[00:12] Riddell, are you going to create a https://help.ubuntu.com/community/NattyUpgrades/Kubuntu page for alpha1? [00:15] Riddell, ScottK - can you confirm that http://releases.ubuntu.com/kubuntu/natty/alpha-1/ (Kubuntu) is where the Kubuntu images will be going tomorrow. [00:16] * cjwatson remembers to disable daily CD builds, in the nick of time [00:16] skaet_: early milestones don't go to releases [00:16] skaet_: http://cdimage.ubuntu.com/kubuntu/releases/natty/alpha-1/ [00:16] cjwatson thanks. [00:16] :) [00:16] will edit. [00:19] cjwatson, my connection started acting up, and I lost part of the last hour of IRC traffic, do we have the packaged fix from desktop team? (ie, have the builds started?) [00:19] the publisher's done [00:20] skaet_: it'd be good if you could do the CD builds though - getting kind of late heere [00:20] *here [00:20] I think the sequence we want is: [00:20] buildlive ubuntu daily-live && for-project ubuntu cron.daily-live [00:20] for-project ubuntu cron.daily [00:20] buildlive ubuntu-dvd dvd && for-project ubuntu cron.dvd [00:20] buildlive edubuntu-dvd dvd && for-project edubuntu cron.dvd [00:21] and possibly, if so probably before the DVDs: [00:21] buildlive ubuntu-netbook daily-preinstalled && for-project ubuntu-netbook cron.daily-preinstalled [00:22] understand its late, and appreciate you're staying up so far. [00:23] let me get started with the first step, and then if environment's good, I'll work through them. [00:23] BTW, the desktop fix: https://launchpad.net/ubuntu/+source/unity/3.2.2-0ubuntu2 [00:24] thanks! [00:24] was going to ask about that... ;) [00:24] I'm doing a source DVD build as well, but that shouldn't interfere, can work in parallel [00:25] I think you should assume that ubuntu-netbook preinstalled needs to be rebuilt unless you hear otherwise from #ubuntu-testing, since it does contain unity [00:25] ok [00:28] hmm [00:31] cjwatson screen session resumed from oct, ok. ubuntu daily-live started... can you check looks ok from your point? [00:33] it's running the right commands thus far [00:33] not sure that says very much, but anyway :) [00:34] heh [00:35] will bug some of the other release team members if it looks like its going in the weeds... sleep well. [02:09] ubuntu desktop i386 and amd64 are available on the iso tracker now (20101202) [02:57] ubuntu alternates for i386 and amd64 are availabe on iso tracker now (20101202) [04:30] ubuntu netbook preinstalled omap4 is available on the iso tracker (20101202) [05:50] Good morning [06:06] good morning pitti [06:06] ubuntu dvd for i386 and amd64 have just been posted to iso tracker. [06:07] edubuntu is now building. [06:07] not sure if highvoltage will want to move up to it or not though. [06:08] haven't seen any bugs flagged against his 20101130 runs, but ... could be impacted by the bug fixed this afternoon. not sure. [06:10] pitti, could you keep an eye on the builds on animony, and when the edubuntu finishes, see if highvoltage want's it uploaded to iso tracker or not? [06:10] antimony even... [06:11] * skaet_ is feeling like its time to zzzz... [06:13] skaet_: sure, will do [06:13] skaet_: I guess we can post it either way, and then choose to release it or not [06:13] thanks pitti! [06:13] highvoltage: ^ does that work for you? [06:58] highvoltage, skaet_: ah, since the current edubuntu DVDs are fully tested already, perhaps I sholdn't post them without highvoltage's exlicit ack [08:11] skaet_, highvoltage: http://cdimage.ubuntu.com/edubuntu/dvd/20101202/ is ready; do you want this as a1 candidate, or the already tested 1130 one? [09:56] skaet_, when doing iso testing which desktop option are we meant to be testing [09:56] (as there are now two, and depending on h/w you get a different one) [10:12] skaet_, so armel is blocked by bug 683683 [10:12] Launchpad bug 683683 in klibc (Ubuntu Natty) (and 3 other projects) "run-init on omap3, omap4 in natty dies with "run-init: Unknown error 17718852" (affects: 3) (dups: 1) (heat: 26)" [High,Invalid] https://launchpad.net/bugs/683683 [10:12] (its not klibc but busybox) [11:31] ogra: why busybox? run-init's in klibc [11:32] oh, right, reading the bug now ... [11:32] ogra: any way I can try this under emulation? [11:49] bug 684059 seems a bit of a showstopper for kubuntu, am investigating [11:49] Launchpad bug 684059 in ubiquity (Ubuntu) "Kubuntu Live Session/Installer: NameError: global name 'QWidget' is not defined in /usr/lib/ubiquity/frontend/kde_components/PartAuto.py (affects: 1) (heat: 6)" [Undecided,Confirmed] https://launchpad.net/bugs/684059 [12:18] Riddell: should be http://paste.ubuntu.com/538959/, right? [12:31] cjwatson: yes that's right [12:31] not sure how it managed to work before [12:31] CD installs fine apart from that issue [12:31] cjwatson: can we do an ubiquity upload then? [12:49] please go ahead [13:09] cjwatson, no idea why busybox, but replacing busybox-initramfs makes it work, i doubt you can try that in emulation easily since its in initramfs [13:10] * ogra just did a build without -marm (though that shouldnt do anything) and under the maverick toolchain [13:10] lets see if i can get it booting [13:24] cjwatson, fyi the above boots [13:24] * ogra has to find out if its -marm or the natty toolchain getting in our way here [13:24] will do another rebuild now [13:25] we only use -marm because you said something else broke beforehand :) [13:26] yes, that was thumb2 back in lucid and maverick [13:26] i guess we can drop it today [13:26] but i really doubt it is the issue [13:26] i bet on the toolchain [13:29] ogra_ac: to drop it you should at least test the reboot -f, that didn't work without marm [13:29] I think we should be fine with it, but would be good to test the same bug as before [13:30] sure [13:30] first of all i want run.init to work though [13:30] it's probably a combination with the new toolchain [13:30] worst case jasper can pull reboot from the system in and not use the busybox one [13:31] sure, if it makes the image usable, go for it [13:32] well, i dont count on A1 anymore anyway [13:32] you could compare assembly of run-init [13:33] i didnt see assembly in it [13:33] would be good to build the new busybox with the older toolchain [13:33] to make sure it's related with it [13:34] thats what i did above [13:34] maverick toolchain and dropped -marm [13:34] now i'm doing the same with the natty toolchain [13:34] oh, ok, I thought you just dropped -marm [13:34] and we'll see if its -marm only [13:34] which i doubt [13:38] ogra: I mean generated assembly from the compiler [13:38] although, you said that run-init didn't change, so I guess it's not that [13:39] right [13:39] klibc is the same as mavericks (not even rebuilt) as well as eglibc [13:39] and downgrading to mavericks busybox-initramfs makes everything work fine [13:41] something to do with how it's executing subprocesses? [13:42] likely [13:42] run-init only calls execve() afaik [13:42] and it seems to fail in that [13:42] or not subprocesses in this case, I hope we do 'exec run-init' [13:43] do you have anything resembling a traceback or anything like that? [13:43] note really and its really hard to get it from the point where run-init runs since the calling pid needs to be 1 [13:43] *not [13:44] i think NCommander managed to get an strace by hacking up /init [13:44] but he seeem to not have attached it to the bug (and isnt up yet) [13:44] *seems [14:07] GRRRR ! [14:07] so it *IS* -marm [14:08] boots fine when built with natty toolchain and -marm dropped [14:11] uploaded [14:11] go ahead and upload it with the -marm dropped, then [14:11] ah yes :) [14:11] thanks [14:11] are you going to shoot for revalidating this in time for an alpha 1 release? [14:11] when is ETA for A1 ? [14:11] ASAP [14:12] then just skip it, i think i saw an oem-config crash too when testing (but didnt pay much attention to that one since i was only validating initrd) [14:12] Riddell: I assume you want a Kubuntu desktop rebuild as soon as this publisher run has finished? [14:12] so its inlikely it will survive [14:12] ogra: OK, presumably might as well do a test run to catch as much as possible now? [14:13] yes, but i'm not tied to release A1 if this one doesnt work [14:14] i know there were some ubiquity fixes (skaet said so) and i was testing yesterdays image [14:14] so it might well be fixed already [14:17] cjwatson: yes please (assuming ubiquity is in it) [14:17] pitti: the already tested 1130 one, please [14:17] highvoltage: ok, thanks [14:18] skaet_: ^ FYI [14:18] * highvoltage needs to get up earlier on alpha days :) [14:24] skaet_: I added Edubuntu to https://wiki.ubuntu.com/NattyNarwhal/TechnicalOverview - ideally I would've liked to be more terse, but pretty much all I added there is important [14:26] Riddell: it is, except for armel [14:27] armel is our for alpha 1 for kubuntu for sure [14:27] so can't build the preinstalled ones yet, but I assume that those suffer from the same issue as ogra's [14:27] heh, yeah [14:27] oh yes of course, KDE is hosed there [14:27] right [14:27] sorry, should have known that [14:27] no QT, no KDE [14:27] nah, its enough if Riddell and i know ;) [14:35] Riddell: rebuilding now [14:40] ogra: will you take care / have you taken care of letting management know that alpha-1 is toast on arm? === bjf[afk] is now known as bjf [14:47] cjwatson, i will, if skaet_ is up, david is aware that we likely wont make it [14:48] * skaet_ waves, with cup of coffee in hand. [14:48] dont wave with a cup if there is coffee in it !!!! [14:48] lol [14:48] * skaet_ ...I've spared the keyboard this time [14:53] hey skaet_, good morning [14:53] good afternon, pitti :) [15:17] pitti, thanks for taking care of the Edubuntu last night. :) [15:17] highvoltage, ack - we'll release with the 20101130 ones. [15:17] skaet_: well, I didn't really do anything :) [15:18] pitti, it was sorted out when I woke up - that is something!! :) [15:49] skaet_: if you do the release announcement, would it be possible that to add that the archive will become a bit uninstallable after the alpha for a day or two, because we update the python default to 2.7? [15:51] doko, ack. seems reasonable to add a warning. (and good to get the python default change debugged) might be a good idea to wait a day or two, rather than tackling it immediately on friday, since others are still trying to get their images into basic working condition. We can talk about at release meeting tomorrow? [15:52] are there any images that expect to release late, rather than just slipping to alpha-2? [15:52] hmm, ok, although I'd like to use the buildds on the weekend, so friday night would be good ... [15:53] cjwatson, just thinking it might be nice to get a workable arm image available, before everything breaks again. [15:53] heh, somebody beat me to posting kubuntu desktop [15:54] (got distracted) [15:54] well, we can try to build one now with the fixed busybox [15:54] but ogra expected other things to be broken [15:54] doko, if everyone agrees tomorrow - I'm cool with using the weekend build reseources. [15:54] * cjwatson runs buildlive ubuntu-netbook daily-preinstalled && for-project ubuntu-netbook cron.daily-preinstalled [15:54] but the good thing is that the arm image will at least boot now [15:55] well, we could test at least [15:55] cjwatson, skaet_ ^^^ [15:55] ogra_ac, ack. [15:55] you should have a build in, er, however long it takes [15:55] but i saw an error popping up with the old oem-config/ubiquity which i didnt pay attention to (due to being busy with initrd issues) [15:55] 90min [15:56] ouch [15:56] but the image was a day old [15:56] don't suppose anyone has a traceback? [15:57] nope [15:57] i can check if there is a log on the SD, gimme a sec [16:06] skaet_, bug 684059 is a showstopper for kubuntu, any chance to fix it today ? [16:06] cjwatson, http://paste.ubuntu.com/539026/ [16:06] Launchpad bug 684059 in ubiquity (Ubuntu) "Kubuntu Live Session/Installer: NameError: global name 'QWidget' is not defined in /usr/lib/ubiquity/frontend/kde_components/PartAuto.py (affects: 1) (heat: 6)" [Undecided,Confirmed] https://launchpad.net/bugs/684059 [16:06] Riddell, ^ ? [16:06] jibel: it's fixed [16:08] jibel: 20101202 [16:09] cjwatson, Riddell, ok thanks. [16:09] * cjwatson wonders why Riddell's upload didn't close the bug [16:10] Riddell: ... you know, it's nice if you upload the same thing as is in bzr. :-) [16:10] archive logs say: [16:10] 2010-12-02 13:10:35 DEBUG * Automatic update of included source packages: partman-auto [16:10] 2010-12-02 13:10:35 DEBUG 93ubuntu2. [16:10] bzr says: [16:10] * ubiquity/frontend/kde_components/PartAuto.py fix namespace on QWidget [16:10] LP: #684059 [16:11] the upload has the fix, it's just the changelog that's wrong [16:11] * cjwatson closes the bug by hand [16:12] * ogra_ac wonders what /var/lib/ubiquity/apt-install-direct is [16:14] I'm fixing that bug [16:14] notice that it fails trying to *create* that file - ergo the directory doesn't exist [16:15] ah [16:15] fixed in bzr. want me to upload? [16:15] and i was grepping myself to death :) [16:15] I don't know if that's the only thing [16:15] well, at least iots one down [16:15] perhaps if you try applying that fix by hand? [16:15] r4446 in ubiquity [16:16] * ogra_ac checks [16:16] grrr, need to bzr upgrade first [16:20] sigh ... slooow [16:26] ah, finlly [16:26] *finally [16:37] good night everyone [16:39] good night pitti. [16:39] sigh, arm build failed because ubiquity wasn't in sync [16:40] * cjwatson restarts :-/ [16:40] yeah, just noticed that too [16:40] test install running btw [16:56] cjwatson, happily looking at a 2D netbook screen here, seems the fix is good to upload [16:57] cool [16:57] ok, good, will do [16:58] beyond that natty seems to run really great [17:21] I wonder if ppc will catch up before alpha1 [17:21] we could try building it on arm ;) [17:22] seems to be faster [17:22] ubiquity was what crabapple was building when it died [18:01] ogra_ac: cjwatson: I can also confirm that the arm image worked fine after porting the busybox and ubiquity fixes [18:01] running normal desktop now [18:02] ah, good to know [18:29] ogra_ac: are you testing http://iso.qa.ubuntu.com/qatracker/test/4865 [18:30] marjo_, wont work without manually applying a ubiquity fix [18:30] ogra_ac: ack [18:31] the build before was tested, fixes for both bugs we identified were uploaded [18:31] ogra_ac: ack [18:32] given that we dont want to hold up A1 by re-rolling and re-testing another image we will skip A1 beyond what was tested and fixed [18:32] ogra_ac: makes sense at this time [18:33] we know that the dailies will work from now on, thats enough as a stop gap [18:34] GrueMaster, do you plan to test the above image ? [18:35] (with the manual ubiquity fix) [18:36] Does this image have the busybox fix? [18:38] yes, buut not the ubiquity one [18:39] ok, pulling now. I had thought it was built before the busybox fix was uploaded (based on earlier discussions on the other channel). [18:39] Otherwise I would have tested it already. [18:41] let me check, but i think its the new one [18:41] skaet_, cjwatson: I'm off out for the evening, kubuntu images look good to release, I don't think we'll have any kubuntu release page done for alpha 1 [18:41] GrueMaster, http://cdimage.ubuntu.com/ubuntu-netbook/daily-preinstalled/20101202.1/ should have the fixed busybox [18:42] Riddell, ok, I'll remove the refs from techoverview for the alpha1 for now. [18:42] Ok. The test on iso.qa points to 20101202 [18:42] and I just finished pulling it down. [18:42] yeah, just seeing that [18:43] Will restart zsync...again. [18:43] well, zsync to the new one should be fast [18:43] only a few byte have changed [19:01] Can the A1 image of Mythbuntu be canned... there is unlikely to be any testing, and the non platform delta with Maverick is none. [19:02] Daviey: yes, superm1 communicated that to us [19:03] we'll skip publishing it [19:03] ahh perfect, i missed that [19:04] Daviey, have gone ahead and disabled it in the tracker as well to make it clear. [19:05] skaet_: awesome. [19:23] Can someone update iso.qa with the correct ubuntu netbook armel image? It is 201011202.1 not 20101202. I don't want to enter data then have someone kill my data because of it being updated. [19:26] I thought that that one was going to fail too, due to the lack of ubiquity 2.5.5? [19:26] but I can certainly update it - doing so now [19:28] done [19:31] ogra_ac asked me to test it after manually modifying the image with the ubiquity patch. It seems to be working. BTW: Is there a bug filed for that patch? Need it for tracker. [19:37] GrueMaster, just add a note about it, there was no bug since it was found and fixed immediately [19:38] I can't pass the image as is. System doesn't work that way. [19:38] sure, mark it failed then, just so we have a datapoint for it [19:38] I need a bug number for that. [19:40] ?? [19:40] file one if you have to [19:40] just check the "failed" radiobutton and leave a note [19:41] though the fix is already uploaded so it's just pushing paper ... [19:41] right [19:43] I understand that it is all paperwork at this point. I didn't design the system. But I can't mark it as passed as the image fails without the fix, and I can't mark it as failed without a bug number. [19:43] So, I will install ubiquity and file a proper bug report. [19:43] * ogra_ac tries [19:43] you don't need to install ubiquity [19:43] wow [19:43] this doesn't need logs, it can just be a placeholder bug [19:43] thats bad [19:43] GrueMaster, i understand what you mean [19:52] Bug filed and updated as Fix Committed. iso.qa updated. I'm breakig for lunch. bbl. [19:53] is it out yet? [19:53] * highvoltage ducks [19:53] GrueMaster, thanks a lot [19:53] GrueMaster: Fix Released [19:53] (ubiquity 2.5.5) [19:54] GrueMaster, i will not ask you such stuff again but just strike it from the tracker in the future, that was silly (i didnt know the tracker is so bad in this regard) [19:56] never mind, I've closed it now [19:59] ogra_ac: This is one of many issues I have with the tracker. The biggest issue is that if we were to respin for any issue that is non-related to any test results previously posted, I can't pull them up and confirm that the status hasn't changed. [20:00] yeah, i remember you saying that before [20:00] do we have a project we can file bugs against ? [20:01] would be great if there could be a link in the tracker itself to file bugs against it [20:01] ubuntu-qa-website, I think? [20:01] though best check with somebody in QA [20:03] ogra_ac, cjwatson, GrueMaster: https://bugs.launchpad.net/ubuntu-qa-website/+filebug [20:03] perfect [20:22] GrueMaster, Bug 684404 and bug 684409 (if you want to subscribe to them) [20:22] Launchpad bug 684404 in ubuntu-qa-website "the isotracker should have a "file bug against this tool" link i its menu (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/684404 [20:22] Launchpad bug 684409 in ubuntu-qa-website "there should be a way to mark an iso test as failed even without a bug (affects: 1) (heat: 6)" [Undecided,New] https://launchpad.net/bugs/684409 [20:27] w/win 26 [20:27] does the first w stand for "whash" ? [20:31] I go through channels quite fast, so I guess it must be 'warp' [20:31] heh [20:54] ogra_ac: that's a new "feature" :) I didn't implement that in the tracker, with my code you could mark a failed without a bug number [20:55] stgraber, heh [20:55] stgraber, ogra_ac: the reason we put that feature in is to be able to track "failed tests with a bug number" i would prefer we don't go backwards! [20:56] well, for this case it cuased extra work for three people [20:56] but i understand that its a corner case [20:57] ogra_ac: understood [20:57] ogra_ac "most" of the time, when a tester claims a test failure, we want to track it [20:58] indeed [20:58] as i said to tobin above, in the future i will just wipe the image from the tracker [20:58] just felt more accurate to mark it properly failed with a comment [20:59] ogra_ac: ack [21:36] smoser: http://uec-images.ubuntu.com/releases/natty/alpha-1/ says "Daily Build" [21:37] yeah. i know. [21:37] i'll see why [21:37] ok, hadn't seen it on this channel [22:05] skaet_, release schedule shows 10.04.2 as Jan 27th, is that still correct? [22:06] bjf, hmm have missed an update or two. :) [22:06] should be in Feb [22:06] cjwatson, fixed now. http://uec-images.ubuntu.com/server/releases/natty/alpha-1/ [22:07] skaet_, maybe i'm looking at the wrong schedule? https://wiki.ubuntu.com/LucidReleaseSchedule [22:08] That's Lucid's... [22:08] skaet_, and lucid is the 10.04 LTS [22:08] bjf, https://wiki.ubuntu.com/NattyReleaseSchedule shows Feb 17 [22:09] I'll go in and update Lucid's if that's where folks will look [22:09] * skaet_ figured I'd missed a couple. [22:09] skaet_, it's also on: https://wiki.ubuntu.com/ReleaseSchedule/LTStoLTS [22:09] bjf, cool. willl take care of them all right now. [22:21] doko: Where did we agree to switch to Python 2.7 as default right after Alpha 1? [22:24] foundations meeting yesterday [22:24] though skaet_ asked that we wait until after the release meeting tomorrow [22:26] I thought we agreed for him to do an archive rebuild test. [22:27] The plan all along has to been to get 2.7 supported first. [22:27] We don't even have Main done yet. [22:35] oh, I missed "as default", sorry [22:36] http://irclogs.ubuntu.com/2010/12/01/%23ubuntu-meeting.txt:[16:41] in that case, I would like to make the test rebuild with 2.7 as the default, so we know immediately what needs to be fixed [22:36] http://irclogs.ubuntu.com/2010/12/01/%23ubuntu-meeting.txt:[16:41] doko: do it [22:36] I find that ambiguous [22:37] at the time, I understood that to mean "switch the default, and immediately do a test rebuild so that we know what we need to plough through" [23:11] cjwatson: yes, the latter is what I meant === bjf is now known as bjf[afk] [23:44] OK. [23:44] Not what I understood at all. [23:45] I thought that meant do a test rebuild so we'd know what needed to be done before we were ready to switch. [23:45] doko, cjwatson ^^^ [23:45] * ScottK doesn't understand the rush to break the archive? [23:48] I don't mind, as long as we're not so leisurely that 2.7 misses natty [23:53] there is no rush. and I'll be surprised to learn how you'll avoid breakage for a short amount of time [23:56] doko: Until you've done the rebuild test and we've at least got 2.7 support in packages that will support both, how do you know it will be short? [23:57] yes, this is what was missed until now, but afaiu barry, the rebuild test was done