[00:01] skaet: the upshot is that neither of these packages should be on the DVD at all; but the best way to get them off of there is probably to remove them from the archive altogether [00:02] slangasek, thanks for figuring it out. [00:24] skaet: ok, those binaries are all cleaned up now [00:33] slangasek, thank you. :) === Ursinha_ is now known as Guest63174 [04:16] pitti: BTW, i'll be in late.. burning the candle here. [04:44] Good morning [04:45] Daviey: urgh, that looks like a complete nightshift [05:29] pitti, Daviey - can you get with jibel when he comes on line and decide if its worth respinning Ubuntu/Kubuntu DVD? I've release noted the issue for now (see linux-backports-module issue in the backscroll) and can live with it, but wanted your thoughts if it was worth a respin. [05:31] due to the oversize? [05:31] oh, ubuntu, too? [05:31] hm, it's not oversized [05:32] pitti, DVD - http://cdimage.ubuntu.com/lucid/dvd/20120214/report.html [05:32] we don't have a meta fix, though [05:33] so a respin alone wouldn't cure it [05:34] slangasek thought cleaning up the archive should sort it out, but if we need that meta fix, yeah, leave it as a release note. [05:35] skaet: yeah, it's a wart, but not really visible to users IMHO [05:36] release note it is then. [06:11] pitti: they were NBS packages on the archs; I've dropped them from the archive now, so a respin would drop them from the DVD as well [06:11] slangasek: hm, how so? linux-backports-modules-wireless-lucid-generic-pae and linux-backports-modules-wireless-lucid-server are still built by current linux-meta [06:12] not on those archs [06:12] oh, I see [06:49] pitti, https://wiki.ubuntu.com/LucidLynx/ReleaseNotes/ChangeSummary/10.04.4 - any review/polishing you can fit in, would be welcome (the linux part was duplicate crazines. :P - I'm still spotting some duplicates, but too tired now, zzz time.) [06:50] skaet: thanks! will do over the day [06:50] skaet: sleep well! [06:51] thanks! === jodh is now known as jhunt [08:42] jibel: hey [08:43] jibel: the (k)ubuntu DVDs have two uninstallable packages from linux-meta on them; they don't really hurt, just show up in report.html; do you think it's worth rebuilding/retesting for this? [08:43] jibel: 10.04.4 I mean [08:47] pitti, what time is the release ? if it's early afternoon, that's too short. if it's end of the afternoon that's fine we can retest Ubuntu. [08:48] pitti, what is the impact if these packages are kept on DVDs ? [08:48] jibel: our evening most probably; but I'd hate to waste any human effort on retesting on that really [08:48] jibel: not much really; they are uninstallable if someone tries to install them from the DVD pool [08:49] but they don't show up in software-center, or are advertised anywhere [09:15] pitti, ok, it is very low impact and is release noted. If really we want to respin we can test DVDs before EOD today but I'd rather spend time on something else than retesting these images. [09:15] jibel: yes, what I thought; thanks for confirming [11:00] pitti: heya [11:01] hey Daviey [11:01] Daviey: got some hours of sleep? I woke up about 30 minutes after you left.. [11:01] pitti: heh. happy times :) [11:01] I just went over https://wiki.ubuntu.com/LucidLynx/ReleaseNotes/ChangeSummary/10.04.4 to beautify it a bit [11:01] pitti: Sorry, for not being here first thing. What is the current status? [11:02] Daviey: we noticed that the DVDs have an uninstallable package each, but IMHO that's harmless enough to not warrant a respin [11:03] Daviey: testing looks good, including Kubuntu [11:03] so please pre-publish if you haven't already [11:03] Daviey: no other good or bad news AFAICS, looks on track [11:03] pitti: the kernel module thing? [11:03] on DVD [11:03] Daviey: yes; they got removed from the archive, but are still on the DVD [11:03] but as it's mostly a cosmetical thing (reports.html), it doesn't really warrant a respin [11:03] as that would void the testing [11:03] pitti: right.. i spotted that last night :) [11:19] pitti: so, i would have expected pre-publish (now done) to create http://cdimage.ubuntu.com/releases/10.04.4/ .. but it hasn't? [11:19] Daviey: no, it's not supposed to [11:20] Daviey: it just adds the new images to a hidden directory, so that mirrors can grab them [11:20] Daviey: adding .html, those dirs, etc. is the actual publishing [11:20] pitti: where does it create it? [11:21] ah, found it [11:23] pitti: --prepublish didn't advice of dvd commands.. is that expected? [11:25] Daviey: yes; prepublish is only for stuff that lands on releases.ubuntu.com [11:25] ahhh [11:25] cool [11:25] Daviey: DVDs, source, derivatives etc. go to cdimage.ubuntu.com which doesn't need prepublishing [11:25] mirroring there is just to our own DC, which is fast [11:25] right, makes sense. [11:26] lunch, bbl [11:46] context menus are often not drawing immediately, is this an issue with compiz 5.4 (under testing)? [12:13] * lamont gets started on that armel downtime window [12:51] lamont: \o/ [12:51] lamont: And, to be clear, we're just knocking out the babbages for now, right? The beagles are staying in place for a bit? [12:52] lamont: (We want the beagles gone at some point too, but just checking what's going on today) [12:53] correct [12:53] we are decommissioning 10 of 13 bbg3 boards [12:53] er, 10 of 15 [12:54] will they get replacements? [12:54] asking because even with the babbages the current queue is 8 days [12:55] oh, sorry, that's already with them taken down [12:55] it was a couple of hours this morning, but constantly non-zero over the last week [12:57] pitti: They're being replaced with Pandas as we speak. [12:57] \o/ [12:58] (And I'll immediately steal a few to finish the armhf rebuild test faster) [12:58] But, like, 4 Pandas should clear up the armel queue in no time. :P [12:59] oh darn, the mrpt build was cancelled [12:59] it already built for 1.5 days and was almost ready :/ [12:59] but *shrug* [12:59] Shrug indeed. [12:59] the pandas should be faster indeed [13:09] lamont: Is caph manual for a reason, or was that an oops? [13:12] that's either not my doing, or an oops [13:12] . o O { what would it take for you guys to accidentally spill a cup of coffee or two into the powerpc buildds? } [13:14] red bull might make them work faster. [13:14] https://rt.admin.canonical.com/Ticket/Display.html?id=48569 made some progress recently ... [13:14] nobody loves ppc :( [13:14] * Laney strokes the old ibook [13:18] lamont: Alright. I'm going to give caph to armel for now, though. [13:18] s/though/then/ [13:19] sure [13:19] infinity: I'm going to assume that you're balancing, and ignore irc for a bit [13:20] cjwatson: I'm actually more eager to get rid of it :) (livefs builds are not much of a concern as long as the bloody thing takes 3 days to get to a package build) [13:20] lamont: Ignore away. ;) [13:21] anyway, it was just trolling, sorry [13:21] pitti: When livefs builds move to sulfur, we get royal as an lp-buildd. [13:21] pitti: (And then when livefs builds move to launchpad, we get sulfur too) [13:21] infinity: still doesn't change the fact that nobody cares about powerpc specific failures [13:22] pitti: I do! [13:22] pitti: And BenC did a mess of uploads. [13:22] pitti: Just because no one's paid to care doesn't mean no one cares. [13:23] ah, so I'll let you worry about build failures like vlc :) [13:23] more seriously, I actually don't want anyone in platform to waste time on it [13:23] we frankly have more urgent things to spend our time on [13:23] if someone actually uses it as a platform and wants to, sure [13:24] but for my part I refuse to spend any more time on it as part of stable+1 [13:24] I'm wasting enough attention on it already [13:25] To be fair, with the exception of really bizarre toolchain issues, if people cared about the packages they upload, it wouldn't ever be much of a mess. :/ [13:25] (And while I realise it's hard to care about powerpc, I could say the same about arm, which we supposedly support) [13:25] arm is way worse off than powerpc in termsof build failures. [13:26] *terms of [13:26] Yeahp. [13:26] like, two or three times as bad. [13:26] yes, but we support it and are interested in the platform, unlike powerpc [13:27] That's sort of my point. [13:27] We "support" it, and we still can't get people outside the ARM team to care much about failures in packages they upload. :P [13:27] and people do test arm images, whereas nobody tests the powerpc ones [13:27] Images, I care less about on PPC. [13:28] But I will commit a fix to debian-cd soon so that PPC images actually boot on one of my machines. [13:28] (Currently, it was installed with a Debian CD, and then cross-debootstrapped to Ubuntu) [13:28] I love the way those nobodies who test the powerpc ones also file bugs [13:28] They just don't *sign up as testers* [13:29] (Which I wish they would and I do occasionally ask ...) [13:29] I'll be sure to actually submit results to the ISO tracker once the ISOs work for me. :) [13:29] well, *shrug*, I still consider it a waste of time, but that's just my personal opinion [13:29] (er, for anyone reading the IRC logs out of context, that was gentle sarcasm, not implying that the people who test powerpc images are "nobodies") [13:30] Up until now, I've had one machine that boots with BootX (clearly can't do installer testing), and one that could work, but needs a tweak. [13:30] pitti: Let others waste their time on it, then. ;) [13:30] pitti: I have more PPC machines in my house than x86. [13:39] cjwatson, a tester just reported bug 933434 in kubuntu 10.04.4 [13:39] I don't think it's new but a consequence of bug 933433 with grub and XFS [13:39] could you look into it ? [13:40] well, certainly isn't grub [13:40] far too early [13:40] is it .4-critical? [13:41] oh - my apologies, I should have read the entire syslog first! [13:41] 933433 is grub, 933434 is not [13:42] the latest comment on 933433 explains the situation accurately === doko_ is now known as doko [14:46] skaet: http://people.canonical.com/~ubuntu-archive/testing-ports/precise_probs.html [15:04] cjwatson: for bug 933433 you set to triaged, does that mean you know the cause? [15:10] what boot option do I add to get oem mode? (bug 645818 means I can't use normal boot loader) [15:15] Riddell: the reporter already explained the cause in his most recent comment [15:15] you can't install grub to a logical partition containing xfs; the problem is just that the installer doesn't tell you this but crashes instead [15:16] OEM mode is oem-config/enable=true [15:18] cjwatson: ok so not a regression then [15:22] indeed not [15:24] Riddell: don't suppose you could have a look at bug 933434? it's in a bit of the KDE UI I'm really not very familiar with [15:24] no idea what an _ElementInterface is [15:26] cjwatson: not off the top of my head, I'll try and test and see if I can recreate and look into it if I can [15:37] skaet, hey, can we have the move to armhf mentioned in the FF announcement ? [15:38] ogra_, give me the words you want and I'll add it in. :) Decision's been made then? [15:38] yes [15:39] any email/announce somewhere already I can crib from? [15:40] infinity, lamont: could we borrow some armhf builders for armel to make it catch up with the long build queue? [15:40] armhf only has the autotest [15:41] skaet, "with the feature freeze of ubuntu precise ubuntu-arm moves to armhf as its default architecture, the armel architecture will go on to exist but not be supported by canonical anymore. image builds will be switched to armhf across the board with the exception of omap3 and ac100 builds still keeping armel builds around for compatibility with their respective binary drivers" [15:41] well, not "only" of course [15:41] but "also" [15:41] infinity, ^^^ does that look ok to you ? [15:41] skaet, we only announced it in the IRC meeting yet, the addition to the FF announcement would be the most official one then :) [15:41] pitti:I was letting infinity do the balancing act - I'm close to throwing 5 or more pandas into the pool, with a total of 9 arriving today [15:42] lamont: ah, that sounds sufficient indeed, thanks! [15:43] pitti: Yeah, I was waiting for the new pandas to be online, but I can rebalance a couple right now. [15:43] infinity: oh, that's ok; I thought we already had all the new ones, and some were moved to armhf [15:44] ogra_ coolio. THanks. [15:44] * pitti is just eager to get some buildable armel images soon again [15:44] ogra_: Other than the complete lack of capitalisation, and the run-on sentence, sure. :P [15:44] ogra_: I'm sure skaet can un-German it a bit. ;) [15:44] infinity, pfft ... grammar, who needs that [15:44] just _DO NOT_ unmanual the new ones [15:45] lamont: Set them to disabled instead of manual. [15:45] :) [15:45] lamont: understood [15:46] There. [15:46] "ishigaq" /me tries to pronounce that in Klingon [15:47] "today is a good day to build"? [15:47] Hahaha. [15:51] ogra_, infinity - I'm happy to mention in the FF announce, but it probably deserves its own announce and searchable subject line ;) with more details (and rallying of folks to help with cleaning up the uninstallable packages, and things not building, wouldn't hurt either... ) Maybe someone on the arm team should issue that later today/tomorrow? [15:51] k [15:51] * ogra_ will care for it [15:51] thanks ogra_ :) [16:01] let's kick this pig, shall we infinity? [16:10] lamont: Once I'm off the phone [16:11] infinity: well, anyway, you have 5 new panda builders building with a track record on ishigaq of success or explainable failures. I declare victory [16:11] lamont: Or you can kick it. :) [16:11] your other 2 will be online sometime soonish, with 2 others being virtual [16:11] #10 is that special board that thinks it has no SD card even though it booted from it [16:12] * infinity nods. [16:12] and that was a tad more painful than I wanted it to be [16:29] does lucid work with alternate installs on netbooks? I get "no common CD drive was detected" [16:39] infinity: so... sycamore... is that still being used for livefs builds? [16:40] lamont: Is it a babbage? [16:40] it is [16:40] so I want to kill it, of course. [16:40] OTOH, that requires a tad bit more coordination [16:41] lamont: It's used, but I can make that not true in pretty short order. [16:41] and would your preference be to replace it with a beaglexm, or a panda? either way you lose one [16:41] A beagle. [16:41] But let me look at things. [16:42] We're dropping two armel images today, so I can probably just lose the builder and shuffle and no one will notice. :P [16:42] ah, that'd be lovely [16:43] if we still need one, I choose adoxaceae [16:43] it pairs most nicely with annonaceae [16:43] Heh. [16:43] How many beagles do we have? [16:44] I don't really want to create extra work for you, but the beagles are less than stellar for package bulding anyway. [16:44] With 6 image targets, we could have 6 beagles. :P [16:44] And make celbalrai an lp-buildd. [16:53] Daviey, ping? [16:57] skaet: hola [17:01] infinity: *aceae are beagles in the buildd world [17:06] lamont: Aye, I knew that much. [17:09] and I think that may be all of them, when you count annonaceae [17:13] Mmkay. [18:31] ev: not still around are you? :) [18:32] Daviey: that depends on what you want me to do [18:32] ev: is there a refreshed wubi for 10.04.4? [18:33] Daviey: yes, it's ready and waiting at http://people.canonical.com/~evand/wubi/lucid/stable [18:34] ev: super, thanks [18:34] when 10.04.4 CDs are spun, that will be automatically included [18:34] sure thing === chrisccoulson_ is now known as chrisccoulson === medberry is now known as med_ === ChanServ changed the topic of #ubuntu-release to: 10.04.4 Released! Precise Feature Freeze in effect. Archive:open | http://pad.ubuntu.com/ubuntu-release | Precise Pangolin Release Coordination. Please don't upload things during freezes where you shouldn't, or be prepared to apologise to the release team | we accept payment in cash, check or chocolate covered ants | melior malum quod cognoscis [22:36] Thank you Daviey! 10.04.4 images are all available for download. :) [22:37] Thank you jibel and Riddell for coordinating the testing and checking that the images were good to send out. :) [22:37] skaet: sorry, NM got uploaded later than freeze time, I got confused by conversion and was still testing it up (and enabling the tests) [22:38] and somehow it fails to build, oh god [22:38] cyphermox, as long as it doesn't cause the daily build to break... [22:39] not really worried, it's an update with some bug fixes [22:39] cyphermox, ok. as long as its fixed before the cron jobs trigger all should be good. [22:39] cyphermox: at least the ftbfs seems easy to fix (missing build-dep) [22:39] yeah [22:40] I should really have caught this early though [22:49] on the bright side if I didn't screw up elsewhere, shutdown will now no longer take a minute to complete [23:42] slangasek, pitti, cjwatson, Daviey, stgraber, updated default_milestone on nusakan to "default_milestone=Precise Daily" === bladernr_ is now known as bladernr_afk [23:55] * ScottK approved the first FFe of the cycle .... [23:56] congrats [23:56] * Laney finds one to deny :-) [23:57] * ScottK cheers Laney on. [23:58] Mine was an easy one to approve because the only reason it didn't get sync'ed pre-FF was it wasn't out of Debian New in time. [23:58] Since we want to encourage going through Debian, easy decision. [23:58] yeah, hyperair was DTRT there [23:59] DTRT?