is his vision?
- I spoke with warren this morning about http://rawhidewatch.wordpress.com and what his idea/plan is around the blog. The summary is Warren intends for the blog to be a low traffic blog that includes imformation on non-obvious (potentially high-impact) problems affecting rawhide users. Warren indicated he will act as editory of the blog and accept story submissions
- adamw: i sent him a suggestion for it (to add a note on the nss issue that caught out some x86-64 users and broke rpm); he basically told me 'patches welcome', i.e. write an entry and he'll post it :) i was going to do that but haven't had time yet
- [jlaska+wwoods] - discuss integration between lab-in-a-box and autoqa
- I failed on this one, wwoods and I spoke about it briefly, but I have not scheduled time to chat yet. I'm still trying to figure out how best to plug the 'lab-in-a-box' into some framework. I don't have a good sense yet whether that's autoqa, something beaker-like, ... or something new that uses f13's watch script. basically looking for a way to automate rawhide installations ... and post the results (all in a virt guest). The missing pieces for me are the scheduler (something that tells it when to start), and results reporting (something that gathers the existing results and does _something_)
- if folks have suggestions, please do drop me a line
Autoqa update
- jkeating completed the rawhide watch script that launches autoqa when rawhide is updated. jkeating plans to hand off work to the QA team and focus on a message bug implementation.
- Next steps:
- continue improving existing test reporting
- interim goal of sending automated test result mails to autoqa-results
- wwoods suggested there are fixes that he would like to get into upstream beaker
- reach out to pmuller for information on packaging rhtslib
F-11 Status
User:jlaska requested some discussion around the schedule and the current planned QA activites for F-11
User:jkeating reminded the Snapshot#1 is scheduled for Friday (see http://poelstra.fedorapeople.org/schedules/f-11/f-11-releng-tasks.html)
- f13 reported that i386 image creation is broken which may impact snap#1 (see https://www.redhat.com/archives/fedora-test-list/2009-April/msg00535.html). He is working with jakub to get a build of glibc that may fix this
User:jlaska asked ... we have 1 month remaining before the final F-11 compose. What can we be doing better between now and that date?
- There are several Test Days planned (including anaconda storage testing part#2) ... see QA/Test_Days/F11.
- BugZapper triage events every Tuesday
- Adamw indicated two new members joined the BugZapper group this week
- f13 recommended we provide a new test results page earlier (e.g. QA:Fedora_11_Beta_Install_Test_Results)
- jlaska will create a new page for use with next weeks Anaconda Storage Test Day
- wwoods expressed concerns around pulseaudio bugs, notably rhbug:472339. lennart is aware of the problems and assisting with debugging. They are seeking assistance from jaroslav (alsa).
- several others indicated they are having sound (or no sound) problems
- There are also reports of DRI memory leaks - http://www.mail-archive.com/xorg@lists.freedesktop.org/msg06116.html
- f13 recommended scheduling several F11Blocker bug review meetings
- jlaska will get these scheduled
- adamw can solicit QA and BugZapper assistance during these meetings
- f13 also recommended someone look at Live image creation+installs
Open discussion
Fedora Board Meeting Update
User:jkeating asked how useful was the board meeting yesterday?
- f13 noted I think the board was was happy to hear about the status, but I was more concerned with the community at large
- jlaska noted it was nice to hear that some of the community events (test+triage days) are being received well
- If follow-up is requested or additional information needed, please pass along to User:jlaska
Test Day Live Image Kickstart
User:jlaska referenced recent fedora-test-list discussion around improving the Test Day live image experience. He has been looking at making a simple live image kickstart to place into the spin-kickstarts package so that others can also build and host Test Day live images.
- wwoods suggested included debugging tools (gdb/strace/ltrace) in the live image
- adamw advised keeping it simple ... make sure it's not something that starts eating up too much time
- f13 recommended a kickstart structure like: livecd-base -> livecd-testday-base -> <package set you care about>
Upcoming QA events
- 2009-04-09 - UEFI
- 2009-04-14 - BugZappers/Triage_days
- 2009-04-14 - Anaconda Storage Rewrite part#2
- 2009-04-16 - Yum presto
Action items
- [jlaska] - reach out to pmuller on packaging rhtslib
- [adamw] - send details of nss rawhide issue to warren for posting to rawhidewatch.wordpress.com
- [jlaska] - talk to poelcat about adding a few F11Blocker review meetings to the schedule
- requested a review before and after final development freeze
- [adamw] - review Test Day X11 bugs to ensure they are represented on F11Blocker
Next QA meeting
The next meeting will be held on 2009-04-15 16:00 UTC
IRC Transcript
| --- | jlaska has changed the topic to: Fedora Quality Assurance meeting | init | Apr 08 12:01 | |
| jlaska | | adamw: wwoods: f13: viking_ice (whom I believe is relaxing at the moment): you folks around? | Apr 08 12:02 |
|---|---|---|
| * | jlaska organizing notes from last meeting | Apr 08 12:02 | |
| f13 | | oh is that now? | Apr 08 12:02 |
| adamw | | here, sir | Apr 08 12:02 |
| f13 | | oh hrm, I think I'm missing a lunch and learn aren't I? | Apr 08 12:02 |
| adamw | | [adamw@adam ~]$ date -u | Apr 08 12:03 |
| adamw | | Wed Apr 8 16:02:58 UTC 2009 | Apr 08 12:03 |
| adamw | | looks like time to me! :) | Apr 08 12:03 |
| * | jlaska tracking notes at QA/Meetings/20090408 | Apr 08 12:03 | |
| jlaska | | f13: do you have a conflict? | Apr 08 12:04 |
| f13 | | only in my head | Apr 08 12:04 |
| f13 | | the L&L is tomorrow | Apr 08 12:04 |
| jlaska | | I think wwoods must be rebooting | Apr 08 12:05 |
| adamw | | what, did you crash him *again*? :) | Apr 08 12:05 |
| jlaska | | okay, he should be joining shortly ... let's go ahead and get started and we can move the autoqa stuff till he returns | Apr 08 12:06 |
| jlaska | | plus f13 is around for the autoqa update as well | Apr 08 12:06 |
| jlaska | | I have f13, adamw around ... show of hands for anyone else? | Apr 08 12:07 |
| jlaska | | okay, then it'll be quicker :) | Apr 08 12:08 |
| --- | jlaska has changed the topic to: Fedora Quality Assurance meeting | previous meeting follow-up (QA/Meetings/20090401#Action_items) | Apr 08 12:08 | |
| jlaska | | first I have is ... | Apr 08 12:09 |
| jlaska | | * [wwoods] - autoqa: work on better (clearer) reporting from existing tests | Apr 08 12:09 |
| jlaska | | but we'll circle back on that when wwoods returns | Apr 08 12:09 |
| jlaska | | next up ... | Apr 08 12:09 |
| jlaska | | * [f13] - autoqa: work on monitor for post-tree-compose | Apr 08 12:09 |
| fhornain | | Belgian Guys ? | Apr 08 12:09 |
| f13 | | yay! | Apr 08 12:09 |
| f13 | | We can mark this as "done". | Apr 08 12:09 |
| jlaska | | f13: it seemed that you were very active with mails+commits this past week! | Apr 08 12:09 |
| f13 | | we have a monitor script that watches for new rawhide trees and calls autoqa accordingly | Apr 08 12:09 |
| f13 | | it's a bit hardcoded for a path, but that's fine for now | Apr 08 12:10 |
| jlaska | | sweet, is that something others can play with from the git repo now? | Apr 08 12:10 |
| f13 | | sortof | Apr 08 12:10 |
| f13 | | they'd have to adjust the path used or make symlinks so that the paths match | Apr 08 12:11 |
| f13 | | basically I wanted to get my QA deliverables done so that I can move on to work on some other things | Apr 08 12:11 |
| jlaska | | no worries, that makes sense | Apr 08 12:11 |
| jlaska | | are there any next steps on that front .... or what would you see as the next step? | Apr 08 12:11 |
| f13 | | the test we run for post-tree-compose is basically just verifytree right now | Apr 08 12:13 |
| f13 | | and it has some interesting issues that need working on. It won't verify the package checksums for packages outside your running arch | Apr 08 12:13 |
| f13 | | IE it won't verify ppc packages if you're running verifytree on an i386 box | Apr 08 12:13 |
| jlaska | | ah okay | Apr 08 12:14 |
| f13 | | and I don't know if verifytree exposes any exit code API that would help us determine pass/fail on the tree | Apr 08 12:14 |
| jlaska | | for the logs ... the autoqa-results mails are being sent to https://fedorahosted.org/pipermail/autoqa-results/2009-April/thread.html | Apr 08 12:14 |
| f13 | | so basically work on the tests, and some work on the reporting | Apr 08 12:14 |
| jlaska | | f13: by reporting ... were you thinking of the autoqa-results list ... or somethign else? | Apr 08 12:15 |
| jlaska | | okay ... we can come back to that | Apr 08 12:17 |
| f13 | | which is pretty much the same story for post-repo-update | Apr 08 12:17 |
| f13 | | man, I'm lagging bad, something wrong with the proxy | Apr 08 12:17 |
| jlaska | | wwoods: you're back! | Apr 08 12:18 |
| f13 | | Ok, I think the lag storm is gone | Apr 08 12:18 |
| * | wwoods is back! | Apr 08 12:19 | |
| jlaska | | f13: did you have any other updates on the post-tree-compose? if not, we can move back to wwoods | Apr 08 12:19 |
| f13 | | I don't think so. I'd like to hand it off for now, the rest of the work should focus on the test being ran and the reporting, and that's sort of "your" side of the sandbox | Apr 08 12:21 |
| jlaska | | f13: you got it, thanks for your help in jump starting! | Apr 08 12:21 |
| jlaska | | okay ... switching back to the top ... | Apr 08 12:21 |
| jlaska | | # [wwoods] - autoqa: work on better (clearer) reporting from existing tests | Apr 08 12:21 |
| wwoods | | at the moment we're just working on getting the reporting to actually happen | Apr 08 12:22 |
| wwoods | | but yeah, the emails from e.g. repoclosure are kind of verbose | Apr 08 12:22 |
| * | jlaska posted this earlier ... https://fedorahosted.org/pipermail/autoqa-results/2009-April/thread.html | Apr 08 12:23 | |
| f13 | | right, we may want to wrap repoclosure | Apr 08 12:23 |
| wwoods | | it would be nice if we actually had some kind of reports server to store simple pass/fail/warn with a link to the full results | Apr 08 12:23 |
| jlaska | | we've got reports being mailed, that's a great start | Apr 08 12:23 |
| f13 | | or work on a different output mode for being piped into things | Apr 08 12:23 |
| wwoods | | the client-side of that is already part of rhtslib | Apr 08 12:23 |
| f13 | | wwoods: all in good time. | Apr 08 12:23 |
| wwoods | | but the result server ain't there yet | Apr 08 12:23 |
| f13 | | part of me handing off autoqa work is so that I can be free to work on the messaging bus | Apr 08 12:23 |
| wwoods | | and actually the report_results stuff is broken ATM 'cuz rhtslib changed the name of the binary | Apr 08 12:24 |
| f13 | | so that we could easily hook up the testsystem with a reporting system. | Apr 08 12:24 |
| jlaska | | wwoods: should we focus on improving the existing mails and getting that process automated ... or should efforts focus on a formal db for capturing/organizing/querying results? | Apr 08 12:24 |
| wwoods | | the messaging bus isn't really required for that | Apr 08 12:24 |
| jlaska | | and of course, when I say "we" ... I mean the work you guys have been doing | Apr 08 12:25 |
| wwoods | | jlaska: I don't know - not sure who the consumers of this data are or what they need | Apr 08 12:25 |
| wwoods | | beyond, you know, us | Apr 08 12:25 |
| jlaska | | having something available for review without needing to kick off an install etc... I think would be beneficial | Apr 08 12:26 |
| jlaska | | whether it's just a mailing list that requires inspection at the start ... or something a bit prettier (aka www.doesrawhidework.com) | Apr 08 12:26 |
| f13 | | A big win will come when we can do post-build testing | Apr 08 12:26 |
| <-- | mcepl [n=mcepl@49-117-207-85.strcechy.adsl-llu.static.bluetone.cz] has left #fedora-meeting ( ) | Apr 08 12:26 | |
| f13 | | those are ... harder tests, but that gives developers some feedback on their builds well before a rawhide compose | Apr 08 12:27 |
| jlaska | | I see ... so what's hte expected frequency of these mails | Apr 08 12:27 |
| jlaska | | it's not just a few a day for rawhide ... but possible for every koji build? | Apr 08 12:27 |
| f13 | | yeah, but we don't have to deliver those to the list necessarily | Apr 08 12:28 |
| f13 | | those can be delivered directly to <package>-owner@fedoraproject.org | Apr 08 12:28 |
| f13 | | I mean, we can keep delivering them to the list if we want, or just failures to the list or something. | Apr 08 12:29 |
| f13 | | but our goal was to get the test system under control a bit before we unleash it upon every build and every user's inbox | Apr 08 12:29 |
| jlaska | | true ... so unless any objections, the next steps are to ... | Apr 08 12:30 |
| jlaska | | - continue improving existing tests | Apr 08 12:30 |
| jlaska | | - interim goal of sending automated test result mails to autoqa-results | Apr 08 12:30 |
| jlaska | | wwoods: f13: those seem reasonable for now? | Apr 08 12:30 |
| f13 | | - continue improving existing test reporting | Apr 08 12:31 |
| jlaska | | gotcha, thanks for correction | Apr 08 12:31 |
| f13 | | that seems fine with me. | Apr 08 12:31 |
| jlaska | | okay, wwoods any thoughts/comments/concerns around those next steps? | Apr 08 12:31 |
| f13 | | we've got two semi-useful things to react to that are well understood, now it's time to fine tune everything else. | Apr 08 12:32 |
| wwoods | | yeah, those are fine | Apr 08 12:32 |
| jlaska | | wwoods: you had mentioned working some possible fixes into rhtslib ... I can put that down too if needed | Apr 08 12:33 |
| wwoods | | yeah, there's some fixes I need to get back into beaker | Apr 08 12:33 |
| jlaska | | okay | Apr 08 12:33 |
| wwoods | | and I should probably see if I can help get that reviewed and accepted into Fedora | Apr 08 12:33 |
| jlaska | | okay, I'll reach out to pmuller to how he's making out on that | Apr 08 12:34 |
| wwoods | | at some point we might want to package autoqa but I'm not feeling a strong need for that | Apr 08 12:34 |
| f13 | | yeah we can just work from a git clone for now | Apr 08 12:34 |
| jlaska | | okay ... changing topics ... | Apr 08 12:34 |
| f13 | | we'd wind up doing that anyway to avoid tones of unnecessary package shuffle | Apr 08 12:34 |
| jlaska | | * [adamw] - will file a F10 bug for the syslinux issue, initiate discussion with pjones or jeremy for building a updated F10 package | Apr 08 12:34 |
| wwoods | | f13: yeah exactly | Apr 08 12:34 |
| adamw | | bug was filed, marked as a dupe, we added an entry to the release notes | Apr 08 12:35 |
| adamw | | i think last time i saw the bug a workaround was being worked on... | Apr 08 12:35 |
| adamw | | https://bugzilla.redhat.com/show_bug.cgi?id=492370 | Apr 08 12:35 |
| buggbot | | Bug 492370: medium, low, ---, katzj, CLOSED RAWHIDE, livecd-iso-to-disk.sh requires syslinux 3.73 for burning rawhide images | Apr 08 12:35 |
| jlaska | | jeremy indicated that root cause was still being determined but he had a custom livecd-iso-to-disk.sh script available if needed | Apr 08 12:36 |
| adamw | | it's marked as "CLOSED RAWHIDE", which i don't quite understand | Apr 08 12:36 |
| * | jlaska loads bz | Apr 08 12:36 | |
| adamw | | having a hacked livecd-to-iso-to-disk.sh script available in git isn't really the same thing as resolving the bug, IMO - ah well | Apr 08 12:37 |
| jlaska | | so does that mean the next livecd-tools package rebuild will include the fix ... and there's a custom script available for testing until then? | Apr 08 12:37 |
| adamw | | that's not what the comment seems to say... | Apr 08 12:38 |
| adamw | | well, i've just posted another comment | Apr 08 12:38 |
| jlaska | | okay | Apr 08 12:38 |
| adamw | | let's see what happens now | Apr 08 12:38 |
| jlaska | | thanks for the update | Apr 08 12:38 |
| jlaska | | anything else we need to do on that front? | Apr 08 12:38 |
| adamw | | don't think so | Apr 08 12:39 |
| jlaska | | okay, we'll wait to hear back on your latest comment | Apr 08 12:39 |
| adamw | | yep | Apr 08 12:39 |
| jlaska | | okay ... | Apr 08 12:39 |
| jlaska | | * [jlaska] - check-in with Warren Togami on the rawhidewatch blog ... can Fedora QA help ... what is his vision? | Apr 08 12:39 |
| * | jlaska likes to say "okay" | Apr 08 12:39 | |
| jlaska | | I spoke with warren this morning about http://rawhidewatch.wordpress.com and what his idea/plan is around the blog | Apr 08 12:40 |
| jlaska | | the summary is Warren intends for the blog to be a low traffic blog that includes imformation on non-obvious (potentially high-impact) problems affecting rawhide users | Apr 08 12:41 |
| adamw | | i sent him a suggestion for it (to add a note on the nss issue that caught out some x86-64 users and broke rpm); he basically told me 'patches welcome', i.e. write an entry and he'll post it :) i was going to do that but haven't had time yet | Apr 08 12:41 |
| jlaska | | adamw: yeah, warren said that he'll act as editor to the blog and accept story submissions | Apr 08 12:41 |
| warren | | adamw: you mailed that? I might have missed it | Apr 08 12:41 |
| f13 | | cool | Apr 08 12:42 |
| adamw | | warren: i sent you the suggestion on IRC. i haven't written the story yet i'm afraid, try and get to it today | Apr 08 12:42 |
| warren | | adamw: I need a lot more detail to write a proper story, I asked for a followup via e-mail | Apr 08 12:42 |
| adamw | | oh you did? i think I missed that :) | Apr 08 12:42 |
| adamw | | all the detail is in the bug report | Apr 08 12:43 |
| warren | | glibc's change is somehow involved right? | Apr 08 12:43 |
| adamw | | i'll send you either a story or more details by email | Apr 08 12:43 |
| warren | | corrupting checksums | Apr 08 12:43 |
| adamw | | let's not hold up the meeting for it | Apr 08 12:43 |
| adamw | | no, that's a separate issue | Apr 08 12:43 |
| warren | | oh | Apr 08 12:43 |
| warren | | yes, please summarize in blurb form in e-mail | Apr 08 12:43 |
| jlaska | | adamw: thanks, I've added that for next week | Apr 08 12:44 |
| jlaska | | okay last up .. | Apr 08 12:44 |
| jlaska | | * [jlaska+wwoods] - discuss integration between lab-in-a-box and autoqa | Apr 08 12:44 |
| jlaska | | I failed on this one, wwoods and I spoke about it briefly, but I have not scheduled time to chat yet | Apr 08 12:45 |
| jlaska | | I'm still trying to figure out how best to plug the 'lab-in-a-box' into some framework | Apr 08 12:45 |
| jlaska | | I don't have a good sense yet whether that's autoqa, something beaker-like, ... or something new that uses f13's watch script | Apr 08 12:46 |
| jlaska | | if folks have suggestions, please do drop me a line | Apr 08 12:46 |
| jlaska | | basically looking for a way to automate rawhide installations ... and post the results (all in a virt guest) | Apr 08 12:46 |
| jlaska | | the missing pieces for me are the scheduler (something that tells it when to start), and results reporting (something that gathers the existing results and does _something_) | Apr 08 12:47 |
| jlaska | | so ... yeah suggestions, please drop me a line :) | Apr 08 12:47 |
| jlaska | | okay tha'ts it for last week | Apr 08 12:47 |
| --- | jlaska has changed the topic to: Fedora Quality Assurance meeting | F-11 | Apr 08 12:48 | |
| wwoods | | yeah, we'll discuss that sometime this week | Apr 08 12:48 |
| jlaska | | I wanted to spend a few minutes just discussing the schedule and the current planned QA activites for F-11 | Apr 08 12:49 |
| jlaska | | but if folks have topics to discuss prior to ... we can hold that | Apr 08 12:49 |
| f13 | | SNAPSHOT | Apr 08 12:49 |
| jlaska | | f13: I see just 1 snapshot on the release, is that right (http://poelstra.fedorapeople.org/schedules/f-11/f-11-releng-tasks.html) | Apr 08 12:49 |
| jlaska | | fyi ... F-11 schedule is available at http://fedoraproject.org/wiki/Releases/11/Schedule | Apr 08 12:50 |
| f13 | | yeah | Apr 08 12:50 |
| f13 | | originally there were 2 snapshots, but we lost a week due to beta, and now there is only one | Apr 08 12:50 |
| jlaska | | f13: I suspect the issue you & panu are chasing might affect that/ | Apr 08 12:50 |
| jlaska | | ? | Apr 08 12:50 |
| f13 | | yes | Apr 08 12:51 |
| jlaska | | f13: https://www.redhat.com/archives/fedora-test-list/2009-April/msg00535.html | Apr 08 12:51 |
| f13 | | i386 image creation is broken | Apr 08 12:51 |
| f13 | | we suspect that its broken due to glibc memory alloc options that were introduced recently (april 1) | Apr 08 12:51 |
| f13 | | in conjunction with some nss breakage. | Apr 08 12:51 |
| f13 | | I'm working with jakub to get a build of glibc that may fix this | Apr 08 12:51 |
| jlaska | | f13: okay, please shout if you need anyone to help on the testing front | Apr 08 12:52 |
| --- | nman64 is now known as nman64_away | Apr 08 12:52 | |
| f13 | | Will do. | Apr 08 12:52 |
| jlaska | | f13: wwoods: you guys have the most experience in this space ... | Apr 08 12:52 |
| jlaska | | but looking at the schedule http://fedoraproject.org/wiki/Releases/11/Schedule | Apr 08 12:52 |
| jlaska | | we've got about 1 month before the final compose date | Apr 08 12:53 |
| jlaska | | what can we be doing better between now and that date? | Apr 08 12:53 |
| jlaska | | we have several test days planned: QA/Test_Days/F11 | Apr 08 12:53 |
| jlaska | | adamw: you have triage events scheduled? | Apr 08 12:53 |
| adamw | | triage days are regular every tuesday at present | Apr 08 12:53 |
| adamw | | had a good one yesterday, two new guys, one who will be triaging hal | Apr 08 12:54 |
| jlaska | | there's been a lot of positive activity on that front, that's great news | Apr 08 12:54 |
| f13 | | jlaska: one lesson from beta, we need to get the test grid up early (prior to the freeze) and people chipping away at it to discover what is broken before we go into freeze | Apr 08 12:54 |
| jlaska | | f13: "test grid" .. oh you mean the wiki results? | Apr 08 12:55 |
| f13 | | (in lieu of having automation to run through the entire suite each night) | Apr 08 12:55 |
| f13 | | yeah, the wiki results | Apr 08 12:55 |
| wwoods | | "test grid" is an overloaded phrase in RHEL QE | Apr 08 12:56 |
| jlaska | | for next week's anaconda storage testing, perhaps that would be a good target ... fleshing out that matrix | Apr 08 12:56 |
| jlaska | | wwoods: heh :) | Apr 08 12:56 |
| f13 | | sorry (: | Apr 08 12:56 |
| wwoods | | just needed some clarification. heh. | Apr 08 12:56 |
| wwoods | | yeah, should add some things to it and get it up early | Apr 08 12:57 |
| wwoods | | so we can start testing using the first available rel-eng-approved rawhide tree | Apr 08 12:57 |
| jlaska | | adamw: I'd really like to know what we're up against in terms of bugs for F-11 ... the post-beta bug focus mainly is F11{Blocker,Target} ... but are we getting everything on that list that should be there? What's your take | Apr 08 12:57 |
| adamw | | jlaska: i don't really have a great overview...at least i'm not aware of any major issue that's not getting the attention it deserves, no | Apr 08 12:58 |
| wwoods | | I'm personally worried about the pulseaudio/alsa bugs | Apr 08 12:58 |
| adamw | | which ones? :) | Apr 08 12:58 |
| wwoods | | (bug 472339 and associated bugs) | Apr 08 12:58 |
| buggbot | | Bug https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=472339 medium, medium, ---, kernel-maint, NEW, snd-intel8x0: timing unstable (snd_pcm_avail() overflows, signals POLLOUT when it shouldn't) | Apr 08 12:58 |
| wwoods | | trying to help lennart chase down the problem (and/or ALSA devs to look at it) | Apr 08 12:59 |
| wwoods | | sound is broken on every test machine I have :/ | Apr 08 12:59 |
| f13 | | my pulse dies after about a day or so | Apr 08 12:59 |
| * | jlaska +1 | Apr 08 13:00 | |
| f13 | | but I hardly notice because I still swapstorm after about a day or so | Apr 08 13:00 |
| f13 | | that's a big concern of mine, I don't think we should release F11 in that state | Apr 08 13:00 |
| wwoods | | the swapstorm? | Apr 08 13:00 |
| f13 | | yeah | Apr 08 13:00 |
| wwoods | | are you using desktop effects / watching lots of video? | Apr 08 13:00 |
| jlaska | | wwoods: so is this a matter of getting lennart some help or are the bugs not on his radar? | Apr 08 13:00 |
| f13 | | using compiz yes, not watching videos | Apr 08 13:00 |
| wwoods | | there's a couple known resource leaks in DRI2 | Apr 08 13:00 |
| f13 | | right | Apr 08 13:00 |
| f13 | | we should get those fixed before releasing F11 | Apr 08 13:00 |
| wwoods | | right - there's a mail thread where one major one was found last week-ish | Apr 08 13:01 |
| adamw | | wwoods: oh. that's why i'm not getting any sound? heh/ | Apr 08 13:01 |
| adamw | | wwoods: still, it's on F11Blocker, i'm not sure what else we can do... | Apr 08 13:01 |
| wwoods | | http://www.mail-archive.com/xorg@lists.freedesktop.org/msg06116.html | Apr 08 13:01 |
| f13 | | we can trim the Blocker list for the things we'll actually delay the release for | Apr 08 13:01 |
| f13 | | which gives developers (and managers) a better priority list | Apr 08 13:01 |
| adamw | | yeah, it may be an idea to go through Blocker and move things to Target | Apr 08 13:01 |
| wwoods | | jlaska: yeah, no, lennart is well aware and he's helping me trace this stuff | Apr 08 13:01 |
| f13 | | its about that time to do so | Apr 08 13:01 |
| wwoods | | but he's gotten as far as he can - once you start wading into alsa-libs or kernel space it gets very hazy | Apr 08 13:02 |
| jlaska | | wwoods: okay, thx | Apr 08 13:02 |
| f13 | | we typically have a day long meeting to prune the blocker list | Apr 08 13:02 |
| f13 | | should kick that up again before we hit final freeze | Apr 08 13:02 |
| jlaska | | f13: are those in the schedule? | Apr 08 13:02 |
| wwoods | | we're trying to get pressure on jaroslav (alsa guy, works for RH) to help with this | Apr 08 13:02 |
| f13 | | jlaska: I don't think so | Apr 08 13:02 |
| f13 | | everybody should take an action item to review the current F11Blocker and prune as they see fit | Apr 08 13:02 |
| wwoods | | and yeah, we should probably have a blocker bug to track the DRI2 resource leaks / swapstorm of doom | Apr 08 13:02 |
| f13 | | then we'll try to get sig leaders together to do a walk through | Apr 08 13:02 |
| adamw | | if we can get a couple days notice of that meeting it'd be great, we can try and get the qa and bugzappers communities out for it | Apr 08 13:02 |
| jlaska | | f13: want me to propose a few time slots and schedule? | Apr 08 13:03 |
| f13 | | yeah | Apr 08 13:03 |
| wwoods | | ajax et. al. seem to be aware that there are some, but most of 'em are thought to be closed (with the possible exception of the one I linked to moments ago - not sure if that fix has landed anywhere yet) | Apr 08 13:03 |
| wwoods | | definitely a blocker review is in order | Apr 08 13:03 |
| jlaska | | f13: okay, I'll reach out to the schedule wrangler | Apr 08 13:03 |
| f13 | | it's gotta be late this week or early next week | Apr 08 13:04 |
| f13 | | or after the final freeze | Apr 08 13:04 |
| jlaska | | how much time will these need? | Apr 08 13:04 |
| jlaska | | a few hours? | Apr 08 13:04 |
| adamw | | i should probably go over the video test day matrices and make sure critical issues there are on blocker | Apr 08 13:04 |
| wwoods | | once we have a fairly solid set of blockers we should start ringing alarm bells to get maintainers / owners / etc. working on 'em | Apr 08 13:04 |
| jlaska | | adamw: notting mentioned that to me earlier, that would be great! | Apr 08 13:04 |
| adamw | | ok, stick that down as an action item for me then | Apr 08 13:05 |
| f13 | | jlaska: yeah, probably a few hours | Apr 08 13:05 |
| jlaska | | adamw: roger | Apr 08 13:05 |
| f13 | | depending on the length of the blocker list | Apr 08 13:05 |
| jlaska | | okay ... so we've got a lot of blocker bug review that's planned | Apr 08 13:05 |
| jlaska | | we have several test days leading up to the Preview and after | Apr 08 13:06 |
| fhornain | | Belgian Guys ? | Apr 08 13:06 |
| jlaska | | any other large holes or areas of concern for folks? | Apr 08 13:07 |
| f13 | | oh, we could use somebody making up live images today/tomorrow and trying to install from them. | Apr 08 13:08 |
| f13 | | since snapshot is just live images | Apr 08 13:08 |
| jlaska | | f13: while live images won't be needed for UEFI testing, I cna build some and play tomorrow | Apr 08 13:08 |
| f13 | | live images are pretty inexpensive to make | Apr 08 13:09 |
| f13 | | but expensive to try and transport | Apr 08 13:09 |
| jlaska | | definitely | Apr 08 13:09 |
| jlaska | | okay ... let's move on to open discussion | Apr 08 13:09 |
| --- | jlaska has changed the topic to: Fedora Quality Assurance meeting | open discussion | Apr 08 13:10 | |
| f13 | | how useful was the board meeting yesterday? | Apr 08 13:10 |
| jlaska | | I'm probably the wrong person to ask ... as a first timer, the forum was odd for me | Apr 08 13:10 |
| jlaska | | f13: what's your take, did you get any feedback? | Apr 08 13:11 |
| f13 | | I didn't, but I ran out half way though | Apr 08 13:11 |
| f13 | | I think the board was was happy to hear about the status, but I was more concerned with the community at large | Apr 08 13:12 |
| jlaska | | it was nice to hear that some of the community events (test+triage days) are being received well | Apr 08 13:12 |
| jlaska | | f13: keep me posted if there is anything folks would like a follow-up on | Apr 08 13:13 |
| f13 | | will do | Apr 08 13:13 |
| jlaska | | okay ... any other open discussion topics? | Apr 08 13:14 |
| jlaska | | I'll toss one into the ring ... | Apr 08 13:16 |
| jlaska | | there was some discussion around creating test day live images on the mailing list last week | Apr 08 13:16 |
| jlaska | | I was looking into adding a new kickstart file that extends the current fedora-livecd-desktop.ks included in spin-kickstarts | Apr 08 13:17 |
| wwoods | | adding some test tools? | Apr 08 13:17 |
| wwoods | | (strace/ltrace/etc) | Apr 08 13:17 |
| jlaska | | I'm not sure if that's the right place for that type of kickstart file, but I wanted tomake the test day live images as *close* as possible to the spin-kickstarts used for official releases | Apr 08 13:17 |
| jlaska | | yeah ... to add some debugging tools, perhaps a desktop shortcut to the page stickster created Test_Day:Current | Apr 08 13:18 |
| jlaska | | am I barking up the right tree here? | Apr 08 13:18 |
| f13 | | its not a terrible place to put it | Apr 08 13:18 |
| adamw | | sounds good to me, only caveat is - as i said to you before - make sure it's not something that starts eating up too much time | Apr 08 13:18 |
| adamw | | we need the live cd to be good enough for people to run tests, not super-pretty with a bow on top :) | Apr 08 13:18 |
| jlaska | | adamw: here here :) | Apr 08 13:19 |
| jlaska | | adamw: I'd like the kickstart to be somewhere where you, or others, could build and post test day live images if needed | Apr 08 13:19 |
| f13 | | perhaps building it on top of livecd-base is better? | Apr 08 13:19 |
| f13 | | that way it isn't heavy on the gnome side or kde side or any side | Apr 08 13:20 |
| jlaska | | true, that's an idea | Apr 08 13:20 |
| f13 | | livecd-base -> livecd-testday-base -> <package set you care about> | Apr 08 13:20 |
| jlaska | | oh, I like that | Apr 08 13:20 |
| jlaska | | okay thanks ... I'll try to pull something basic together to play with | Apr 08 13:20 |
| jlaska | | okay folks let's call it a wrap in 5 ... 4 | Apr 08 13:21 |
| jlaska | | and apologies for taking up 1:21 of your time :( gotta get these meetings to <= 1hr | Apr 08 13:21 |
| jlaska | | 3 ... 2 ... | Apr 08 13:21 |
| f13 | | *shrug* they should take as long as they need to | Apr 08 13:22 |
| jlaska | | 1 ... thanks folks for your time! | Apr 08 13:22 |
| jlaska | | I'll send minutes out to the list shortly | Apr 08 13:22 |
| --- | jlaska has changed the topic to: Channel is used by various Fedora groups and committees for their regular meetings | Note that meetings often get logged | For questions about using Fedora please ask in #fedora | See http://fedoraproject.org/wiki/Meeting_channel for meeting schedule | Apr 08 13:23 | |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!