=== jamesh_ is now known as jamesh [09:31] hey folks [10:06] om26er: hello! Are you free for some maguro testing of image 76? :) [10:06] sil2100, sure, let me upgrade [10:06] om26er: thank you! [10:16] sil2100: what was the blocker bug for promotion yesterday? something related to setting or upgrades? [10:16] sil2100: is it fixed now? [10:18] lool: hm, there was no particular promotion blocker, since we promoted all we could - we didn't promote any other image because we didn't manual-test those through ;) [10:19] lool: there was an online settings test crash on maguro, but it's just flaky [10:19] can someone test video playback on the latest image? [10:20] it's broken here [10:20] I'll try upgrading [10:21] also. why does mediaplayer do facebook/twitter/friends stuff? [10:22] sil2100: ah ok [10:22] I just realised that I only get notification of tweets / facebook messages _after_ I open a video [10:22] * popey will file a bug [10:23] popey: video playback >> regression from :73? [10:23] yes [10:25] sil2100: did we keep the URL of the online settings issue? would be good to escalate [10:26] lool: not yet, someone noticed it late late evening yesterday, so it was already my EOD [10:27] sil2100, on the latest image video player does not play videos, seems it crashes [10:27] sil2100, on both mako and maguro that is [10:28] Ouch [10:28] Ok, so it's confirmed then, let's see the changes between those images [10:29] in one of the image gstreamer0.10-ffmpeg was dropped, probably that ? [10:29] http://people.canonical.com/~ogra/touch-image-stats/20131218.changes [10:29] Yes, see that as well, hm hm [10:29] sil2100, where is the official place to check changes between images ? [10:30] I wonder why? [10:30] om26er: I always use the pages generated by ogra_'s magic [10:31] haha seems popey already reported that ^^ :p [10:37] http://paste.ubuntu.com/6598991/ [10:39] om26er, popey: anyway, try filling out a bug for this one, in the meantime I'll try to check what dep-change could have really caused this [10:40] ya [10:40] will file a bug [10:47] om26er: bug 1262600 [10:47] bug 1262600 in mediaplayer-app (Ubuntu) "mediaplayer no longer plays video in image #76" [Undecided,New] https://launchpad.net/bugs/1262600 [10:53] also bug 1262607 [10:53] bug 1262607 in friends (Ubuntu) "Social notifications only appear after mediaplayer is opened" [Undecided,New] https://launchpad.net/bugs/1262607 [10:57] sil2100, did ralsina tell you want was needed for udm? I had to EOD because it was very late and told him to take care of that [10:57] mandel: he got the rebuild started, it's in the landing plan, manual testing was discussed [10:57] ralsina, awesome! [10:57] ralsina, buenos dias! ;) [10:57] mandel: but it all happened in like 3 minutes after you left :-) [10:58] ralsina, great! [10:59] also bug 1262611 [10:59] bug 1262611 in unity8 (Ubuntu) "Network indicator doesn't display until you pull down" [Undecided,New] https://launchpad.net/bugs/1262611 [11:00] mandel, ralsina: hello! [11:00] So, I expected that Ken or Robert will handle the release of this one yesterday, but sadly it didn't happen [11:01] I'll pick it up in a moment and do it till EOD [11:01] sil2100: awesome, thanks! [11:01] (before EOD) [11:01] Just need to finish something first and deal with the mediaplayer-app regression [11:02] sil2100, superb! thx a lot [11:08] if you have an image which still has the gst stuff installed, install aptitude and then run aptitude why [11:22] sil2100: lool: maguro tests completed, re-running unity8 now. pls let me know if you want me to rerun anything next [11:23] psivaa: thanks! [11:23] lool: I think we can kick a new build in the meantime [11:24] But no promoting until the mediaplayer-app flock is resolved [11:24] sil2100: is the gst thing fixed then? [11:28] sil2100: ok [11:29] sil2100: buildin [11:30] popey: no, but we won't make it for the morning build anyway - if we fix it we'll catch it in the evening [11:44] sil2100: who needs to help on the flock issue? [11:44] or is it all sorted? [11:45] asac: for now I don't have time for that, we have a big regression with video playback and I'm trying to get to the root cause [11:45] sil2100: i referred to: [11:45] 12:23 < sil2100> But no promoting until the mediaplayer-app flock is resolved [11:46] hence i thought that was the main blocker :) [11:46] sil2100: who owns mediaplayer? [11:46] sil2100: was that regression in the promoted image? [11:46] or is that new? [11:47] asac: from what popey said it wasn't in the promoted image [11:47] popey: you tested on 73 and it was ok there, right? [11:48] 73 works, 76 does not [11:48] did not test 74 or 75 [11:48] popey: ok, so it's not gst lack's fault [11:49] Damn, it doesn't make much sense - I tried it on my device and I get a segfault as well [11:50] asac: bfiller and renato are responsible for mediaplayer, both are not on - but let's see what we can do by ourselves [11:50] sil2100: by ourselves? thats not really the way we should do it [11:50] get them involved [11:50] of course if you have identified a source through bisecting fior them thats good [11:50] but rather escalate early and often [11:51] than holding on to things for ourselves to clean [11:51] we got this: http://people.canonical.com/~ogra/touch-image-stats/20131218.changes [11:51] dropping stuff [11:52] Yes, get them involved, but I'm not going to block on them not being around [11:52] asac: but it's 73, which was good [11:52] sure not block [11:52] thought they were around [11:52] if we are sure [11:53] then its probably lxc-android-config and friends [11:53] and lxc [11:53] try downgrading those [11:53] http://people.canonical.com/~ogra/touch-image-stats/20131218.1.changes [11:53] other candidate (maybe even more related frfom topic) is https://www.flyn.org/projects/libdmapsharing/ [11:53] we upgraded that [11:54] http://people.canonical.com/~ogra/touch-image-stats/20131218.2.changes [11:54] sil2100: try downgrading those two things [12:14] asac: I already tried libdmapsharing earlier but didn't help, now I'm trying to recover from my device being broken by a downgrade of lxc [12:14] But I'm slowly triaging [12:19] sil2100: check with -touch folks [12:19] i think its not trivial to downgrade lxc [12:19] etc. [12:19] ogra_: ^^ do you nkow :) [12:19] (sorry) [12:20] asac, hmm, ask stgraber ? ... (he is in europe) [12:20] is that about video playback ? [12:21] (why are you discussing lxc ?? ) [12:21] ogra_: yeah. video playback regression [12:21] ogra_: because its pretty much the only thing that changed [12:21] but look :) [12:21] right, must be a dependency change somewhere [12:21] oh wait [12:21] i think i misrad the changes :) [12:21] gstreamer0.10-ffmpeg was dropped for whatever reason [12:22] its either http://people.canonical.com/~ogra/touch-image-stats/20131218.2.changes [12:22] i see that xnox did demote it recently to a recommends in another package (which would remove it from touch if that was the dep we need) but thats was end of nov and doesnt seem related [12:22] http://people.canonical.com/~ogra/touch-image-stats/20131219.1.changes [12:22] or http://people.canonical.com/~ogra/touch-image-stats/20131219.changes [12:22] ogra_: the image produced based on 18.1 was good [12:22] It's not lxc's fault anyway [12:23] definitely not [12:23] Image 74 plays fine [12:23] ogra_: it's dropped for very good reasons, which are outlined in the bug report. [12:23] right [12:23] hence [12:23] http://people.canonical.com/~ogra/touch-image-stats/20131218.2.changes [12:23] xnox, might be, it is essential for touch [12:23] http://people.canonical.com/~ogra/touch-image-stats/20131219.changes [12:23] http://people.canonical.com/~ogra/touch-image-stats/20131219.1.changes [12:23] are the candidates [12:24] we checked libdmapsharing [12:24] guess if lxc we can rule out [12:24] there is only upstart-app-launch [12:24] left [12:24] sil2100: tried that? [12:24] ogra_: it is not. we legally cannot ship ffmpeg, it's blacklisted, it's not installable in the archive and about to fall out. If you need back the support for those codecs, the app in question must be ported to gstreamer1.0 and use libav [12:24] i guess if its about "cant start playback" [12:24] it smeels like its again upstart-app-launch [12:24] ogra_: the bug has been set at critical priority, since saucy, yet it hasn't been touched nor fixed yet. [12:25] xnox, well, without gsdtreamer.-ffmpeg, no video playback [12:25] ogra_: the image 73 and 74 are good [12:25] asac: I'll check that, but damn the app is starting fine, just crashes suddenly [12:25] the drop happened for 73 [12:25] so... [12:25] ogra_: it's not installable either any more. it's about to fallout on all architectures from the archive. [12:25] sil2100: are you sure its not on 74 and 73? [12:25] xnox, we added it on request of management [12:25] asac: sure, it's not gst [12:25] xnox, that means no video playbaclk on touch then [12:25] ogra_: good, it's also blacklisted from all ubuntu releasable images as per TC. [12:25] asac: I just have 74 on my device and it's fine, 73 was fine as well as popey said [12:25] xnox: yes. it was added explicitely. who triggered the removal? [12:25] ogra_: please use other codecs. [12:26] xnox, there were long discussions with legal and management before pulling it in [12:26] xnox, we cant [12:26] I'll just continue digging, no need for everyone to look into that [12:26] if you want to override the TB you should talk to the TB [12:26] not Canonical management [12:26] ogra_: in that case please use gstreamer1.0-libav as soon as possible. let me look up the bug report. [12:26] asac: what's the progress on https://bugs.launchpad.net/gallery-app/+bug/1221968 ? [12:26] Ubuntu bug 1221968 in gallery-app (Ubuntu) "GStreamer 1.0 port of video thumbnailing" [Critical,In progress] [12:26] cjwatson, lool and pmcgowan were involved in the discussion ... i even refused ot have my name on the seed change [12:27] asac: this has been critical since september and still not resolved. [12:27] please talk to them [12:27] xnox: dot know the status [12:27] all i know is that ffmpeg is currently needed [12:27] asac: this is the correct way to pull in gstreamer ffmpeg backend in trusty [12:27] xnox: still, who triggered the removal>? [12:27] ogra_: I didn't address you specifically; anyone who's attempting to override the TB for something within the domain of the Ubuntu project must talk to the TB [12:27] rather than working around them [12:27] asac: can you please get it fixed faster? [12:27] i dont know [12:27] it's not OK to just ignore it [12:27] i am not familiar with the technical reasons [12:28] i just want to ensure that we dont shart shoot each others and talk to each other [12:28] before doing things like this [12:28] asac: the demotion of ffmpeg1.0 dependency was blocking approximately 200 packages migration in -proposed, including gtk stack, server stack, samba4 stack. This was done part of the elaborate libav9 transitions. [12:29] i dont understand [12:29] asac: it's part of the plan to resolve _not_ having two sets of gstreamer stacks on the touch & desktop images. [12:29] i dont want to know the reaons, just who did it and why there was not even a single line of talk about it [12:29] xnox, but we cant remove the working stuff before we have the new stuff ready [12:29] are you sure there was not a single line of talk about it, rather than talk you didn't see? [12:30] I saw lots of chatter about the libav transition [12:30] asac: the problem is that gallery-app is not ported to gstreamer1.0 and all requests to get that sorted have been ignored by gallery-app upstream, which is holding back the rest of the ubuntu. [12:30] noone appraoched those that are doinmg the touch image and try to bring a phone together [12:30] at least i didnt hear it [12:30] asac: i've appoach lool and assignee of bug report bug 1221968 about it. [12:30] bug 1221968 in gallery-app (Ubuntu) "GStreamer 1.0 port of video thumbnailing" [Critical,In progress] https://launchpad.net/bugs/1221968 [12:31] xnox, so instead of trying harder to get uphstream to fix it you just make the touch images unusable ? [12:31] asac: who should I have talked to about it? [12:31] ok [12:31] well. [12:31] xnox: landing team [12:31] ogra_: how is this a productive approach [12:31] xnox, upstreams manager ? [12:31] asac: landing team does not right code? [12:31] xnox: if you feel you might change the features of the touch image, alwasy go to the lnding team [12:31] ogra_: how would i know who that is? [12:31] asking ? [12:31] xnox: its not about the solution. its about telling someone that you are about to kick ffmpeg out of our iamge [12:32] anyway. so whats the fallout? [12:32] ogra_: ? [12:32] cjwatson, it is productive to find a way to solve this differently than by making touch images unreleasable the next time it happens [12:32] do we have no working audio? [12:32] people said that image 73 was good [12:32] so not sure if they didnt test it [12:32] ogra_: shouting at people is not productive [12:32] i didnt shout [12:32] shouting? [12:33] didnt hear that :) [12:33] i asked why it was not tried harder to get upstream to fix it [12:33] I assume that gallery-app must have gone through the landing pipeline, since it was daily-released [12:33] asac: this was discussed on #ubuntu-release and #ubuntu-devel and was proposed and reviewed and accepted upstream which you should see plenty of bug-mail and code review changes, which has passed all of the CI and autopilot on premerge and post-merge. [12:33] asac: and it was done a long time ago. [12:33] however due to the abysmally crappy spreadsheet I cannot check history [12:33] cjwatson, it sure did [12:33] therefore, the landing team was consulted [12:33] kk [12:33] if that was done then ignore [12:33] but you have chosen to micromanage everything so you probably can't see the wood for the trees as a result ... [12:34] so meh [12:34] I should go on holidy [12:34] a [12:34] * ogra_ should go back to that again too [12:35] sil2100, so did you try to roll back gstreamer and reinstall ffmpeg already ? [12:35] ogra_: the image that has the gstreamer change didnt have the regression [12:35] hence this whole discussion is slightly moot :) [12:35] how do I check this bug? [12:35] ogra_: yes, actually - using an image that's already missing the gstreamer stuff it still works fine [12:35] ogra_: but you said you knew that there is now something broken, what exactly is it? [12:35] asac: you made me feel really bad just now. i do not feel welcome. And the tone of above discussion and how it was started, is over the top for me, what i consider a healthy working environment. Especially, when it's unwarranted. [12:35] ogra_: so it's not gst [12:35] * sil2100 is bisecting further [12:36] One by one, I'll get to it [12:38] ogra_: i tried reinstalling gstreamer on #76 and it didn't fix the issue [12:38] k [12:38] I still couldn't play video [12:38] popey, ffmpeg as well ? [12:38] it pulled in all the dropped packages [12:38] the ones listed in the bug [12:38] and rolled back the updated ones ? [12:39] no, i just added gstreamer and the libav friends [12:39] didnt roll anything else back [12:39] ogra_: but as I said, image 74 which is already after the dropping works fine [12:39] oh, please do that too [12:39] revert what? specifically? [12:40] This regression makes me sad :< [12:40] meh, shit happens [12:40] libgstreamer0.10-0 [12:42] revert to 0.10.36-1.2ubuntu2 right? [12:42] as per http://people.canonical.com/~ogra/touch-image-stats/20131218.changes [12:42] yep [12:42] k, ta [12:42] xnox: yeah sorry, wasnt very well done well by me. you are very welcome [12:43] i didnt want to know about the WHO because i want to complain about him doing something wrong, i just wanted to understand who to talk to to figure if the coordination on this was done right or if we have a process problem. [12:43] asac: apologies accepted. [12:44] E: Version '0.10.36-1.2ubuntu2' for 'libgstreamer0.10-0' was not found [12:44] bah [12:44] xnox, sorry, i think it was my fault steering it in that direction ... i didnt mean to blame you specifically [12:44] ☃ ☃ ☃ ☃ ☃ ☃ ☃ ☃ ☃ ☃ [12:44] popey: you have to dl it from LP manually, sadly [12:44] oof [12:45] xnox: so by now i think i guessed it was you and you did it properly and we dont even have a known regressiono from the issue. so i am happy :) [12:45] popey, https://launchpad.net/ubuntu/+source/gstreamer0.10/0.10.36-1.2ubuntu2 [12:45] ogra_, popey: the sad thing... I think I upgraded every package that makes sense from image 74 and I still have working video! All that is left is autopilot-related packages o_O [12:45] https://launchpad.net/ubuntu/+source/gstreamer0.10/0.10.36-1.2ubuntu2/+build/5326960/+files/libgstreamer0.10-0_0.10.36-1.2ubuntu2_armhf.deb looks good [12:45] asac: right, but that would be WHAT and WHERE =))) e.g. upload changelogs, merge proposals, landing spreadsheet / bug tracker mail / history. The more audit logs, the better. Which of all the listed, landing spreadsheet is the one that has the least audit history. [12:46] asac: anyway, let's work on resolving any anomalies that are present. [12:46] sil2100, thats why i think it is gstreamer .... probably there is some user db in ~/ thats not properly set up on fresh installs ... none of the package changes after r73 can have caused this [12:47] nope [12:47] downgraded libgstreamer and added gstreamer0.10-ffmpeg and friends, and video still doesn't play [12:47] k [12:47] hm, maybe ogra_ is right, like... [12:47] the one thing we dont reflect are android changes ... [12:47] Maybe I should do a whole wipe of my system and then try 74 [12:48] were there any android uploads the last days [12:48] that could have broken the codecs on the android side ? [12:48] ogra_: do you know of a place where we could check that? [12:48] well, android is a package like any other :) [12:49] uploads should have been asked at the landing meetings [12:49] * ogra_ checks the changes ML [12:49] is that in /system/vendor/lib ? [12:49] I just did a apt-get dist-upgrade on my device even and no breakage ;p [12:50] not sure where the codecs actually live [12:50] looks like /system/lib [12:50] sil2100, that kind of points to android, there were two uploads yesterday, but neitrher should have any impact [12:50] hm hm [12:50] popey, rsalveti or jhodapp should be able to tell you what to look for [12:51] ogra_: thanks for your help! [12:51] ogra_: now, I think it's time to get back on holidays ;) We'll try to poke people around for this [12:51] heh, k [12:51] ogra_: at least now we know a possible direction in which to go! [12:52] http://paste.ubuntu.com/6214995/ ... thats the way to replace the android bits [12:53] (grabbed from the android deb you want to use) [12:53] you want to test 20131202-2236-0ubuntu2 [12:54] * ogra_ goes back to play with his home automation ... [12:54] sil2100: daker says #75 has the problem too [12:54] which narrows it down a bit [12:55] android changed between 74 and 75 [12:56] http://ci.ubuntu.com/smokeng/trusty/touch/ ... 20131211.2 in 74 vs 20131218.2 in 75 [12:56] Ok [12:56] So... [12:56] ok so i guess we dont have a good view on those changes? [12:56] We at least know that there is a 90% chance that it's android [12:56] (that should really use the package version ... not sure why we add a date there) [12:57] asac, trusty-changes has ... first one was just emulator related ... second replaces the opendjk version that we dont really use in the device images (also emulator related) [12:57] asac, but it could be that i.e. something in the toolchain changed or some such [12:57] doesnt sound like it could cause issues, but well. what else could it be [12:58] how can we test this? [12:58] right, someone needs to roll back the android part on a i.e. 76 image [12:58] see the pastebin above [12:58] just downgrading android package and run some tool to update the container? [12:58] garb the android deb, unpack and then do the steps from the past [12:58] kk [12:58] *paste [12:59] sil2100: you think thats good enough to try? http://paste.ubuntu.com/6214995/ [12:59] .... we did have eglibc 2.17 -> 2.18 changed recently. it was validated a lot, but there are subtle breakage here & there (e.g. i did have to fix cross toolchains, and java exposed a random / non-deterministic crash) [12:59] can eglibc down-grade be tried? [12:59] asac: I'm reading this up, I guess I can try this - but I would have to upgrade to the latest image [12:59] popey: or maybe you feel adventurous? What image does your device have? [13:00] xnox: when did that change happen? [13:00] you can theoretically omit the boot.img stuff from the paste btw ... [13:00] system.img is the important change you want to look at [13:01] (kernel and initrd didnt change in a while) [13:02] sil2100: 76 currently [13:02] xnox, http://people.canonical.com/~ogra/touch-image-stats/ nothing there shows a libc update ... (between 20131218 and 20131219.1) [13:02] i see a libneon though [13:02] * popey lunches [13:02] ogra_: neon only had a ftbfs fix on !armhf [13:02] but only in the last image [13:03] right [13:03] ogra_: I tried upgrading libneon as well, so damn [13:03] cant be it [13:03] well, try the android rollback [13:03] Will do [13:03] popey: can you try that ^ I'm upgrading now to something newer so that I can try downgrading [13:03] only line 6 to 15 [13:04] (and 1+2) === cjohnston changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cjohnston | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [13:05] ogra_: is this the package's deb we need to fetch? https://launchpad.net/ubuntu/+source/android/20131202-2236-0ubuntu2 [13:05] right [13:06] https://launchpad.net/ubuntu/+source/android/20131202-2236-0ubuntu2/+build/5325024/+files/android_20131202-2236-0ubuntu2_all.deb in fact [13:07] void QGstreamerPlayerSession::processInvalidMedia(QMediaPlayer::Error, const QString&) [13:07] void QGstreamerPlayerSession::stop() [13:07] 0:00:01.014680298 3457 0x1a9c600 WARN mirbufferpool mirpool.c:318:gst_mir_buffer_pool_release_buffer: delegate is NULL, rendering will not function [13:08] void QGstreamerPlayerSession::setSeekable(bool) false [13:08] AbstractPlayer: Internal data stream error. [13:11] sil2100: try what? [13:11] aha, android rollback [13:12] GST_DEBUG=*:3 mediaplayer-app /home/phablet/Videos/small.ogg --desktop_file_hint=/usr/share/applications/mediaplayer-app.desktop [13:12] is useful [13:15] Laney++ [13:16] not sure why reverting stuff is the first step in debugging [13:17] Laney: yeah, reverting individual pieces takes one into a _new_ state, not into a _known_ state (good/bad) [13:26] It wasn't the first step in debugging, I actually tried looking at the output but since I'm completely unfamiliar with the codebases, I moved on to a different way of finding the problem [13:26] Not to block on waiting for upstream to appear, as it's the upstream who should help in finding the bug essentially [13:34] ogra_: where can I get simg2img ? [13:34] android-tools-fsutils [13:35] ogra_: thanks again! [13:35] Since dlocate was silent today [13:44] \o/ [13:44] ogra_, popey, asac: after downgrading android, video playback works! [13:44] good [13:45] sil2100: nice one :) [13:45] rsalveti: ^^ [13:45] :) [13:45] now the question is why xnox' uploads with only unrelared changes can have caused this ... i guess you need to ask rsalveti [13:45] * ogra_ guesses there are additional in-tree changes that were picked up [13:46] Strangeness, but at least the component is locked on and we can poke specific people [13:47] ogra_: thanks for all the good advice, I wouldn't know how to downgrade android completely ;/ [13:47] now you do :) [13:51] ogra_: well no change rebuilds do trigger bugs. [13:51] ogra_: we've had that in the past. compare build-logs / build-deps used? [13:52] xnox, for android ? [13:52] ogra_: yeah. it fetches stuff from the archive at build-time. [13:52] i dont think the deps changed at all [13:53] though indeed someone should check the logs ;) [13:53] but thats up to someone not on vacation ;) [13:57] morning [14:16] ogra_, didn't the kernels change? [14:17] sergiusens, not within this week i think [14:17] ogra_: I'll try taking a look after lunch [14:26] ogra_: hey [14:26] ogra_: thought you were in vacation [14:26] what's up? [14:26] let me read the backlog [14:27] rsalveti, i'm working on my home automation system (well, mainly on the heating control) [14:27] so i'm around on the kbd ;) [14:28] rsalveti, looks like one of the recent android rebuilds broke video playback [14:29] rsalveti, rolling back to 20131202-2236-0ubuntu2 makes it work again [14:30] wow, interesting, wonder how [14:31] hm, maybe because the video compat side changed as well as it used the latest hybris version [14:31] which device? [14:31] yeah, i was thinking about hybris too [14:31] mako mainly ... not sure someone tested on maguro [14:32] yes, maguro too [14:32] i tested mako, om26er tested maguro [14:33] great, let me take a look [14:36] Back [14:36] rsalveti: it's on both mako and maguro, only tested those two [14:37] yeah, I should have enough now to debug it further [14:37] thanks a lot [14:38] rsalveti: thanks for taking a look into that! Let us know once you have some info ;) [14:39] sure :-) [14:53] hm, the hybris compat side didn't change comparing with the previous version [14:58] rsalveti, culd it be that something needs a rebuild against the new headers on the ubuntu side ? [14:59] shouldn't be needed, as it was working with previous android [14:59] just reverting android made it work, right? [14:59] I'll confirm this locally as well [15:00] yeah [15:00] rsalveti, following http://paste.ubuntu.com/6214995/(without the boot.img bits since that didnt change) [15:01] ogra_: right, thanks [15:15] can anyone else reproduce bug 1262711 ? [15:15] bug 1262711 in unity8 (Ubuntu) "Unity crashes with lots of music displayed in expanded music category in home scope" [Undecided,New] https://launchpad.net/bugs/1262711 [15:17] sil2100: ogra_: guess I found the issue, testing the fix in a sec [15:18] rsalveti: ! [15:18] rsalveti: great news! Could you shed some light on it when you have a moment? === alan_g is now known as alan_g|tea [15:19] sure === alan_g|tea is now known as alan_g [15:46] sil2100: ogra_: pushed the fix (libhybris), once in we need to bump the android package again and trigger a new image [15:46] asac, ^^^ [15:46] rsalveti: ACK! [15:47] rsalveti, awesome !! [15:47] will send the details over email [15:47] Excellent timing, I guess this way we'll have the regression out of the way for our evening image [15:47] rsalveti: big thanks! [15:47] sound there is good news? :) [15:47] thanks so much everyone! :) === gatox is now known as gatox_lunch [15:55] \o/ === cjohnston changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: doanac | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [16:09] woah ... tapping on activity in the browser takes a century to come up [16:10] the same for trying to go to the bookmarks inside the activity window .. nearly a minute [16:17] ogra_: yeah, that's a bit problematic, would be nice to have it fixed somewhen ;/ [16:17] yeah [16:18] really annoying [16:18] Our browser is not super usable [16:18] i rarely use bookmarks, but i'm just working on a heating control webapp over here and dusring development i sometimes use bookmarks for testing the webapp pages [16:18] really slows me down [16:31] * ogra_ curses perl [16:33] doanac: hiya! just need some guidance...we're trying to do a full smoke dry run on a feature branch manually... [16:33] and one of the tests listed is "sdk" [16:33] kgunn: okay. let me find that test [16:33] http://ci.ubuntu.com/smokeng/trusty/touch/mako/77:20131219.1:20131218.2/5572/sdk/ [16:33] i'd guess it means ubuntu-ui-toolkit, or is there an entry for that as well? [16:33] doanac: just need to know how to run it [16:33] xnox: there's a seperate test for ubuntu-ui-toolkit [16:34] kgunn: right. sorry. [16:34] np [16:34] kgunn: jenkins says: #NOTE: Automatically created from a script as part of daily smoke testing [16:34] lp:ubuntu-test-cases/touch [16:34] kgunn: https://jenkins.qa.ubuntu.com/job/trusty-touch-mako-smoke-sdk/ [16:35] xnox: ta [16:35] ../run_test.sh check-packages [16:35] or some such. [16:35] kgunn: looks like we grab this branch: http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/tests/sdk/setup.sh [16:35] lp:ubuntu-qtcreator-plugins/tests/device [16:36] and then run: ./qt_tests on everything [16:36] i think there's only a single "check-packages" test case for that [16:41] psivaa, asac: 77 finished so I'm rerunning some things now [16:50] robru: hello! Did you fill in a bug for that calendar-app test regression with new qtorganizer? [16:56] sil2100, i took that landing-plan bug and added a qtorganizer5-eds task for it.. [16:57] robru: could you poke the related people about that? [16:58] sil2100, ok [16:59] Meeting time! [17:01] * kenvandine joins [17:02] plars: hi! Want to join the meeting? === gatox_lunch is now known as gatox [17:02] cyphermox: ping! [17:03] balloons: do you want to attend our meeting? [17:03] sil2100: brt [17:23] rsalveti: hello! Did the android package bump already happen? [17:23] rsalveti: I'm asking because we would like to kick an image with the regression fixed soon [17:31] sil2100: why do you ask, instead of checking launchpad? [17:31] sil2100: uploaded 37 minutes ago, and still building, https://launchpad.net/ubuntu/+source/android [17:32] sil2100: and it needs to fully migrate to release pocket, this can be checked with $ rmadison -S android [17:32] (launchpad internal mirror is slightly ahead of on-disk, thus rmadison is when image builders see it published) [17:33] xnox: right! I over-hasted and didn't even look, thanks! [17:35] sil2100: anyone, one shouldn't trust people telling things, one should check that it's actually published =) cause from rsalveti point of view, he did everything and is free to go ;-) [17:35] s/anyone/anyway/ [17:36] xnox: right right, just thought that maybe he would poke us that everything is 'done' [17:36] But that's a valid point [17:36] sil2100: waste of engineering productivity time to poke each other. just like context less pings ;-) [17:38] xnox: understood, I got the point [17:44] plars: I see many many failures on mako in latest testing - did anything happen in the infra? [17:44] sil2100: yeah, I ran webbrowser tests again and they had a big problem - I'm rerunning *again* now [17:45] sil2100: we went from 3 to 27 failurs on the first rerun === om26er_ is now known as om26er [17:55] kenvandine: ok, I sent out the update e-mail - could you poke Bill and others about the dialer-app problem? And maybe, if you have a moment, retrace some info to fill in a bug about that crash? [17:56] robru: once you have a free moment, could you then poke someone regarding that flaky testo in ubuntu-clock-app on maguro? (as per e-mail) [17:56] And, of course, if both of you could somehow get ahold of someone responsible for calendar app and the flaky tests for maguro as well [17:56] robru: aaand that one additional flaky test introduced by qtorganizer5 [18:00] kenvandine: android has finished building, now just waiting for it to move to the release pocket - after it does, let's kick an image [18:00] ok [18:00] kenvandine: if I would EOD before that happening, could you have a look-out for rmadison -S android and once 20131202-2236-0ubuntu5 is out build an image? [18:01] kenvandine: would be really grateful ;) [18:01] sure! [18:02] Thank you! And thanks everyone for help with everything that was going on today ;) === alan_g is now known as alan_g|EOD [18:07] sil2100, already poked renato about qtorganizer. haven't read emails (just got out of last meeting right now) [18:11] ev, hey, if you're still around - if apport-cli fails when collecting data with "There is not enough disk space available to process this report.", do you know which FS does it mean? === plars changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: plars | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [18:29] sil2100: webbrowser looks ok now [18:40] sil2100: new android is already in [18:40] sil2100: can trigger an image anytime you want [18:44] rsalveti: ok, can you please trigger the image? :) [18:45] sure [18:45] kenvandine: rsalveti will kick the image it seems, so you're free here [18:45] Ok guys, see you tomorrow! [18:45] just triggered [18:45] Thanks again [18:45] image is now building [19:05] doanac: did you play with tribunal at all? [19:05] thomi: no [19:12] can someone look at the landing asks for click-apparmor and apparmor-easyprof-ubuntu? they are very low risk. apparmor and dbus would also be nice [20:26] plars, can you give me the link to the gallery-app failure ? [20:26] I have a few spare cycles to investigate into that [20:27] om26er: sure, one sec [20:29] om26er: http://q-jenkins:8080/job/trusty-touch-mako-smoke-gallery-app-autopilot/85/ is the one on mako from this morning, http://q-jenkins:8080/job/trusty-touch-maguro-smoke-gallery-app-autopilot/92/ most recent on maguro [20:30] plars, q-jenkins ? [20:31] om26er: yes, or you can replace all that with jenkins.qa.ubuntu.com if you need the public-facing jenkins copy [20:32] plars, thanks, q-jenkins didn't work for me. [20:40] image built fine, let me give it a shot [20:41] om26er: I just saw olivier's email, looks like he's looking at the gallery-app failure already [20:42] plars, ok, great. [20:56] popey: yup, just confirmed that video playback is now working again with 78 [21:00] rsalveti: sweet, will update [21:05] cihelp can someone trigger a daily-build PPA build of unity-scope-click please? [21:07] dobey: let me look, do you happen to have the url? [21:08] plars: i don't know how the daily buids are triggered. given the uploader is listed as "PS Jenkins Bot" i'd presume there is some jenkins job for it [21:10] i don't know what url it'd be though [21:10] fginther: do I just need to kick the job on q-jenkins for that? or is there a preferred way to do it by hand? [21:15] plars, you just need to trigger the job, one moment [21:16] fginther: ack, it looks like someone got it already, for future reference, it would have been just http://q-jenkins.ubuntu-ci:8080/job/cu2d-click-package-head-1.1prepare-unity-scope-click/ directly? [21:17] plars, you need to trigger the stack job, in this case http://q-jenkins.ubuntu-ci:8080/job/cu2d-click-package-head/ [21:17] fginther: ok, that's what I was afraid of.. it wasn't clear to me how far up the chain I needed to go to start it [21:18] plars, right, cu2d-click-package-head is at the top of the chain [21:18] fginther: doesn't that pull in more than we want? what are the conditions for when it is/is not appropriate to run this by hand? [21:20] plars, It would pull rebuild any other packages that are due to be built, but I don't think there is anything adverse to running the whole thing [21:20] fginther: cool, thanks [21:20] plars, something weird is happening though, lots of jobs have been running for too long [21:21] fginther: yeah, I see http://q-jenkins.ubuntu-ci:8080/job/cu2d-click-package-head-2.1build/264/ has been running for 4 min [21:21] yay! video playback works again :) [21:23] fginther: but it looks like it normally takes a bit of time when it fails [21:24] kenvandine, can you explain what's going on here: http://q-jenkins.ubuntu-ci:8080/job/cu2d-hud-head-2.1build/385/console [21:24] kenvandine, hud has been building for 15 hours [21:24] kenvandine, is it supposed to wait on powerpc? [21:26] yay, the ppa build works again [21:26] so looks like the test failure was due to network access :( [21:27] fginther, it's not supposed to block on ppc [21:28] fginther, maybe it waits for ppc to finish, but doesn't fail if ppc fails to build [21:28] kenvandine, I thought the tools were set to completely ignore them. In this specific case the ppc would has already failed [21:29] it's in a dependency wait state [21:29] on libgsettings-qt-dev [21:29] ah... right [21:29] shouldn't do that [21:31] fginther, maybe because hud is built for ppc in the archive [21:32] kenvandine, ARCHS_TO_EVENTUALLY_IGNORE = set(['powerpc', 'arm64', 'ppc64el']) [21:32] that was recently changed to add ppc64el, but powerpc has been there for a while [21:33] i wonder what EVENTUALLY means there :) [21:33] ignores it after some timeout? [21:33] kenvandine, right :-) [21:34] comment says: "an arch we will ignore for publication if latest published version in dest doesn't build it" [21:34] ah [21:35] so it sound like if the package arch is in the archive, we wait for it, others not. The next line has ""ARCHS_TO_UNCONDITIONALLY_IGNORE = set(['arm64', 'ppc64el']) [21:35] so the latest in dest does have powerpc [21:35] hud | 13.10.1+14.04.20131205-0ubuntu1 | trusty | source, amd64, arm64, armhf, i386, powerpc [21:35] kenvandine, I also see it waiting on powerpc builds in prior jobs [21:35] kenvandine, so this appears to be rooted in the broken dependency [21:36] i'm confused though... looking at the history [21:37] that build dep wasn't added since the latest build in the archive [21:37] but there are no ppc builds of libgsettings-qt-dev [21:38] i was wrong [21:38] that was just added [22:01] kenvandine, shouldn't there be a gsettings package in the daily-build ppa? === plars changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: cihelp | Landing instructions: http://paste.ubuntu.com/6292280/ | Known issues: - [22:04] kenvandine, gsettings-qt in the settings stack which has a dependency on the sdk stack. the sdk stack doesn't build because of missing gsettings-qt [22:04] recipe for deadlock? [22:07] kenvandine, can we do a direct upload of gsettings-qt? [22:08] we should be able to do a build [22:09] fginther, if we do a force build of gsettings-qt in the settings stack [22:09] it will go in the ppa [22:09] the stack is running now [22:09] i need to head out... can you trigger a build? [22:09] just force it to build gsettings-qt even if prepare doesn't need it to [22:09] kenvandine, I can try [22:09] cool [22:09] that should do it [22:10] i need to leave though... good night! [22:10] kenvandine, bye