=== fader_ is now known as fader|away === asac_ is now known as asac === dholbach_ is now known as dholbach === ogra_ is now known as ogra === alsroot__ is now known as alsroot === Pici` is now known as Pici === fader|away is now known as fader_ === imlad|away is now known as imlad === robbiew-afk is now known as robbiew === bjf-afk is now known as bjf === Keybuk_ is now known as Keybuk [15:56] * james_w waves [15:57] hiya [15:58] * mterry waits until #startmeeting [16:00] hi [16:01] what's the main airport for Dallas? [16:01] Fort Worth? [16:01] hello [16:01] DFW [16:01] #startmeeting [16:01] Meeting started at 10:01. The chair is robbiew. [16:01] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [16:02] * robbiew apologizes for no agenda...up past 4am working on xsplash bug :/ [16:02] robbiew: congrats on the bug fix ;) [16:02] hello all [16:03] * robbiew has to keep is "street cred" with the developers from time to time ;) [16:04] ehlo [16:04] helo [16:04] ok [16:05] is slangasek off today? [16:05] * robbiew should probably know that :/ [16:06] he was around earlier ... [16:06] so what do we have scheduled for Beta release? [16:06] anything slip out? [16:06] feature wise [16:06] let's see, https://wiki.ubuntu.com/FoundationsTeam/ReleaseStatus/Karmic [16:06] hi [16:07] I think technically my 'oem-config improvements' spec slipped, but that was expected. remaining work should be deferred [16:07] eh...I don't count that ;) [16:07] boot performance; archive reorg; deb/rsync; kerneloops; oem-config; wubi updates [16:07] I think BootPerf can be considered done [16:07] deb/rsync and wubi updates were deferred [16:07] just fixing bugs now [16:07] kerneloops is done [16:07] major change expected from archive reorg this cycle is done though we're still working on ongoing changes [16:08] nice [16:08] I'm quite pleased with the SSD boot performance [16:08] mvo: more-robust-python-packaging, we should settle the status of that task on releasestatus/karmic one way or another [16:08] I think everything else is as out of the way as it's supposed to be [16:09] cjwatson: its partly done, party defered, I would lean toward defered [16:09] james_w: are you expecting anymore progress with regard to Dist Dev or Daily Builds? [16:09] or are we "done" for this cycle [16:09] * slangasek holds the network wires together with one hand and waves with the other [16:09] hopefully [16:09] heh [16:09] nothing in the archive though [16:09] slangasek: I feel your pain [16:10] slangasek: isn't that what teeth are for? [16:10] Keybuk: is the Packaging and Upgrade Policy for Upstart stuff done? [16:10] robbiew: no [16:10] will it be for 9.10 :P [16:10] probably not ;) [16:10] because it'd be changed again by 10.04 :P [16:11] heh [16:11] ok...will defer [16:11] [ACTION] robbiew to defer Packaging and Upgrade Policy for Upstart [16:11] ACTION received: robbiew to defer Packaging and Upgrade Policy for Upstart [16:12] and where are we at with bug 432089 [16:12] Launchpad bug 432089 in sreadahead "performs poorly on slow HDD" [High,Triaged] https://launchpad.net/bugs/432089 [16:13] Keybuk: ^ [16:13] deferred === marjomercado is now known as marjo [16:14] cf. my comment there [16:14] well, deferred though substantially improved, right ... [16:14] (door) [16:14] the deadline scheduler change does certainly improve it [16:14] amongst other things [16:14] the real problem is that HDD performance optimisation is *hard* [16:14] sreadahead is "at least as good" as readahead-list on HDD [16:14] though I wouldn't use the word "good" there [16:15] okay [16:15] my main fear was regressing people with HDDs [16:15] if we can say we haven't done that...then I'm okay [16:15] I've not seen any evidence that it's worse [16:16] yeah...there's the usual "stories" of people claiming slowness [16:16] but without bootcharts to back it up...nothing we can do [16:17] indeed [16:17] and yes, jaunty boots faster [16:17] but jaunty did not have couch potato based fat bastard db [16:17] lol [16:17] I thought they were fixing that [16:17] slangasek: did you get any late fixes around couch? [16:17] for beta? [16:17] system daemon, or per-user? [16:19] I believe the default install no longer starts the former [16:19] robbiew: there were some uploads, but they arrived late in the beta freeze with nothing I could see that was beta-critical [16:20] (in fact, the one bug report that was being tracked was a non-bug) [16:20] nice [16:20] ok [16:20] the previous upload was the one that was supposed to fix that [16:20] so those are in the queue and will be released post-beta [16:20] james_w: so I shouldn't see couchdb start at all? or does it start after user-login? [16:21] robbiew: it will start in your session if something uses it [16:21] * slangasek nods [16:21] but perhaps the fix for the system daemon was incomplete [16:22] hmmm [16:23] "perhaps"? [16:23] * Keybuk hasn't looked in the last week [16:24] just hypothesising, I believe the system daemon shouldn't start on the beta images [16:24] it doesn [16:24] argh US keyboard [16:24] it doesn't look like it started on this [16:25] ok [16:26] cjwatson: so are we looking better for the cloud installer stuff [16:26] * robbiew backs away to avoid a swing by cjwatson :P [16:27] there are some startup race difficulties at the moment, which ttx and kirkland are working [16:27] and Keybuk commented on a short while ago in #ubuntu-devel [16:27] the installer as such seems to be largely OK no [16:27] w [16:28] it's eucalyptus itself and its startup that are busted [16:28] right...I saw that, thanks [16:28] yeah, and the eucalyptus scripts are doing strange things that I don't even know the behaviour of ;) [16:29] what was strange about eucalyptus-cloud BTW? [16:29] just the timeout? [16:29] it uses a post-start script [16:29] I already suggested to mdz that that should use respawn limit [16:29] without any main process, etc. [16:29] I'm not even sure that works [16:29] Keybuk: I commented on that yesterday, and ion immediately said it worked [16:29] or was supposed to [16:29] and I'm certainly not sure of upstart's behaviour when it exits ;) [16:29] now, he might have been wrong, I didn't check further after that [16:30] since there are bugs in RECENT upstart history which resulted in a kernel panic when using post-start, it's not very well tested code [16:30] hah [16:30] why is it post-start anyway? [16:30] poke ion for me, then ;) [16:30] I couldn't find any reason why it shouldn't be pre-start [16:31] in fact, if you don't have a main process, there really isn't much difference between pre-start and post-start [16:31] I don't know of a good reason for that, no [16:31] mdz just said "it worked for me" when I asked about it [16:31] :D [16:32] I guess I need someone to explain to me what they're trying to do [16:32] Keybuk: cjwatson: if it helps, i have the precise order that the scripts should run, according to nurmi [16:33] kirkland: that would help [16:33] ok...I don't have anything else [16:33] [TOPIC] Good News? [16:33] New Topic: Good News? [16:34] * robbiew can still use gdb :P [16:34] I inadvertently tested the GRUB2 menu on recovery feature [16:34] and it worked [16:34] my home file server's 8 hard disks and SATA controller haven't totally failed yet! [16:34] "oh, that's odd, I got the GRUB menu ... ah, it failed to boot" [16:34] heh [16:34] Keybuk: (might be nice if it told you why it was showing the grub menu, but I haven't figured out how yet ...) [16:34] Ah, is that why I've been given the grub menu? :) [16:34] I thought the timeout got removed somehow [16:35] I think I nailed about 75% of the remaining clock handling bugs in the installer [16:35] yeah [16:35] this is your weekly dose of extremely qualified good news [16:35] sadly the reason I failed to boot was one of the other 25% :) [16:35] Keybuk: hmm? [16:35] I actually, genuinely, think there might be another ext4 bug here [16:35] it doesn't look like it updates the last mount time timestamp when you *unmount* [16:36] so if you boot with a fast clock [16:36] I managed to spend time on something other than softtware-center last week [16:36] and ntpdate fixes it [16:36] I also have plans to actively compare NTP with the hardware clock + timezone, and try to guess better [16:36] then on shutdown we fix the hardware clock to be right [16:36] mvo: \o/ [16:36] but unmounting the ext4 filesystem doesn't fix that [16:36] so you reboot with a mount time in the future === zul_ is now known as zul [16:36] but ext4 is bug-free! [16:36] ahem [16:36] Keybuk: http://pastebin.ubuntu.com/282246/ [16:37] mvo: now you can help with compiz stuff instead ;) [16:37] good news: I can almost make nfs-common start via upstart [16:37] stopping is another matter [16:37] heh [16:37] Amaranth: heh :) [16:37] Amaranth: uh...no [16:38] kirkland: right [16:38] if they're supposed to run in series, these conf files won't do that :p [16:38] slangasek: hmm? [16:38] Keybuk: I was prodding at bug #431248... did you not get my voicemail? :) [16:38] Launchpad bug 431248 in portmap "NFS not mounted" [High,Triaged] https://launchpad.net/bugs/431248 [16:39] I don't have voicemail? [16:39] * Keybuk turned it off [16:39] oh cool [16:39] kayak.co.uk makes looking for flights significantly less painful [16:39] the trick to those is to break nfs-common into lots of different conf files [16:39] thus /etc/init/rpc.statd.conf [16:39] too bad the message isn't turned off, inviting me to leave one :P [16:39] etc. [16:39] Keybuk: yes, I'm aware [16:39] :) [16:39] it's slow going though, to keep my eyes from bleeding too much [16:40] (from the existing nfs-common code) [16:40] Keybuk: well, theoretically some of these can run in parallel, but the euca guys aren't comfortable with that; they've long trusted their series of init scripts [16:40] okay...I think we're done. Don't forget to setup your travel for the release sprint (if you are attending) and UDS [16:40] #endmeeting [16:40] Meeting finished at 10:40. [16:40] slangasek, ifyou turn off your monitor, your eyes will bleed less [16:40] thanks [16:40] kirkland: OOI, why was this converted to Upstart? [16:41] thanks all [16:41] muchas gracias [16:41] thanks [16:41] Keybuk: i expected to convert to upstart for Lucid, mdz decided to do it immediately [16:41] ta [16:42] Keybuk: i'm not positive of his rationale, but i suspect it's something along the lines of "those init scripts were an absolute mess" [16:42] Keybuk: and that we thought we could enforce the dependencies better using upstart === ian_brasil_afk is now known as ian_brasil [17:59] * fader_ waves. [18:00] #startmeeting [18:00] Meeting started at 12:00. The chair is marjo. [18:00] Commands Available: [TOPIC], [IDEA], [ACTION], [AGREED], [LINK], [VOTE] [18:00] Hi folks! [18:00] QA Team Meeting Agenda [18:00] hello [18:00] Agenda: [18:00] * SRU testing -- sbeattie (Steve Beattie) [18:00] * Bugday highlights -- bdmurray (Brian Murray) [18:00] hola! [18:00] Are there any new topics you want to add to the agenda? [18:01] *cough* beta testing *cough* [18:01] ah, but of course! [18:01] just don't mention wubi [18:01] ouch [18:02] Wubi? Never heard of it O:-) [18:02] heh [18:02] #topic SRU Testing - sbeattie [18:02] [TOPIC] SRU Testing [18:02] New Topic: SRU Testing [18:02] SRU Activity for the past week (since 2009-09-16): [18:02] * Jaunty: 5 new packages in -proposed (apport, dbconfig-common, language-packs, net-snmp, pythonmagick) and 8 packages pushed to -updates (apport, dbconfig-common, firefox-3.5, gitosis, kdenetwork, linux, opencity, xulrunner-1.9.1) [18:02] * Intrepid: 1 new package in -proposed (net-snmp) and no packages pushed to -updates [18:02] * Hardy: 1 new package in -proposed (net-snmp) and 1 package pushed to -updates (linux-ubuntu-modules) [18:02] * Dapper: no changes this week [18:03] sorry, since 2009-09-23 [18:03] Thanks to Alok Kataria, A. Karl Kornel, Pascal DANEK, stgraber, and Jeremy Wilkins for testing SRU packages this week. [18:04] sbeattie: anything else? [18:04] That's it for SRUs this week. [18:04] thx [18:05] [TOPIC] Bugday highlights -- bdmurray [18:05] New Topic: Bugday highlights -- bdmurray [18:05] I think that agenda item is stale / left-over from last week's meeting. [18:05] oops, sorry [18:05] No problem, I should have looked at it in advance too. [18:05] ara: are you able to give Beta status? [18:06] [TOPIC] Karmic Beta Testing Status [18:06] New Topic: Karmic Beta Testing Status [18:06] yes, the server is a bit slow now, as it has been respun a couple of times due to a UEC bug [18:06] but the rest is looking good [18:06] Yeah! [18:07] we shouldn't expect (finger x'ed) another respin [18:07] but we do need help in testing the untested test cases [18:07] http://iso.qa.ubuntu.com/qatracker/build//inprogress [18:07] LINK received: http://iso.qa.ubuntu.com/qatracker/build//inprogress [18:07] ara: is it true only moblin remix is untested!? [18:07] http://iso.qa.ubuntu.com/qatracker/build/all/untested [18:07] LINK received: http://iso.qa.ubuntu.com/qatracker/build/all/untested [18:07] marjo, no, wrong link [18:08] http://iso.qa.ubuntu.com/qatracker/build/all/inprogress [18:08] LINK received: http://iso.qa.ubuntu.com/qatracker/build/all/inprogress [18:08] those two [18:08] ah, got it! [18:08] FYI I'm pulling down the kubuntu DVD now and will start on some of those tests shortly [18:08] *amd64 [18:08] Folks: Can anyone on the meeting help out? [18:08] * pedro_ starts to sync ubuntu studio i386 [18:09] marjo: I'll make sure automated testing is poised for tomorrow [18:09] fader, pedro, cr3: thx! [18:09] ^^ And I plan to head into Lexington tomorrow to hit the netbooks there :) [18:09] * sbeattie is working on xubuntu and ubuntu-server upgrade tests [18:09] I'm about to have tea then kill the last of the ubuntu alternate tests. Then kill edubuntu then whatever is left [18:09] looking forward to good beta test coverage [18:10] sbeattie, fader: thx [18:10] davmor2: great! [18:10] any ideas on how to get upgrade coverage? [18:11] that's what I'm working on [18:11] thx! [18:11] marjo: edubuntu is being respun to kill sbeatties bug [18:11] looks like it was just a seed issue [18:12] davmor2: yes, as you suspected [18:12] or hoped for [18:12] * Riddell hugs fader_ [18:12] ara: if everyone does what they just said, what would be left uncovered? [18:12] :D [18:13] well, upgrades needs more people, I guess [18:13] sbeattie: how much of the upgades list can you cover? [18:14] download info Upgrade Edubuntu amd64 (20090929) 0/1 [18:14] download info Upgrade Edubuntu i386 (20090929) 0/1 [18:14] download info Upgrade Kubuntu amd64 (20090929) 0/1 [18:14] download info Upgrade Ubuntu Server amd64 (20090930) 0/1 [18:14] download info Upgrade Xubuntu amd64 (20090929) 0/1 [18:14] download info Upgrade Xubuntu i386 (20090929) 0/1 [18:14] marjo, and also, when edubuntu is done, we need to retest [18:14] marjo: I'm working on 3 of those 6 right now, will move to other 3 when I'm finished with these. [18:15] sbeattie: nice! [18:15] sbeattie: Out of curiosity, do you have images of previous installs handy that you might be able to share? [18:15] E.g. put on a server somewhere? [18:15] (I'm assuming you're using virtualization for the upgrade testing) [18:15] fader_: not... easily. my outbound bandwidth is pretty slow. [18:16] sbeattie: Okay, I thought that might be the case but it didn't hurt to ask :) [18:16] ara: any ideas for edubuntu DVD amd64? i don't see a subscriber to it [18:17] marjo: I will grab that one if nobody else does by the time I finish the kubuntu DVD [18:17] fader_: thx! [18:17] fader_, nice [18:18] so, does that mean all is covered?! [18:18] that would be great, after edubuntu is respun [18:19] all: can you please go ahead and subscribe so that the subscription summary is updated? [18:19] thx [18:20] for RC it will be nice to have more people on the platform team testing [18:20] marjo: FYI, the subscriptions go past the immediate testing run and indicate that you're interested in testing every new spin of an image. I don't mind testing any of them but it might be a good idea to try to find people who use Edubuntu regularly to help test it, for example [18:20] ara: i agree; let's work on that [18:20] I do try to grab images that haven't been tested even if I'm not subscribed [18:20] fader_ : oic [18:21] fader_, yes, I think that's what most of us do [18:21] ara: Indeed, I didn't mean to imply that it was just me ;) [18:21] fader_, I know :D [18:21] aren't people on the platform team busy frantically fixing bugs, or are there so few bugs that they really have spare cycles for testing? [18:21] fader_ : yes, that's one of the things ara is working on - recruiting regular testers [18:22] cr3: the idea is to get more coverage by folks who know and use the product well [18:22] cr3, one test case per person per milestone is achievable, I think [18:22] but in general, the more the better and merrier [18:23] cr3, we are about 80 people in the platform team [18:24] marjo: I would suspect that the coverage you speak of is already done before pushing a new package to the repository [18:25] cr3: you're right, but I'm thinking more of "test coverage" as in hitting all the cases called out in the iso tracker [18:25] ara: will we be assigning specific cases to members of the platform team or will they pick cases according to their expertise? [18:25] ara: nevermind, "iso tracker" comment by marjo answers my question :) [18:26] cr3, I think that the pick what's untested in the iso tracker is a good strategy [18:26] currently there are 7 red bugs showing [18:27] according to davmor2, https://bugs.launchpad.net/wubi/+bug/439279 is being worked [18:27] Launchpad bug 439279 in wubi "There seems to be some instability with the installer" [Undecided,New] [18:28] davmor2: sorry, i had to mention wubi [18:28] any new status on the other 6? [18:29] https://bugs.launchpad.net/ubuntu/+bug/431206 is already milestoned by the mobile team [18:29] Launchpad bug 431206 in Ubuntu Karmic "partman-uboot needs to exist for check script for supporting manual partitioning" [Undecided,New] [18:30] ok, folks, anything else on Beta? [18:31] imho, it's looking pretty good, no? [18:31] any other topics for today? [18:32] ok, if nothing else, i propose we adjourn [18:32] thx to all and good luck for Beta release! [18:32] #endmeeting [18:32] Meeting finished at 12:32. [18:32] Thanks everyone [18:33] thanks [18:49] Hi, I have a question about the upcoming jams. I'm a programmer and ubuntu user, but haven't done any particularly ubuntu-specific work - is there anything it would be useful to do to prepare for the jam and increase the odds I can do something useful there? === swoody_ is now known as swoody === nxvl_ is now known as nxvl === starcraftman is now known as starcraft === starcraft is now known as starcraftman === robbiew is now known as robbiew-afk