=== yofel_ is now known as yofel [09:14] infinity, cjwatson I did a quick verification last week and the auto-upgrade-test doesn't support the upgrade from an LTS to the current dev release out of the box. It upgrades to the latest stable. I'll spend more time on it this week. [10:20] CUPS 1.6.2 is out. It is a bugfix-only release. Can I upload it? [10:21] tkamppeter, if it's only bug fixes and no new feature, yes [10:28] seb128, OK, so I will go ahead. [10:28] tkamppeter, thanks === mmrazik is now known as mmrazik|afk [10:45] infinity, cjwatson re upgrade from LTS to current dev release, I had a look and it's a bit counter-natural for the upgrader. Here are the changes required to allow this upgrade path: [10:45] 1. Add raring to http://changelogs.ubuntu.com/meta-release-lts-development [10:45] 2. in update-manager-core on Precise change the logic in UpdateManager/Core/MetaRelease.py lines 228-232 to consider the latest dev release as a valid upgrade instead of stopping at the first upgradable release (quantal) [10:45] 3. in ubuntu-release-upgrader on Raring add DistUpgrade.cfg.precise with From=precise and To=raring in the Sources section [10:46] With these changes I could upgrade a server image from precise to raring [10:46] infinity, cjwatson I can force 1 and 3 with the upgrade-tester but not 2 === mmrazik|afk is now known as mmrazik [10:58] * cjwatson fixes the overnight build failure and respins [10:58] jibel: that doesn't seem *too* awkward; shame it'll require an SRU === mmrazik is now known as mmrazik|lunch === mmrazik|lunch is now known as mmrazik [12:17] I think the kernel armhf build has stalled. 18 hours seems like quite a lot. https://launchpad.net/ubuntu/+source/linux/3.8.0-13.23/+build/4379671 [12:18] You'll need to ask #webops (internal IRC) [13:02] Could somebody look at NEWing tk8.5 and tcltk-defaults, so that I can move on with the ruby1.8 change based on those? [13:02] Trying to unstick a fairly long chain of stuff [13:03] cjwatson, looking [13:15] thanks! [13:16] yw ;-) [15:05] does this look ok for a seed change to add touch: http://paste.ubuntu.com/5628377/ [15:05] ? [15:05] Why Task-Seeds: ubuntu-minimal? [15:05] thats how the quantal image is currently built [15:06] The effect of that is to incorporate entries from minimal directly into the ubuntu-touch metapackage [15:06] Maybe, but it's wrong :) [15:06] standard has to much overhead i was told [15:06] You shouldn't use Task-Seeds for things that already have their own metapackage [15:06] minimal vs. standard is a different issue [15:06] oh, ok [15:06] What I'm saying is that the Task-Seeds line shouldn't be there at all [15:06] yeah, got that, the STRUCTURE line is enough [15:06] Other than that it looks fine to me [15:07] great, committing [15:07] and pushed ... [15:08] Oh, one missing thing - you need to add touch to the inheritance of some other seed, probably supported [15:08] Otherwise the packages in it won't be kept in main [15:08] not sure we want them in main yet [15:08] since the packages that are pulled in arent either [15:08] Well - it's kind of invalid to have things in that seed collection that aren't inherited by supported [15:08] ok [15:08] But I guess it might work temporarily [15:09] Is there a project somewhere to get everything in touch MIRed? [15:09] I mean, yes, adding touch to supported would trigger immediate component-mismatches for everything in it, so if you don't want that then it might not be the best idea [15:10] I didn't think that touch in main was a goal for raring - is it? [15:10] you mean adding it to the last line in STRUCTURE right ? [15:10] Laney, well a goal was to have a regular overview [15:10] of whats missing etc [15:10] so component mistmatches might be a good tool here [15:10] Maybe best leave it out of the last line in STRUCTURE for now [15:10] ok [15:11] We need component-mismatches for other things, and if it isn't a target for raring then making it harder for ubuntu-archive to do other jobs isn't a good plan [15:11] we wont have everything merged by raring release [15:11] there are intrusive changes to such minor things like NM :) [15:11] * cjwatson goes to update tasksel [15:11] which wont make it before release [15:11] Not that it matters, but [15:12] ubuntu-meta will just magically pick it up without any intervention, right ? [15:12] * ogra_ hasnt changed./added seeds in quite a while [15:13] ah, update.cfg ... [15:13] * ogra_ remembers ... [15:14] hmm, but that wants main indeed [15:17] Without main inclusion, I'd recommend a separate -meta source package [15:18] hmm well, after all it is supposed to end up on the same meta in the end [15:18] i would even go that far to claim that we'll have images containing touch and desktop at the same time at some point [15:18] (convergence) [15:18] I'm not saying it should be in a separate source package permanently [15:19] But for now, it will be easier [15:19] and lool and slangasek actually want me to start building images even of they fail [15:19] As long as you keep ubuntu-touch-meta's version number lower, it won't be a painting-self-into-corner problem [15:20] (which i think makes no sense just to see whats missing) [15:20] ubuntu-meta can just take over the binary once it moves to main [15:20] (Yes, ubuntu-meta can technically produce a binary package in universe. But getting germinate-update-metapackage to DTRT without breaking the other binary packages' dependencies is going to be a pain, and I don't think it's worth it) [15:20] just filed bug 1157227 whenever someone else on the release team has a few minutes to look at it. The main goal here is to get us closer to the final upstream release (currently on an alpha) so we can release 13.04 without having to worry too much about maintenance. [15:20] Launchpad bug 1157227 in lxc (Ubuntu) "[FFe] Get LXC 0.9~rc1 to replace the current 0.9~alpha1" [Undecided,New] https://launchpad.net/bugs/1157227 [15:21] I was looking at it [15:21] let me build packages to see if my stuff continues to work ;-) [15:21] ogra_: well I've suggested that and you've pushed back; we didn't discuss the specifics, but yes, I think we could either build against the target seeds even if that fails and use the failures as a todo list, or we could create the seed with some of the missing packages as commented out and build semi-broken images [15:22] My main goal was to parallelize the efforts of getting the images building with the efforts of pushing stuff to raring [15:22] It doesn't make any difference to image building whether ubuntu-touch is built from ubuntu-meta or ubuntu-touch-meta [15:22] lool, the issue is that it will spam component-mistmatches all the time (though as i understood that was said todo list) [15:22] No, that also isn't affected by image building [15:23] no, but by moving the packages to the archive and main [15:23] Since we've agreed not to add it to supported's inheritance list, it won't spam component-mismatches [15:23] We could produce a separate mismatches report [15:23] It wouldn't be the first time [15:23] quite some effort for just some weeks [15:24] Some weeks spanning a release [15:24] And component-mismatches is expected to be empty at release [15:24] well, S opens in what 6 weeks ? [15:24] i would just go with universe for now if lool and slangasek agree [15:24] ogra_: sure thing [15:24] instead of putting effort into creating separate reports etc [15:25] Well, sure, I was just saying that having a bunch of stuff known-not-ready in component-mismatches across raring release isn't a viable option [15:25] ogra_: we don't need main/universe distinction, at least not in the short-term [15:25] right and i agree [15:25] i was just echoing what i understood i was supposed to do :) [15:25] yup [15:25] * ogra_ creates ubuntu-touch-meta [15:26] Let me know if you need help with that [15:26] Done it a few times now :) [15:26] i'll just grab xubuntu and fix it up [15:31] stgraber: I still haz working containers, at least ;-) [15:32] Laney: good to hear ;) [15:32] yeah, seems fine to me [15:32] goferit [15:32] yay, thanks! [15:33] ogra_: do you know what the jenkins hostname/IP address is for my WI in https://blueprints.launchpad.net/ubuntu/+spec/foundations-1303-cdimage-android-builds ? [15:33] yep === rsalveti_ is now known as rsalveti === Ursinha_ is now known as Ursinha [15:37] cjwatson, http://10.97.0.6:8080/job/ubuntu-touch-image [15:37] ogra_: http://people.canonical.com/~ubuntu-archive/component-mismatches.txt [15:37] I probably need to look at why that's happening :) [15:37] cjwatson, more details are in my sync-phablet-images script [15:38] in ~ on nusakan [15:38] Right [15:38] urlpath=$jenkinsurl/job/$project/$ver/artifact/archive [15:38] thats the relevant bit [15:38] I'll fix up c-m after this call [15:39] hmm if you black/whitelist that anyway we could as well with the ubuntu-meta change [15:40] Nah, easier to use the component-mismatches command-line option to exclude a seed [15:40] ah k [16:10] ogra_: Ah, hmm, it's actually really tricky to exclude this from component-mismatches [16:11] I've used '-e touch', but that only excludes things that are direct run-time/build-time dependencies - it doesn't help with the case of extra binaries from the same source which then pull in other (build-)dependencies [16:11] hmm [16:12] And I'm not sure it's successfully excluding build-dependencies [16:12] ogra_: Would it be possible to move this to a separate seed collection for now? [16:12] I'm happy to set that up for you if you like [16:12] (I'd need to have some involvement anyway) [16:12] sigh, yeah ... still i think its a waste of manpower for just a few weeks [16:13] So is making c-m useless. [16:13] we'll merge them once S opens [16:13] It's a bit of effort, but not horribly much, and I can see the point of convergence here [16:13] I mean, I think in general having the definitions in seeds earlier rather than later is a good thing [16:13] infinity, right, i would rather postpone [16:13] I'll just move it to a separate branch for now, then [16:14] having everything prepared to merge with a finger snip [16:14] It'll be easy to move back later (though we might not get to keep the history of the seed) [16:14] ok, tell me the branch name for the meta then [16:14] * ogra_ hates such throw away work [16:15] (regardless who does it ... it currently doesnt serve any purpose and will be thrown out before it gets useful) [16:16] Does lp:~ubuntu-core-dev/ubuntu-seeds/ubuntu-touch.raring look OK? [16:17] ogra_: Branches are cheap, it's not really that much work. [16:17] (Less work than the discussion that preceded it) [16:18] cjwatson, looks fine [16:18] infinity, heh [16:19] infinity, btw, i guess we need to sit down at some point and come up with something completely new for the ondemand script .... it causes some havoc on phablet (my phone glows until i remove it and reboot for example) [16:19] That said, the component-mismatches explosion wasn't as bad as I expected, which bodes well [16:19] well, its only ~60 packages [16:20] I mean, still non-trivial work, but could be worse [16:20] and many are there already [16:20] OK, so I've removed touch from ubuntu.raring again [16:20] yep, thanks [16:20] ogra_: Your phone glows? Neat. [16:20] You'll want to adjust ubuntu-touch-meta [16:20] I made sure ubuntu-touch.raring was mirrored in the standard location (http://people.canonical.com/~ubuntu-archive/seeds/) too [16:21] infinity, well ... :) [16:24] ogra_: Do you know if 10.97.0.6 is also the host we'd need to prod with an ssh trigger? [16:24] (I can ask Sergio if you're not sure) [16:24] i think thats where all the test image runs are, yeas === tkamppeter_ is now known as tkamppeter [16:42] [i386] Loading seed lists... [16:42] * Checking out bzr+ssh://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/ubuntu-touch.raring/ [16:42] * Checking out bzr+ssh://bazaar.launchpad.net/~ubuntu-core-dev/ubuntu-seeds/platform.raring/ [16:42] ! Could not open supported from checkout of (any of):ile 30/36 [16:42] hmm [16:42] ogra_: I am only a bot, please don't think I'm intelligent :) [16:42] that doesnt look happy [16:42] i guess i need to wipe supported from STRUCTURE [16:46] * ogra_ does so and tries again [16:50] ogra_: no [16:50] ogra_: better to add an empty supported seed [16:50] ah [16:50] the last seed has weird semantics [16:50] ok [16:51] sorry, my bad for getting that wrong in the first place [17:04] xnox: → ubuntu-quality@ too, perhaps? [17:05] Laney: oh, i had no idea about ubuntu-quality@ not sure I am subscribed. [17:05] I'm sure balloons can moderate it for you [17:06] xnox, Laney i sure can moderate.. I'll forward if you don't want to crosspost xnox, but many on the list will be keen to read it [17:07] balloons: ack, cool. [17:08] balloons: and thanks. [17:10] ogra_: doesn't serve any purpose> it serves the purpose of making sure any phablet images we're building today have their definition where it should be - in Ubuntu, not in some out-of-the-way jenkins job definition. Please do not underestimate the importance of this [17:11] slangasek, well, i'm nearly done anyway no worries ... [17:11] just waiting for the ./update script to finish and will upload the new meta [17:12] slangasek, i dont underestimate the importance but i think it could have waited four weeks [17:13] no, it can't [17:13] now its done though :) [17:13] four weeks is a LONG time [17:22] stgraber: ubuntu slideshow - READY =) [17:30] hmm, i wonder why i dont seem to have any WI for creating the touch seed [17:32] ogra_, not sure about touch, but you have one on https://blueprints.launchpad.net/ubuntu/+spec/desktop-r-arm-reduce-footprint for the common/nexus seed [17:32] seb128, well i consider that obsolete/suspended atm === mmrazik is now known as mmrazik|eod [17:33] i just added one for touch [17:33] ogra_, right, you can either recycle or it POSTPONED it [17:33] i'll postpone ... i guess it will come up for convergence again at some point [17:34] right [17:34] thanks [17:37] ubuntu-touch-meta-1.001 sits in NEW ... i would appreciate a review and pass :) [17:41] looking [17:42] ah there it is [17:42] its completely empty atm .. "? Unknown touch package: ..." for all the packages ... [17:42] which should be right atm [17:42] $ cat metapackage-map [17:42] touch ubuntu-touch-touch [17:42] eek [17:43] was fine when the seed collection was ubuntu [17:43] does the script add that ? [17:43] needs a seed fix [17:43] one moment [17:43] ah k [17:44] * ogra_ wishes ./update wouldnt take ages ... it wrangles with an x86-android repo sync i'm running here [17:44] you should drop override_dh_germinate_metapackage too [17:45] cjwatson: hey, do you happen to have a list of dependencies for run_test in lp:ubuntu-cdimage? [17:45] ogra_: rejecting for both the above - could you re-update? I've fixed the seeds [17:45] done [17:45] yep, running [17:47] stgraber: should be just python-mock, plus python3 to run the python3 tests as well [17:47] cjwatson: actually, I just figured it out. It's: python-mock, python3, procmail, squashfs-tools and having your system use the GMT timezone [17:48] oh, I forgot about procmail [17:48] and yeah, squashfs-tools for test_check_installable [17:48] GMT timezone is a bug [17:48] what fails if you don't? [17:49] cjwatson: a bunch of string match fail [17:49] cjwatson: test_sync_lock_failure is one of them [17:49] huh, let me see if I can reproduce that [17:50] cjwatson: there's basically a timestamp in the log file which contains the timezone abreviation [17:50] cjwatson: might be worth calling "date -u" or similar, to generate timezone neutral timestamps [17:51] It's meant to but I obviously missed something [17:51] I'll fix that, thanks [17:52] It's already using time.gmtime(), so I think I just shouldn't have used %Z [17:54] stgraber: r1140 fixes timezone handling; r1141 documents dependencies [17:55] cool, thanks === mmrazik|eod is now known as mmrazik [18:03] jbicha: next daily build of Ubuntu GNOME will be auto-published on the iso tracker. [18:03] there was a small bug in the publishing script (wrong product name) so the past couple of builds didn't publish [18:33] there we go, next try [18:34] i left ppc out btw ... i hope nobody minds :) [18:36] stgraber: aha, thanks [18:39] ogra_: LGTM [18:39] thx [18:39] yay === highvolt1ge is now known as highvoltage [19:43] was just reading the TB meeting logs from yesterday, and it looks like it was decided that support period would be 9 months starting with 13.04. is that correct, or still need votes/opinions from the absentee TB members? [19:47] that's correct [19:48] IIRC 3 members + chair is the minimum requirement for TB quorum and we had that yesterday. Anyway, the 3 other members indicated their support to the proposal by e-mail to the TB mailing-list before the meeting. [19:48] Indeed. [19:49] ok [19:49] right, that's how i read the log. i just wanted to confirm. thanks :) === yofel_ is now known as yofel [21:54] So I saw Bug 1157191 today and suddenly my ability to maintain state of PS generated FFe requests has collapsed. [21:54] Launchpad bug 1157191 in ubuntu-ui-toolkit (Ubuntu) "[FFe] API updates for MainView, PageStack and Tabs " [Medium,Confirmed] https://launchpad.net/bugs/1157191 [21:55] seb128: Could we perhaps get a complete list of the stuff they still have to land? It seems like it's something new every day. [21:57] ScottK, I though there was an agreement/understanding that ubuntu-touch work in universe could go in without issue? [21:57] slangasek, ^ [21:57] ScottK, that shouldn't impact on the release or on any flavor... [21:57] seb128: There was some discussion about a general FFe request for such things, but I've never seen the request. [21:58] Right. That one's not particularly a problem. It's just they keep coming. [21:58] seb128: right, what ScottK said... I think I was supposed to file the FFe request and dropped the ball, sorry about that [21:58] It'd be nice to get everything on the table and reviewed once. [21:58] wrt the new packages, it should all be done as a single blanket FFe [21:59] slangasek, new like "not in the archive yet"? or like "new from this cycle that is under active work" (like the ui toolkit)? [21:59] But the blanket FFe should list them all, so we know what's approved already and what's not. [22:00] ScottK, afaik the "only" 2 FFe from the #ps team are the scopes one (that sabdfl overruled) and the in-dash-payment one [22:00] seb128: new like "new this cycle for the touch stack and not touching any Ubuntu images" [22:00] slangasek, ok, thanks [22:13] * ScottK gives up and approves ALL the FFes. [22:13] Actually not, but am tempted. [22:34] what is the actual size (disk space) of a full repo mirror as per https://launchpad.net/ubuntu/+archivemirrors and where should I head to for more details?