[01:14] <phillw> when bdmurray is away all the bug team scarper... :D
[05:47] <infinity> superm1: Is anyone doing 12.04.3 testing for mythbuntu?
[05:49] <infinity> Riddell: Kubuntu 12.04.3 testing seems a bit light.
[05:49] <infinity> knome: Are Xubuntu 12.04.3 alternates being tested too?
[07:42] <ara> hello!
[07:42] <ara> is 12.04.3 release happening today?
[07:43] <ara> there was some conversations about maybe delaying it?
[08:21] <ara> wgrant, is the 12.04.3 release happening today?
[08:22]  * Laney wonders how to unstick RUNNING tests
[08:24] <seb128> Laney, retry the test in jenkins?
[08:24] <seb128> Laney, which one is that?
[08:24] <Laney> how?
[08:24] <Laney> see glib2.0 @ excuses
[08:25] <seb128> Laney, e.g http://10.98.0.1:8080/view/Saucy/view/AutoPkgTest/job/saucy-adt-chromium-browser/10/matrix-reloaded/?
[08:25] <seb128> Laney, click "rebuild matrix"
[08:25] <Laney> hahaha
[08:25] <Laney> seriously?
[08:25] <seb128> well,  that's what pitti told me the other day
[08:26] <Laney> ok i tried it for glib
[08:28] <seb128> I did it for chromium
[09:11] <TheDrums> infinity: Re: Xubuntu: < elfy> kn0me: not getting far with the alternate tests for 12.04.3 -> http://imagebin.org/268366
[12:43] <superm1> infinity: yeah we'll make sure iso's get checked today
[12:59] <tkamppeter> jdstrand, you have moved bug 711061 to sarnold and there is still no reaction and we have FF next week. I have already updated to GS 9.09 final.
[12:59] <ubot2`> Launchpad bug 711061 in openjpeg (Ubuntu) "[MIR] libopenjpeg2" [High,Confirmed] https://launchpad.net/bugs/711061
[13:18] <diwic> Hi, I uploaded a PulseAudio 12.04 SRU almost a month ago - what is the current waiting time for this to be approved and move into proposed
[13:18] <diwic> ?
[13:20] <stgraber> diwic: 12.04 has been frozen for the past couple of weeks in preparation for 12.04.3, and AFAIK that SRU wasn't a blocker for that
[13:20] <stgraber> diwic: I'd expect things to get back to normal next week
[13:20] <diwic> stgraber, no, not really, even if it would be nice to have it in 12.04.3, it isn't really a blocker
[13:20] <jdstrand> tkamppeter: it is at the top of sarnold's list
[13:21] <diwic> stgraber, it's more a blocker for 12.04.4, but I hope you'll have it through by then ;-)
[13:22] <stgraber> we definitely should ;)
[13:22] <diwic> thanks.
[13:24] <smartboyhw> stgraber, slangasek if we found a mandatory testcase failing, we can't release that image right?
[13:27] <stgraber> smartboyhw: it's usually up to the product owner
[13:27] <smartboyhw> stgraber, OK.
[13:28] <superm1> infinity: things arent' looking that good.  it look like 3.8 caused a problem with our installer https://bugs.launchpad.net/ubuntu/+source/ubiquity/+bug/1215440 so i'll need to investigate
[13:28] <ubot2`> Launchpad bug 1215440 in ubiquity (Ubuntu) "Mythbuntu 12.04.03 candidate install fails" [Undecided,New]
[13:31] <smartboyhw> stgraber, then, would it be possible to fix that failing bug if it is found in debian-installer?
[13:31] <smartboyhw> I don't know about d-i, but you guys most certainly do.
[13:32] <xnox> do you have a bug #?
[13:32] <smartboyhw> xnox, Bug 1215418
[13:32] <ubot2`> Launchpad bug 1215418 in debian-installer (Ubuntu) "Guided install with KVM and encryption failed in Kubuntu 12.04.3 pre-release alternate image" [Undecided,Confirmed] https://launchpad.net/bugs/1215418
[13:35] <stgraber> superm1: can you get /var/log/upstart/* on that system to see why mysql failed to start?
[13:36] <superm1> stgraber: it's actually during install that it failed to start in it's bootstrap environment.  i think it is because we had a workaround for bug 663069 in it but that bug has been fixed by 3.8 now
[13:36] <ubot2`> Launchpad bug 663069 in linux (Ubuntu) ""non-accessable symlink" errors when using aufs-shaddowed read-only root filesystem" [Undecided,Fix released] https://launchpad.net/bugs/663069
[13:41] <superm1> er maybe not, that should be removed but not hte root failure
[13:41] <superm1> i'll see if i can get the upstart log
[14:17] <xnox> stgraber: infinity: seems like kubuntu-alternate is a bit mismatched, upon install I get "no packages matching running kernel 3.2.0-52-generic in archive" and in the cdrom pool there seems to be two versions: linux-meta at 3.2.0.52.62 & linux at 3.2.0-52.78, no idea if it matters.
[14:19] <xnox> udebs are 3.2.0-37.58
[14:19] <xnox> uname -a is 3.2.0-52-generic
[14:19] <xnox> slangasek: ^
[14:19] <smartboyhw> xnox, that's not jjust Kubuntu-alternate, it's Xubuntu-alternate too
[14:20] <smartboyhw> I think...
[14:20] <smartboyhw> Since phillw is testing that and he found the bug there
[14:26] <xnox> stgraber: infinity: seems like ubuntu switched to enabled stack, and before kernel bumps were done in "ubuntu" seed and inherited "for-free", which is not the case for Kubuntu/Xubuntu alternates anymore since they got a static copies of boot/installer. I'm guessing a kernel bump is needed which i'm not sure how to do.
[14:26] <xnox> ( update seed, reupload meta, rebuild the cd?! )
[14:28] <phillw> yikes!
[14:28] <smartboyhw> yikes~!
[14:28] <phillw> xnox: should I abort testing until re-spin?
[14:29] <xnox> well, that "only" affects alternate images, the desktop images should be all fine. Also images that use enablement stack are not affected (ubuntu/edubuntu).
[14:29] <xnox> which boils down to kubuntu/xubuntu alternates, it's not a bug in d-i.
[14:29] <xnox> so not respinning the "world"
[14:30] <phillw> xnox: that's fine. I'll zsync up the xubuntu-alterante-amd64 if it is going to be respun to sort out the kernel mismatch?
[14:31] <smartboyhw> xnox, can you fix the encryption bug at the same time?
[15:27] <rtg> sarnold, infinity, jdstrand: uploaded apparmor according to https://bugs.launchpad.net/ubuntu/+source/apparmor/+bug/1214979/comments/6
[15:28] <ubot2`> Launchpad bug 1214979 in apparmor (Ubuntu Precise) "Feature buffer full in precise with LTS kernel" [Undecided,In progress]
[15:48] <phillw> xnox: are the alternates for k/xubuntu going to be respun?
[15:50] <xnox> phillw: how would i know. (a) i'm not product driver (b) i can't respin anything
[15:50] <phillw> okies :)
[15:51] <smartboyhw> phillw, I think knome and Riddell can request for re-spins
[15:54] <Riddell> except I can never remember how :)
[15:54] <smartboyhw> Duh
[15:54] <smartboyhw> Riddell, in the ISO QA Tracker
[15:54] <smartboyhw> Tick the Kubuntu alternates
[15:54] <smartboyhw> At the bottom you see "Request for rebuild"
[15:54] <Riddell> got it thanks
[15:55] <Riddell> kubuntu and xubuntu alternates rebuilding
[15:55] <knome> Riddell, thanks :)
[15:59] <phillw> Riddell: thanks, I can only do lubuntu ones :)
[16:00]  * smartboyhw can only do Studio and Kylin...
[16:04] <xnox> Riddell: from the build log looks better.
[16:04] <xdatap1> Hi guys, do you know if Ubuntu Desktop i386 and amd64 is going to respin again before the 12.04.3 release?
[16:55] <Riddell> I need to go out for about 3 hours, back about 2000UTC
[17:03] <infinity> xdatap1: Almost certainly not, unless there's some massive showstopper.
[17:04] <xdatap1> infinity, thanks
[17:05] <infinity> xdatap1: slangasek would know for sure, though, this point release is his baby.
[17:07] <xdatap1> slangasek, Hi, do you know if Ubuntu Desktop i386 and amd64 is going to respin again before the 12.04.3 release? I'm in charge of the italian loco team image and waiting to build the final one
[17:07] <xnox> infinity: i think i successfully managed to bump kernel installer seeds for kubuntu & xubuntu alternate images. Since they are not using enablement kernels, they need to be bumped individually on precise now that they don't inherit them from platform seeds.
[17:07] <xnox> infinity: not sure who should do that in the future, but shotgun not me.
[17:10] <infinity> xnox: I should have caught it, I didn't think about the fact that they had their own installer seed due to not using enablement kernels.
[17:10] <xnox> infinity: ok. well they have been respun and look good now.
[17:10] <infinity> xnox: The proper solution is likely to just move them to enablement kernels before .4 and clean up that mess, but if they'd prefer not to, I just need to remember they're unique snowflakes. :P
[17:10] <xnox> =))))
[17:10] <xnox> speaking of unique snowflakes.
[17:11] <infinity> (I know kubuntu wanted to, it's just that no one's done the work to make it so, I have on idea what xubuntu wanted to do)
[17:11] <infinity> s/on/no/
[17:11] <xnox> I've updated wubi to stop using disk-preinstalls (which have not been build since 12.04.0), moved the metalinks to 12.04.3, tested on Windows7 and now waiting for IS to sign the new wubi.exe.
[17:11] <knome> i have no objections on that, as long as it works
[17:11] <infinity> xnox: Anyhow, thanks for hunting it down and fixing while slangasek and I were asleep. :)
[17:12] <xnox> infinity: hehe.
[17:12] <xnox> infinity: so moving xubuntu should be easy, but kubuntu is using pae gernels which don't exist in enablement stacks....
[17:12] <knome> i think we use pae for 12.04 as well
[17:12]  * xnox giggles at gernels, I meant kernels
[17:12] <knome> if that's what you're referring to
[17:13] <infinity> xnox: Yeah, there would be a few options there.  We could just do a hard cutover and say "if you want -pae kernels install <= 12.04.3), or we could use enablement only on amd64 images (ew) for secure boot.
[17:14] <xnox> knome: not sure, kubuntu seed has "-generic-pae" and xubuntu seed has "-generic"
[17:14] <infinity> xnox: But it's really up to the flavours in question to twiddle that, fix their installer support as necessary, and move themselves forward, not much for ubuntu folks to do there, except advise and review.
[17:14] <xnox> infinity: yeah, agree. I looked into this to buy time, and nobody else seemed to be aroundish
[17:14] <infinity> xnox: For xubuntu, it should be simple enough, given it's the GTK ubiquity, kubuntu might need to put a bit more effort in, if we failed to abstract the SB stuff well enough.
[17:14] <knome> if it doesn't matter if xubuntu uses what it uses in 12.04 now, let's not change that
[17:15] <infinity> knome: It's not world-ending if you stay on 3.2 forever, no, and it's just a matter of me remembering you have a different installer seed. :)
[17:15] <knome> :)
[17:15] <infinity> knome: Not moving to enablement kernels means you can't install xubuntu/precise on secureboot machines, but if your users are the types who'd be happy to use Q/R instead for new amd64 hardware (or mess with their BIOSes), it's pretty much a non-issue.
[17:16] <xnox> knome: sure, it's just ~xubuntu-dev should watch for security&updates uploads of the original ubuntu kernel & bump the ABI numbers when that happens in the seeds.
[17:16] <knome> i would say it's a non-issue, nobody told they need to do that:)
[17:16] <knome> xnox, right... so it's either that or move to the new enablement stuff?
[17:16] <infinity> xnox: I tend to handle all those seed changes, I don't expect them to track them.
[17:16] <xnox> ok.
[17:16] <infinity> xnox: I'll just remember next time. :P
[17:16] <knome> okay. infinity, thanks :)
[17:16] <xnox> infinity: i shall remind you ;-)
[17:17] <infinity> (I don't think I ever knew they had a forked installer seed, since it was Colin who did this all for 12.04.2)
[17:17] <infinity> And knowledge is power.  Or something.
[17:17] <infinity> So a few more of us know now. :)
[17:17] <knome> secrets unraveling!
[17:19] <xnox> infinity: can you add hint to ignore autopkg results for ubiquity 2.15.14, it's stuck "RUNNING" even though it clearly finished and passed.
[17:19] <infinity> I have better "never have to touch the seeds again" plans for kernel/d-i syncing half done right now in S, and by 14.04, it should be hands-off, so this really is only an issue for one more precise point release and then we can stop faffing about with it, which will be pleasant.
[17:19] <phillw> bug 1215453 can be marked as working, by xubuntu is in the final stages and that was an early pop-up alert :)
[17:19] <ubot2`> Launchpad bug 1215453 in xubuntu-meta (Ubuntu) "No Kernel Modules were found" [Undecided,Fix released] https://launchpad.net/bugs/1215453
[17:20] <phillw> s/by/my
[17:21] <infinity> xnox: Yeahp, can do.  Inconvenient to have both jibel and cjwatson on vacation at the same time, when they seem to be the only two people who grasp how this all fits together.
[17:21] <xnox> indeed =)
[17:21] <knome> vacation? cool.
[17:21] <phillw> I'll try bug 1215418 next.
[17:21] <ubot2`> Launchpad bug 1215418 in debian-installer (Ubuntu) "Guided install with KVM and encryption failed in Kubuntu 12.04.3 pre-release alternate image" [Undecided,Fix released] https://launchpad.net/bugs/1215418
[17:22] <xnox> infinity: actually vacations is a good thing to test bus-factor.
[17:23] <infinity> xnox: Generally, I agree, sucks when you're testing the shared knowledge of a brand new system/process, though, cause it takes a while to filter that stuff around.
[17:23] <infinity> xnox: Anyhow, skip hint committed.
[17:24] <xnox> thanks =) !
[17:45] <bdmurray> slangasek: do you have any opinion on the verification of bug 1003296?  Its rather weak.
[17:45] <ubot2`> Launchpad bug 1003296 in samba (Ubuntu Precise) "lightdm crashed with SIGSEGV in _pam_winbind_change_pwd() when password is expiring" [High,Fix committed] https://launchpad.net/bugs/1003296
[18:15] <slangasek> stgraber, xnox: do you know why the kubuntu precise amd64+mac rebuild is "pending" instead of "current"?
[18:16] <slangasek> xnox: erm, what's this about wubi not using disk preinstalls?  I thought we were meant to be using those exclusively
[18:18] <slangasek> xnox: furthermore, there *are* 12.04.2 tarballs on release.u.c
[18:18] <slangasek> precise/ubuntu-12.04.2-wubi-i386.tar.xz
[18:19] <bdmurray> Is the release of verified Precise SRUs being held back for the point release?
[18:19] <stgraber> slangasek: where are you seeing amd64+mac kubuntu precise?
[18:20] <slangasek> stgraber: kubuntu/precise/daily/20130822//precise-alternate-amd64+mac.iso; whereas the current/ dir symlinks to the 20130820.2 version
[18:20] <slangasek> bdmurray: yes, please
[18:21] <stgraber> slangasek: that'd be because they triggered a rebuild on the tracker and only selected amd64 and i386 so only those two got rebuilt
[18:21] <slangasek> bdmurray: 1003296> I wouldn't consider that a proper verification
[18:22] <slangasek> stgraber: erm, confusing.  So the image in the 20130822 directory is also the one from 21030820.2, sigh
[18:22] <stgraber> slangasek: yep, that's what happens with partial rebuilds...
[18:22] <slangasek> stgraber: if you're on the page already, could you kick a rebuild of amd64+mac?
[18:23] <stgraber> slangasek: I can but why do we care if they don't intend to relesae it for 12.04.3?
[18:23] <stgraber> if they do intend to release it, it should be on the manifest and added to the 12.04.3 milestone or it won't be picked up by publish-release
[18:24] <slangasek> stgraber: ah, if it's not in the manifest then fine
[18:24] <stgraber> all good then
[18:25] <slangasek> still, it's very confusing to have the image copied forwarded to $serial+1, have pending symlinked to $serial+1, and have current directory created with symlinks to a mixture of $serial and $serial+1
[18:30] <stgraber> yep, it can be confusing. I think the tracker is doing the right thing as it asks for rebuilds of what the user selected, the current symlinks also do what you'd expect (point to the latest built version for the given file), it's just the whole carying stuff forward that's a bit odd and we may be able to drop now
[18:31] <stgraber> (as people should be using /current or know to go look at a previous build if a given file isn't there)
[18:31] <slangasek> yeah... let's not rush to judgement while in the middle of a milestone release, though
[18:32] <stgraber> sure, it's the kind of change that cjwatson probably has an opinion on and that should be discussed with the rest of the usual release/cdimage people
[18:32] <slangasek> xnox: so for the record, we have wubi preinstall candidates: http://cdimage.ubuntu.com/precise/wubi/20130820/ - I think your change to drop disk preinstall support needs to be reverted.
[18:33] <slangasek> xnox: I'm poking the RT to ask IS to hold off
[18:33] <stgraber> slangasek: should I get wubi back on the manifest and get those published to the tracker?
[18:33] <slangasek> stgraber: yes please (why/when was wubi removed from the manifest?)
[18:33] <slangasek> strange, I can't view the RT
[18:35] <stgraber> not sure why and no idea when. I didn't tweak the manifest for 12.04.3 so I'm assuming we inherited it from 12.04.2 unless someone did some cleanup in between without actually thinking about wubi
[18:35] <slangasek> ok
[18:47] <infinity> stgraber: With current now doing vaguely the right thing, I'd be in favor of serial directories only containing images that belonged to that build set.  Definitely would want Colin's input too, though.
[18:48] <infinity> (And, of course, that change would make current stop doing the right thing again, so that would need a bit of munging :P)
[18:53] <xnox> slangasek: there were none on cdimage, are there tarballs for 12.04.3?
[18:53] <xnox> slangasek: http://cdimages.ubuntu.com/precise/wubi/current/ last modified 16-march-2013
[18:54] <slangasek> xnox: yes, http://cdimage.ubuntu.com/precise/wubi/20130820/ + http://releases.ubuntu.com/precise/ubuntu-12.04.2-wubi-amd64.tar.xz
[18:54] <slangasek> xnox: can you please revert that bit quickly?
[18:54] <xnox> sure.
[18:54] <slangasek> xnox: thanks :)
[18:59] <xnox> slangasek: one sec checking something.
[19:01] <xnox> slangasek: that tarball (i386.tar.xz) says 12.04.2 in /etc/lsb-release, whilst normal desktop.iso has 12.04.3
[19:02] <slangasek> xnox: so I guess we should respin the tarballs themselves
[19:02] <slangasek> not sure why that wasn't done already, but kicking off a build right now
[19:02] <slangasek> ok?
[19:02] <xnox> slangasek: ok. thanks.
[19:02] <infinity> Should be DIST=precise SUBPROJECT=wubi for-project ubuntu cron.daily-live --live
[19:03] <infinity> Maybe you missed it in your last run?
[19:03] <infinity> Or maybe there's no way to do it from the tracker, and you're not driving manually?
[19:03] <xnox> yeah, tracker doesn't list wubi as far as I can see.
[19:15] <slangasek> infinity: I was driving from the console; and I did do a wubi build, so dunno why it had the right contents
[19:16] <xnox> slangasek: wubi build updated, emailed RT as well.
[19:20] <slangasek> xnox: thanks!
[19:22] <xnox> oh and wubi back in the manifest, thanks stgraber.
[19:22]  * xnox really should read scrollback first, then talk.
[19:23] <xnox> instead of just highlights.
[19:28] <slangasek> :-)
[19:28] <phillw> does wubi need testing? I've got a loan XP machine I can use, but my download speed would be ~3 hours to grab an entire ISO
[19:31] <phillw> can wubi use an iso on CD?
[19:32] <xnox> phillw: we want to test the tarballs, not the iso. it can, if you copy wubi off the cd and launch it from the desktop.
[19:32] <xnox> (whilst having the cd inserted)
[19:33] <xnox> phillw: it's ok, i can test fast in windows7 VM
[19:33] <phillw> xnox: okies :)
[19:40] <rtg> slangasek, linux-goldfish uploaded ^^
[19:40] <xnox> rtg: \o/ awesome
[19:41] <xnox> rtg: armhf & i386 ?
[19:41] <rtg> xnox, I haven't been able to test it, so lemme know if it works. armhf only
[19:41] <rtg> are you wanting i386 as well ?
[19:41] <xnox> rtg: ok. cool. Once it's build, I'll test it and will let you know how it goes.
[19:42] <xnox> rtg: I don't want i386, but plenty of people are thinking about using i386 in the emulator..... we don't have ubuntu-touch-rootfs for i368 yet, so it's of lower priority at the moment.
[19:43] <rtg> xnox, ok, I'll wait until someone hollers (or bribes me)
[19:43] <xnox> rtg: ;-) good plan
[19:55]  * xnox giggles "Windows 7" has turned my desktop black with a sign "This copy of Windows is not genuine" because I opted to not activate it online.  Yet it still offers to download updates =)
[19:56] <sarnold> xnox: before long it'll start turning off every few hours
[19:56] <xnox> sarnold: nice. Well I snapshot my install, cause spending ~40 minutes installing + applying updates is a pain.
[19:56] <sarnold> xnox: good plan :)
[19:56]  * xnox ubuntu installs in <<8minutes from boot to final login screen on the same VM setup
[21:41] <xnox> slangasek: infinity: something odd is with wubi build.
[21:41] <xnox> http://cdimages.ubuntu.com/precise/wubi/20130604/MD5SUMS
[21:41] <xnox> http://cdimages.ubuntu.com/precise/wubi/20130822/MD5SUMS
[21:42] <xnox> both are identical.
[21:42] <xnox> slangasek: verified that updated wubi.exe downloads & does tarball based install.
[21:48] <slangasek> xnox: so nusakan is lying about having built it? hmph
[21:48] <slangasek> xnox: anyway, thanks for the verification that wubi.exe itself works; now let's look at some logs
[21:49] <slangasek> hmm, livefs build dispatched successfully, then the download failed
[21:50]  * xnox was under impression that wubi tarball builds were killed with fire and we switched / support cd-rom based builds only..... but then that was for post-precise. no idea what happened with precise.
[21:50] <slangasek> xnox: yes, killed with fire in that we didn't want to continue supporting it as a separate build time going forward and wanted to downplay wubi as an install method... but for precise it's meant to still work
[22:03] <slangasek> http://people.canonical.com/~ubuntu-archive/livefs-build-logs/precise/ubuntu-wubi/20130822/livecd-20130822-amd64.out looks quite complete and correct
[22:06] <xnox> agree
[22:08] <slangasek> the target directories on the livefs builders likewise have correct contents
[22:09] <slangasek> so something's going pearshaped on nusakan itself
[23:26] <slangasek> xnox: wubi fixed (manually); now to figure out the source of this bug
[23:35] <Riddell> kubuntu images all good for me (although not tested wubi)