[01:36] cjwatson: /me ponders where did I miss it. [01:36] checking. [01:42] yes. [01:43] * xnox had a strange feeling something is wrong when plymouth migrated without dri-tools. [02:06] please accept colord-gtk, split of libcolord-gtk library from the colord source package. [02:20] xnox: Why the split? [02:21] infinity: ask the uploader. /me has no clue, but as far as I can see something to do with not liking gtk portion of it. [02:22] I thought you had some insight, since you were demanding it be accepted. :P [02:22] * xnox just wants empty britney [02:22] +commit fca4196ed897e086fd917f489c38d95278e4b1df [02:22] +Author: Richard Hughes [02:22] +Date: 2012-06-18 [02:22] + [02:22] + Split out colord-gtk to a new sub-project to prevent a dep loop [02:22] + [02:22] + At the moment GTK requires colord to build, but colord-gtk needs [02:22] + GTK to build. [02:22] + This makes bootstrapping a distro (or using jhbuild) harder than it [02:22] + needs to be. [02:22] That seems to answer it. [02:23] =))) [02:23] colord: - Remove colord-gtk packages that are now in a different source package [02:23] Yeah, your paste doesn't answer anything. :P [02:23] As I'd still ask "and why are they in a different source package?" [02:24] * infinity grabs the source to get all reviewy on its ass. [02:24] =))))) [02:24] ack, ack. [02:28] thanks a lot =) [03:38] why does britney not consider glew for transition? [06:27] xnox: because there are binaries in raring-proposed that are out-of-date (because they're NBS), and those need removed first; fixing === doko__ is now known as doko [09:08] slangasek: interesting. thanks. === mmrazik is now known as mmrazik|afk [09:34] xnox: Just wondering about the ETA of the fix for bug 1068178 :) [09:34] Launchpad bug 1068178 in ubiquity (Ubuntu Precise) "ubi-usersetup failed with exit code 1 on preseeded - encrypted-home installations on precise" [Undecided,Confirmed] https://launchpad.net/bugs/1068178 [09:35] psivaa: let me check. [09:36] Also i take precise desktop images being oversized is a known issue. === didrocks1 is now known as didrocks [09:58] cjwatson: I have cherrypicked the fix for the ^^^^^ ubiquity issue for precise. Are you planning to make a new upload of ubiquity to precise again soon or should the current upload pass verification first? [09:59] xnox: Let the secure boot crap all pass and get promoted first, unless this bug's a showstopper. [09:59] xnox: Pretty please. [10:01] infinity: ack. It's a show-stopper for one test-case: automatic pre-seeding of default install with ecryptfs enabled. [10:01] infinity: targetted for 12.04.2 to make sure we include it. [10:02] xnox: Well, I'm sure it won't be the only bug we're trying to squeeze into .2 ;) [10:02] xnox: (or bugfix rather) [10:02] xnox: But yeah, the SB stuff kinda all needs to land in a big (hopefully working) group. [10:04] psivaa: right. we currently have sru for ubiquity with 7 bugs to verify and I am sorry this bug did not make it into this sru. http://people.canonical.com/~ubuntu-archive/pending-sru.html After those are verified, I'll upload the fix for the encrypted-home bug. Sorry for the delay. [10:08] xnox: ok, hope it gets included in the next batch. thank you. [10:08] it will. === mmrazik|afk is now known as mmrazik [11:48] bdrung: In case you were wondering, I'm working on the reason that gworkspace didn't get copied to the release pocket (bug 1083131). [11:48] Launchpad bug 1083131 in Launchpad itself "Closing bugs OOPSes on invalid UTF-8" [Critical,New] https://launchpad.net/bugs/1083131 [11:49] cjwatson: i haven't noticed it yet :) [11:55] * ogra-cb__ wonders why he didnt see the plymouth issue when upgrading his nexus7 === ogra-cb__ is now known as ogra-cb [11:56] (I've confirmed that gworkspace is currently the only case where a publication was deleted from proposed without a successful copy to release and without having already been superseded by a later upload.) [12:08] ogra-cb: btw. should there not be libdrm-omap? [12:08] i think its called libdrm2-omap (not sure, would need to check) [12:09] and yeas, theer shoulld eb the omap one and the radeon and nvidia ones shoud go away on arm :) [12:09] (as well as intel) [12:10] ogra-cb: for the debs in plymouth. Ah.. it's libdrm-omap1 [12:10] ah, right [12:10] i knew there was a number in the name [12:12] hmm, the nexus7 is really unhappy with the raring kernel :( === lamont` is now known as lamont [12:18] cjwatson: FYI, a fixed gworkspace is uploaded (the changelog had ISO-88... and UTF-8 mixed) [12:19] I kind of wish you hadn't [12:19] I was going to use it as a test case for the fix [12:19] I guess I can still do that on dogfood [12:19] But it wouldn't have been necessary if you'd asked ;-) [12:21] cjwatson: sorry. feel free to reject that upload [12:22] Can't === greyback is now known as greyback|lunch === lan3y is now known as Laney [13:42] hmm, the nexus7 build still fails on the plymouth issue [13:43] is something stuck in NEW by chance ? [13:45] http://paste.ubuntu.com/1389006/ [13:45] ogra-cb, NEW is https://launchpad.net/ubuntu/raring/+queue?queue_state=0 [13:46] ogra-cb: well I didn't add libdrm-omap1 dep on armhf, yet. I guess you need it like _now_ [13:46] yeah, nothing there [13:46] * xnox didn't understand the urgency earlier. [13:46] xnox, no hurry [13:46] ogra-cb: did it work before? [13:47] ogra-cb, try to sudo apt-get install libdrm-intel1 libdrm-radeon1 libdrm-nouveau2 plymouth-theme [13:47] ogra-cb, to see what is the issue [13:47] 0.8.8-0ubuntu1 [13:47] i have that one installed on the last nexus7 image [13:47] didnt cause build issues [13:48] ogra-cb: and what is the current plymouth version? [13:49] hmm, i dont get that [13:49] (for you / in that log failing to install) [13:49] so the build from the 24th failed [13:49] err [13:49] 25th [13:49] and todays too [13:50] bur the one from the 24th worked with the same plymouth version [13:51] http://cdimage.ubuntu.com/daily-preinstalled/20121124/raring-preinstalled-desktop-armhf+nexus7.manifest [13:51] I see only 24th log here: http://people.canonical.com/~ubuntu-archive/cd-build-logs/ubuntu/raring/ [13:51] libplymouth2 0.8.8-0ubuntu1 [13:51] where are the 25/26th logs? [13:51] failed [13:52] not sure why they arent copied, i got them by mail [13:52] both have the above pastebin error [13:52] ogra-cb: can you forward them to xnox@u.c ? [13:53] well it should be 0.8.8-0ubuntu2 across the board [13:54] ubuntu2 was only uploaded this night [13:55] 3am according to raring-changes [13:55] * xnox was awake.... [13:55] ogra-cb: so libdrm-intel1 is not built for armhf since quantal and up. [13:55] well, i would assume a libdrm issue [13:55] doesnt smell like plymouth [13:55] ogra-cb: but the later two are available, so why does the resolver tries to use the first one instead of any of the other two. [13:56] well, why does it use them at all :P [13:56] you will never find any compatible HW for either of these libs on arm [13:57] but iirc the prob was that the initrd stuff needs to be rewritten to be able to drop them [14:13] ogra-cb: Hmm. I assumed that that was due to needing s/libdrm-nouveau1a/libdrm-nouveau2/, which is why I asked xnox about that yesterday; but that isn't the problem any more. [14:13] xnox: The resolver message indicates that none of the three are installable. [14:14] * cjwatson sets up some raring-only chdist environments [14:14] yeah, whats really so confusing is that i got one successfull built on the 24th with the accused to be broken plymouth version [14:16] there was a libdrm upload on thesday it seems [14:16] *tuesday [14:17] ogra-cb: and that was stuck in proposed until plymouth and dri-utils got rebuilt/fixed. === davmor2_ is now known as davmor2 [14:18] Hmm, can't seem to reproduce in chdist [14:19] * ogra-cb tries a dist-upgrade on the 20121124 image [14:19] Wonder if I can run a test build on my Nexus 7 :-) [14:20] It only really has to manage debootstrap, which should be fine ... [14:20] you surely can, but i would suggest using a USB key/disk via the otg cable ... might be faster [14:20] Meh, debootstrap output isn't that large [14:21] well, yes, and it wont be eath shattering faster either [14:21] *earth [14:21] I'll give it a try in a few minutes once I've upgraded it sufficiently [14:21] just a question -- are we now shipping kernel 3.5 by default on Precise? [14:22] everything but GLES stuff should be buildable just fine on the nexus [14:22] hggdh: Yes, the quantal enablement stack [14:22] and i bet even way faster than on any panda [14:22] cjwatson: thanks [14:22] (just disk io isnt thrilling) [14:22] ogra-cb: Once it has overlayfs support it should be a nice enough sbuild machin [14:22] e [14:22] heh [14:23] if you could charge while having a HDD attached we could use a stack of them as buildd farm :) [14:23] hggdh: For new installs, at any rate; nothing auto-upgrades existing installs [14:24] cjwatson: yes, makes sense. We were just surprised to see new installs with 3.5 in QA. I do not remember receiving any notice it would be done now [14:25] ogra-cb: Are you running nexus7 builds by hand? I don't see them in nusakan's crontab [14:25] dist-upgrade seems to work fine here btw [14:26] hggdh: I probably assumed the kernel team would tell you :) [14:26] cjwatson, i added them for 13:32 [14:26] right next to the lubuntu preinstalled ones [14:26] ogra-cb: Oh, never mind, failure to read default-arches [14:26] I was grepping crontab for nexus :) [14:26] heh [14:27] cjwatson: Hey, psivaa said they our precise daily images have quantal kernel? [14:29] Daviey: Scroll up :) [14:29] aha [14:29] https://launchpad.net/ubuntu/raring/+source/libdrm/2.4.40-1 says it was only published at 24th [14:29] sad that it doesnt tell the time [14:29] Hover [14:29] but i would guess its libdrm then === didrocks1 is now known as didrocks [14:30] As in, if you mouseover the date, you'll get a tooltip [14:30] since the image build of the 25th was the first that failed [14:30] Or whatever you call it [14:30] hmm [14:30] cjwatson: I understood that this would not impact server. [14:30] 03:33 CET [14:30] os that would mean it was available on the 20121124 image [14:30] *so [14:31] (thanks btw, didnt know about the tooltip) [14:32] ARGH [14:32] that dist-upgrade made my initrd go from 1.5M to 3M [14:35] Daviey: Perhaps it would make sense if all the people involved talked to each other. It will be very very inconvenient to accommodate your request. [14:35] Daviey: What is the reason behind your request? [14:39] cjwatson: I did just send a mail to those involved checking.. I thought that was what was agreed. [14:39] cjwatson: Having a higher reliance on dkms and other stuff that reaches into kernel spae, concerns me that we are doing this. [14:39] I thought server was purely sticking with the LTS kernel.. But i am checking for details now. [14:41] Then y'all need to tell me very clearly what you actually need. [14:42] And you won't get reliable secure boot support. (You probably don't care, but I should be absolutely clear about this.) [14:42] cjwatson, could it be that there is some issue with the mirror used to build the images ? i cant really figure out whats wrong, manual dist-upgrade just works [14:42] as well as installing in a chroot [14:42] ogra-cb: No, it's reproducible in a local build here [14:42] ah, good [14:43] Just working on debugging it now. [14:43] cjwatson: Okay, yeah.. I'll stay shum until we have clarification. Thanks [14:44] Daviey: It may not be too horrible to fix, I guess, but I need total clarity or else there'll be chaos as we go back and forth. [14:45] cjwatson: yeah, just hold out. It might be what was expected all along, but it wasn't what i heard. So lemme get clarification [14:46] (Since I did go to the effort of making it parameterised by a variable in CONF.sh, so I could make that project-specific.) [14:59] ogra-cb: Ah. I demoted libdrm-radeon1 from Priority: required to optional a bit too enthusiastically, apparently. [14:59] oh, hard to catch [14:59] ogra-cb: The effect of this is that debootstrap's output has unsatisfied dependencies. [14:59] Fixed in the archive now, for the next publisher run [14:59] k, i'll trigger a manual build in a few hours [15:00] That was the cause of the powerpc image build failures too. [15:01] we really need to clean that up in an arch specific way some day ... on arm none of these libs make sense (apart from omap1) [15:05] ogra-cb: Start your list now for the UDS-S spec of "stuff I wish I'd had time for in Raring." [15:05] ;-) [15:05] heh [15:06] thats from the "list of things i wish i had had time for in natty" [15:08] its not like thats a new issue ... just rather complex due to the initramfs involvement [16:19] infinity: ping === smoser` is now known as smoser [17:14] cjwatson: I think rsyncable is broken on the desktop iso's. Over the past week right now I have hit with zsync "no relevent local data found" will redownload the whole file. [17:14] Don't know, I'm afraid; depends rather on squashfs-tools' behaviour [17:14] cjwatson: and when ogra-cb & infinity were poking (one of the cd building softwares) they may have mentioned lack of gzip --rsyncable options. [17:14] I don't recall there being a special option for it or anything [17:15] But gzip isn't used ... [17:15] hmmm [17:15] At least not anywhere that matters much [17:15] (it may have been something for nexus7) [17:15] Anyway, AFAICS --rsyncable just isn't documented, but exists [17:15] Compare the output of 'gzip --rsyncable' and 'gzip --garbage' [17:16] Shortly (once digikam is updated) we'll be in a position where only two packages block getting all of KDE 4.10 beta 1 (4.9.80) from raring-proposed to raring. Those both need upstream porting due to API changes, so I think we don't want to wait. Would the preferred approach be to force everything in despite those packages or to remove their binaries and then everything should migrate naturally? [17:17] Hmm [17:17] It screws users of those packages either way [17:17] Are the packages on your images? [17:18] One of them is. [17:19] The part I don't understand though is why the impact isn't just two versions of the library for awhile? [17:19] Then it's probably best to remove the binaries so that images can keep building, and file an RC bug as a reminder. I'm concerned that this approach undermines the message about raring being continuously usable, though. [17:19] Oh [17:19] cjwatson: I do wonder how I can debug my zsync problem. [17:19] ScottK: If that's all it is, it should be OK to force it [17:19] xnox, cjwatson, --rsycable isnt set fr tarballs inside live-build [17:19] britney considers the situation as if all NBS were removed [17:19] Ah. [17:19] debian-cd uses it [17:19] Now I understand better. [17:19] Or all NBS from the new source anyway [17:19] for the images themselves [17:20] Who can force stuff? [17:20] * xnox adds a todo item. [17:21] ScottK: ~ubuntu-release [17:22] Is there an ubuntu-archive-tools script for that or more properly, what wiki page or some such should I be reading instead of harassing you? [17:22] ScottK: First time, you need to edit britney.conf in ~ubuntu-release/britney/britney2-ubuntu to add a hint permission for yourself (just follow the pattern) [17:22] OK. [17:22] er with lp: on front [17:22] Gathered that bit. [17:22] ScottK: Then create a file for yourself in lp:~ubuntu-release/britney/hints-ubuntu with the hints you want [17:23] http://ftp-master.debian.org/testing/hints/README has the syntax [17:23] Thanks. [17:23] I should probably wikify this at some point [17:26] Then once the packages move, the hint can be removed .... [17:29] At your leisure, yes [17:29] Thanks. I think I'm all set up now. [17:31] SHRIEK !! [17:31] http://paste.ubuntu.com/1389492/ [17:32] so that build obviously failed [17:48] grmbl [17:48] there was no upload that could have caused live/build to break that way i think [17:49] ogra-cb: well, seb128 did upload pyxdg [17:49] oh [17:50] ogra-cb: and I wonder which file it's processing & fails on. [17:50] seb128, seems that broke update-apt-xapian-index somehow [17:50] obviously a kde one [17:51] but we dont log anything from subprocesses called in hooks during build it seems [17:51] ... and there were plently of kde uploads lately migrating. [17:51] so its hard to get more info beyond the tracback [17:51] ogra-cb, urg, sorry about that ... I was unsure how to test it so I mostly relied on the upstream test suit, seems that was not enough ... looks like a good case to add regression tests [17:51] Run it locally and then you can just chroot in and try again [17:53] ogra-cb, it traceback on /usr/share/app-install/desktop/spout:spout.desktop [17:54] Categories:Application:Game:ArcadeGame [17:54] it doesn't like that syntax error [17:55] bah, silly games [17:55] that's the only issue [17:55] if you move that .desktop away it works [17:57] ogra-cb, want to fix app-install-data? [17:58] ogra-cb, I need to run for ~1 hour but I can have a look later if needed, I will look at making pyxdg robust to those parsing errors but the easy fix is to update spout:spout.desktop to use "Categories=...;...;..;" [17:58] that's the lack of "=" that makes the parser unhappy [17:59] I'll fix the spout source package now [18:00] cjwatson, thanks === LordOfTime is now known as TheLordOfTime [18:05] the pyxdg "bug" was introduced in http://cgit.freedesktop.org/xdg/pyxdg/commit/xdg/IniFile.py?id=39407c25769dd90ab9d5e8ee4c49b08f2d3f7089 [18:05] - index = line.find("=") [18:05] - key = line[0:index].strip() [18:05] - value = line[index+1:].strip() [18:05] + key, value = line.split("=", 1) [18:05] well it was probably buggy before but not hitting an exception [18:05] uploaded; it'll need somebody to refresh app-install-data-ubuntu once that's in place [18:05] Heh, upstream filed a bug on Ubuntu's spout package, even [18:05] do we need/want a pyxdg workaround upload? like revert that diff? [18:06] or is that ok to wait for app-install-data-ubuntu to be refreshed? [18:06] Meh, it was busted, let's just fix a-i-d-u [18:06] works for me [18:06] I can poke it after dinner [18:07] (i.e. manual refresh if necessary) [18:07] thanks [18:07] need to run, bbl === scott-work is now known as Guest83303 === exekias_ is now known as exekias [18:48] GunnarHj: ? [18:51] infinity: Hi Adam, I was about to ask for your help to publish a few im-switch SRUs. ScottK seems to have done it already, though. Possibly he missed the quantal one. Maybe you can check if that's still in some queue? [18:52] GunnarHj: I see nothing named im-* in the quantal queues. [18:53] I think I got that one. [18:53] * ScottK looks [18:53] Oh, unless by "queue", you mean "archive". [18:53] GunnarHj ... [18:53] --- Releasing im-switch --- [18:53] Proposed: 1.22ubuntu2.1 [18:53] Release: 1.22ubuntu2 [18:53] Copied to quantal-updates [18:54] infinity, ScottK: I don't see it yet at https://launchpad.net/ubuntu/+source/im-switch, but I just got an email, so it's probably in the archive soon. Thanks! [18:54] GunnarHj: You want /$version/+publishinghistory on the end of that. [18:54] GunnarHj: It should show you the pending records. [18:55] infinity: Aha, thanks for the tip. [18:55] https://launchpad.net/ubuntu/+source/im-switch/1.22ubuntu2.1/+publishinghistory <-- For example. [18:55] * ScottK went through and released all the verified SRUs within the last hour. [18:56] ScottK: Thanks. Now I get to flood proposed with a bunch of queue reviews today. [18:56] Yes. Please. [18:56] And the cycles continues... [18:57] infinity, thanks in advance for doing queue reviews ;-) [19:02] cycling keeps you fit :) [19:03] great, one of the two powerpc building just picked up libreoffice to build [19:04] building->builders [19:05] Yes, fat software needs building too. [19:06] infinity, I'm more annoyed by the n_builder = 2 and the fact that nothing will move out of proposed to raring until the powerpc backlog is cleared :p [19:07] I've scored down builds on powerpc that I knew would fail to keep them from blocking up stuff until the queue is empty. [19:07] oh, great, it failed [19:07] heh [19:08] well, everything graphical is failing due to the out-of-sync-between-arch of fontconfig [19:08] ok, dinner time, bbl [19:09] seb128: I just scored fontconfig up then. [19:10] seb128: You can be annoyed about it all you want, but I'm not sure what good it will do. ;) [19:11] FYI, powerpc delay annoys the heck out of me too :P [19:13] arent we supposed to be over that soon ? [19:13] All it takes is a DCE to go revive sulfur. [19:13] Tomorrow apparently. === yofel_ is now known as yofel [19:15] Yeah, that's "all it's needed" for a while now. Just a bit light on staff in London right now. :/ [19:57] infinity, well, the issue is that there is always a good reason and a solution "soon" but the fact is that powerpc is an issue again and again and again every cycle [19:57] ScottK, thanks [19:57] At least there's a third builder down and down one = 2 and not =1. [19:57] infinity, it's blocking SRUs, delaying normal work for 99% of users for the benefit of 1% etc [19:59] That last Main builds in the quantal backlog are building now. === LordOfTime is now known as TheLordOfTime [21:20] Of course fontconfig becomes installable on powerpc just as the digikam build that I know will fail half way through starts ... [21:26] Timig: iz everything. [21:26] Timing, too. [21:27] You shouldn't have corrected yourself. I thought that was some new lolcat dialect with which I was not familiar. [21:29] Uploading that app-install-data-ubuntu update now. [21:40] please could the NEW packages for walinuxagent be accepted for raring; start of the SRU process for a critical fix.... [21:48] jamespage: Reading that bug and the fix kinda made me die a little inside. [21:51] please reject the previous gnome-shell upload and I hope I don't lose my place in the SRU line :| [21:52] jbicha: Old one rejected, review rescheduled for January 3rd. [21:53] jbicha: (I'm working on a bunch of reviews this afternoon/evening, we'll so how many I can get through) [21:53] s/so/see/ [21:53] infinity: yay! I should be able to get the paperwork done by January [23:49] OK. Now that I retried everything that failed due to fontconfig uninstallibity, the powerpc build queue is looking suitably crappy again.