[00:00] I haven't yet [00:00] wainting on livecd-rootfs [00:00] bbl, dinner [00:00] and potentially mediaplayer-app [00:00] it's building now [00:00] rsalveti: what's missing? [00:00] lool: just gst-plugins-bad1.0 [00:00] that I just pushed [00:00] libhybris is in? [00:01] yeah it is [00:01] yup [00:10] lool, rsalveti with mir it seems I won't be able to run a large suite on maguro terminate called after throwing an instance of 'boost::exception_detail::clone_impl >' [00:10] what(): buffer allocation failed [00:10] sergiusens: ouch [00:10] sergiusens: is that mm stack? [00:10] plars, doanac hey, next filemanager is going to be click [00:11] lool, that's unity8 [00:11] but that's using all memory? [00:11] lool, I'm just running the filemanager tests again to see if the input stuff is fixed [00:11] that would be cool [00:12] lool, it's seems it is btw [00:23] lool, can you ping me when the build starts? [00:23] I'm off for a bit [00:23] will ping here [00:23] as usual [01:03] == Build #97 == [02:27] sergiusens: just seeing the message. i'll update file-manager now [02:27] doanac, great [02:29] * sergiusens goes to bed [04:24] fginther, thomi - can anyone work out what is going wrong with the autolanding for https://code.launchpad.net/~robert-ancell/lightdm/lp-1231841/+merge/191078? Looking at http://10.97.0.26:8080/job/lightdm-autolanding/ it just seems to be running the same job every 15 minutes [04:25] #169 is another autolanding, but I suspect the queued 170 will be that merge again [04:29] robert_ancell: this gives me a hint: http://10.97.0.26:8080/job/lightdm-autolanding/168/console [04:30] Waiting for the completion of generic-land generic-land #18476 completed. Result was FAILURE [04:30] which failed because: bzrlib.errors.LockContention: Could not acquire lock "(remote lock)": bzr+ssh://ps-jenkins@bazaar.launchpad.net/~lightdm-team/lightdm/trunk/ [04:30] robert_ancell: perhaps you can make sure that the launchpad repository is sane? and if not, ask in #launchpad-ops? [04:30] thomi, aha [04:31] thomi, yeah, I'll check if there's a lock on it [04:31] but, I'm just guessing here - I don't really know anything about this system [04:31] but I guess the CI team are all asleep [04:32] I can't see anything holding the lock [04:32] oh well, I guess I'll just be wasting resources overnight until someone notices [04:42] hey Mirv, how are you? last day before holidays? :) [04:51] yes, but where did didrocks drop :) [05:10] didrocks: so, yes :) [05:11] I guess all hands testing mir is welcome? [05:13] Mirv: exactly! [05:13] let me fw you an email [05:14] Mirv: so, most of the time, I had to merge manually the other branches (the deps) [05:14] Mirv: and kick a build (maybe Mir already built with the latest branch merged, to check) [05:14] Mirv: do you think you can handle that or are you busy with something else? [05:16] didrocks: if I'm reading correctly, mir + platform have already been rebuilt, probably by lool. unity8 is failing to build because of a dependency mismatch, and I'm not sure how to resolve that. [05:17] Mirv: want me to have a look? [05:17] Mirv: I guess the rebuild was automatgic for mir + platform [05:18] but at least, they were merged [05:18] ok, I think unity8 FTBS because of unity-mir [05:18] this one doesn't use the ppa [05:18] didrocks: yes, I was just guessing the same, and that is not merged [05:18] let me push it manually [05:19] there was a mention in the landing plan of retrying needed for the mir build, otherwise probably automatic [05:20] oh right [05:20] so, lp:unity-mir merged [05:20] can you rekick a build in cu2d? [05:20] doing [05:21] seems all the rest built, indeed [05:23] platform-api has one or two more commits, anyway, I think we'll retest all AP tests with it [05:36] lool: seems latest image is still missing the gst-plugins-bad1.0 upload I did [05:37] rsalveti: when did you upload it? [05:37] * didrocks looks [05:38] didrocks: quite a few hours ago, before lool triggered the latest build [05:38] rsalveti: yeah, probably it wasn't fully published yet [05:38] I asked him to wait for it to be published, but maybe it took more time than expected [05:38] rsalveti: we'll rekick an image probably in 4h [05:38] so you will have it then [05:38] awesome [05:38] thanks for the head's up :) [05:43] Mirv: so, unity-mir is published [05:43] Mirv: mind on latest image flashing using mir + unity-mir + platform-api ? [05:43] I'll do the same here [05:44] and dogfooding [05:44] (then, we'll run the AP tests on it) [05:44] I'm still rebuilding the unity8 against it now [05:59] Mirv: you don't need unity8 I guess TBH [05:59] didrocks: ok, it just annoyed that it didn't build [06:00] so upgrading without [06:00] yeah, it should build [06:00] now it does, it just needed the unity-mir to be also published [06:04] definite performance improvement. scrolling not smooth still, but eg. indicators are back to surfaceflinger smoothness. [06:05] yeah, it seems way better [06:05] Mirv: going to give some AP trials? [06:05] didrocks: yes, starting with uitoolkit to see if it still explodes [06:12] still gets Killed, trying individual tests now [06:28] ev, could you have a look at bug 1239811 [06:28] bug 1239811 in whoopsie-preferences (Ubuntu Saucy) "whoopsie-preferences crashed with SIGSEGV in _IO_new_fclose()" [High,New] https://launchpad.net/bugs/1239811 [06:41] E: Method bzip2 has died unexpectedly! [06:41] I wish someone would be able to debug what happens when running uitoolkit tests... [06:51] how does one install autopilot tests for the packages that have been converted to click? [07:23] psivaa: I tried to relaunch the tests on our infra for unity8, but it didn't work [07:24] didrocks: let me take a look [07:24] psivaa: look at unity8 (I'm relaunching ubuntu-ui-toolkit now) [07:24] psivaa: it failed first, and I relaunched it and failed as well [07:27] didrocks: it timed out suggesting a hang somewhere, when such a hang happened with notes test on maguro i saw http://pastebin.ubuntu.com/6219418/ in the kernel logs [07:27] didrocks: let's see if similar issue occurs with this mako [07:27] psivaa: oh, but we had an apparmor change for notes-app [07:30] didrocks: looks like the the notes test on 97 still had this issue. i'm re-running it. let's see if we see the same log [07:31] didrocks: in todays image the sudoku app was dropped/added with the wrong name, do you know what caused that? [07:31] psivaa: I don't see that in dmesg [07:32] popey: it's a click package, not really sure about those, sergiusens or ogra_ should know [07:32] popey: you will tell that during the meeting? [07:32] didrocks: heh, sergiusens asked me ☻ [07:32] ok [07:32] ahah ;) [07:32] ok ogra_ then! [07:33] psivaa: do you see anything in this run: http://10.97.0.1:8080/job/saucy-touch_mir-mako-smoke-unity8-autopilot/23/artifact/clientlogs/dmesg.log ? [07:35] didrocks: no, but i did not see the PVR alloc errors in dmesg earlier as well. for some reason only saw that in kern.log ( i could have missed anything) [07:35] psivaa: where do you see kern.log, I don't see it collected in http://10.97.0.1:8080/job/saucy-touch_mir-mako-smoke-unity8-autopilot/23/ for instance [07:36] didrocks: http://pastebin.ubuntu.com/6239493/ is happening with maguro right now.. so the test has just hung [07:36] didrocks: i logged into the device [07:36] didrocks: log in to phoenix and then connect to the device [07:36] ah, ok, we need to get something better than logging I guess :) [07:36] didrocks: agree [07:36] psivaa: is there a wiki page? I don't know about that one [07:45] * didrocks rebooting [07:58] thostr_: hey, in landing ask, request 201 and 203 are the same, right? [07:59] didrocks: yes... [07:59] thostr_: ok, killing one [08:08] ev: do you know if "powerd-cli display on" is executed before starting the suite in utah? === alan_g|EOD is now known as alan_g [08:20] rsalveti: Ok; I thought I had waited until it was in release pocket [08:21] rsalveti: indeed [08:21] rsalveti: my bad [08:22] Mirv: https://wiki.ubuntu.com/Touch/Testing#Running_Click_tests [08:23] Mirv, didrocks: So I had trouble getting the mir stack in PPA; eventually I got all but unity-mir in PPA manually [08:23] oh, that way [08:23] didrocks: cu2d kind of "lost" qtpowerd somewhere [08:23] lool: unity-mir is now there as well [08:24] lool: we handle the end of mir [08:24] regression on #97, sudoku and dropping letters to not start [08:24] didrocks: it generate a rsync file, but that never resulted in an upload in unapproved [08:24] so apt-get install libunity-mir1 libplatform-api1-hybris libubuntu-application-api-mirclient1 libubuntu-application-api-mirserver1 libubuntu-application-api1 libubuntu-platform-hardware-api1 libmirclient3 libmirplatform libmirprotobuf0 libmirserver7 [08:24] gets you updated mir [08:24] optionally one may also update the unity8 [08:24] lool: I see it, did you refresh the whitelist? [08:24] I dont know how to do that [08:24] didrocks: it's .3 that got lost [08:24] lool: checked the FAQ? [08:24] lool: https://wiki.ubuntu.com/DailyRelease/FAQ#Adding.2BAC8-removing_components_to_a_stack [08:24] The archive admin needs to, after checking that lp:cupstream2distro only adds component for filtering we want: [08:25] popey, for the manifest creation i only run "click list" on the finished image ... to install click packages we read a file that sergiusens creates to get the list [08:25] didrocks: but I'm not ~ubuntu-archive [08:25] didrocks: I pinged cjwatson about it, and infinity tried to pull it I think [08:25] ogra_: yeah, found the commit thanks to dpm, have mailed upstream [08:25] good [08:25] lool: I saw it pulled this morning, but I guess that was after you published it? [08:25] maybe [08:25] in that case, the sync was rejected [08:26] how do you see that? [08:26] hmm, my nexus 7 has been running nearly 3 days and has init using 75% RAM [08:26] PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND [08:26] 1366 phablet 20 0 1039m 730m 488 S 0.0 75.0 12:37.24 init [08:26] popey: sounds like the same bug as we have on mako and maguro [08:27] image 94, so needs an update [08:27] lool: I changed the email to me to MAILTO=ubuntu-unity@lists.launchpad.net [08:27] popey: we only applied a workaround on maguro to limit the number of events in udev and on all devices to limit unity8 traffic [08:27] popey: but the leak remains [08:27] lool: however, I don't see any mail (where we should receive), maybe launchpad not permitting that? [08:28] didrocks: So I also think the changelog is screwed up, but not sure why [08:28] There are no mailing list messages requiring your review. [08:28] lool: well, first one wasn't cu2d issue, just not following the FAQ [08:28] let's look at the changelog [08:28] didrocks: .2 got committed to bzr, then despite .3 not making it to archive, there was a .3 changelog commit! [08:28] ah [08:28] lool: right [08:28] actually that might be the issue [08:28] lool: because the filtering is in the archive copy [08:28] the revisions are wrong [08:29] links? [08:29] didrocks: I did not follow the faq but I asked an archive admin to pull the config amounting to the same [08:29] lool: maybe the FAQ should mentionned to do that before publishing [08:30] publishing also failed due to the branch being stacked for some reason [08:30] Colin provided me with a rune to fix it up in bzr [08:30] didrocks, lool: I'll join the call in a moment [08:31] sil2100: can you try to be on time in the near future for our calls please? ;) [08:35] didrocks: ok! ;) [08:38] ev: http://10.97.0.1:8080/job/saucy-touch_mir-mako-smoke-unity8-autopilot/24/ [09:09] lool, where was the PPA removed ? i still see it in meta [09:10] ah, only from the build scripts yet ... that needs cleanup [09:10] ogra_: well I flashed #97 and at least it's true that it's not enabled [09:11] Mirv, right, but germinate still uses it when processing the seed [09:11] * ogra_ will clean that up after the meeting [09:12] ogra_: please flushes :) [09:13] Upgrading the device, start running unity8 tests as soon as it finishes [09:13] landing 235 [09:17] Mirv: sorry i missed it, did you want one of us to run uuitoolkit on 97? [09:17] Mirv: i just got http://pastebin.ubuntu.com/6239756/ on maguro [09:21] (so we found the reason for dropping-letters to be broken, and it was at the time of the removal of share-app) [09:21] needs a seed update to pull in the hud qml bindings now [09:22] psivaa: anyone who could understand what's going on in there, but maybe on mako since it seems you're not hitting the problem on maguro [09:22] psivaa: on mako I constantly get "Killed" eventually when running phablet-test-run ubuntuuitoolkit (the whole suite at once), kernel log shows processes getting killed (http://pastebin.ubuntu.com/6239756/) and eventually the device reboots itself [09:23] psivaa: I believe it's the same as this general slowdown with Mir, but can be gotten with the uitoolkit test suite easily [09:23] 阿里单位额五日俄方 [09:23] Mirv: what cras files do you get? [09:23] Ok, that was unexpected [09:24] What I wanted to say...: I noticed yesterday that the clock app looks strange, and you cannot set an alarm because the layout is out of place [09:24] asac: nothing (at the time when that problem occurs, unity8 etc otherwise) [09:27] sil2100, it is fixed on 97 [09:28] Awesome [09:29] Mirv: ok, it doesn't look like we are seeing that issue in the smoke runs. [09:29] https://jenkins.qa.ubuntu.com/job/saucy-touch_mir-mako-smoke-ubuntu-ui-toolkit-autopilot/19/artifact/clientlogs/utah.yaml/*view*/ [09:31] psivaa: I know. interestingly I get it also if I try to execute single tests but during the same boot. [09:31] psivaa: the smoke runs of course use utah etc so it's different [09:35] Mirv: ack, that could be the difference === vrruiz_ is now known as rvr [09:51] anyone knows what is usually happenining if we see something failing like this: https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1239646/comments/4 ? [09:51] Ubuntu bug 1239646 in Ubuntu UI Toolkit "CI fails most tests on UITK trunk" [Critical,Confirmed] [09:51] i had those as well (e.g. process not starting or crashing right away) [09:54] in get_autopilot_proxy_object_for_process [09:54] dbus_bus=dbus_bus [09:54] .... [09:54] dbus gone or blocked ? [09:54] didrocks: https://code.launchpad.net/~lool/mediaplayer-app/qtpowerd-keep-screen-on/+merge/191061 [10:18] didrocks: do you know when ogra is back? [10:19] ? [10:19] ogra_: that happens on CI. I don't know where it is executed. Do you know if that environment is different from the CI environment for other projects? [10:19] didrocks: I just wanted to tell him that I took over the -meta update to drop the PPA since I was uploading it [10:19] lool: now? ;) [10:19] ogra_: uh [10:19] hey ogra_! [10:19] ogra_: my completion didn't work for some reason [10:19] ogra_: lool has a message for you :p [10:19] lool, i have seed changes pending too [10:19] * didrocks runs ;) [10:19] didrocks, tell him i have seed changes to merge as well [10:19] heh [10:19] ogra_: Ok; I've uploaded -meta without -ppa and fixing the accidental hud drop from share-app and fixing some poppler seed error [10:20] just wanted to do a quick call tests on maguro before i fiddle with that [10:20] lool: high time for you to take some rest! [10:20] ogra_: want me to do another upload? [10:20] lool, accidential hud drop from share-app ? [10:20] share-app should not be on the image [10:20] (we removed it twice now) [10:21] lool, for the seed change we have to wait for the new location service package [10:21] ogra_: ok, will let you handle it then [10:22] ogra_: yes, we unseeded share-app and that dropped the hud bindings [10:22] which were used by some clicks [10:22] which hud bindings ? [10:22] because it's our platform [10:22] ogra_: we (SDK team) don't have control over the environment where we run the UITK autopilot tests for CI. So what can we do if dbus is gone or blocked? [10:22] ogra_: qtdeclarative5-hud1.0 [10:22] 1.0 is supposed to be completely gone by now [10:22] ogra_: or first, how do we find out whether that is the case? [10:22] lool, right, thats not wanted in the image [10:22] lool, apps need to port to 2.0 [10:23] ogra_: where is 2.0? [10:23] lool, talk to ted before pulling anything in again [10:23] should be shipped by default [10:23] lool, libhud2 [10:23] and libhud-client2 [10:24] ogra_: sorry, this is what dropping-letters was using [10:24] ogra_: I dont see the new hud bindings though [10:25] ogra_: so you're saying libhud-qt should be removed from archive? [10:25] lool, no idea, but i know that libhud2 replaces the old 1.0 stuff [10:26] ogra_: libhud2 is depended upon by libhud-qt1 [10:26] lool, well talk to ted ... it took some effort to get that right, i dont want to mess it up [10:27] ogra_: do you have a reference I could check? [10:27] qtdeclarative5-hud1.0 is definitely dead and supposed to not be on the image [10:27] ogra_: ted is only coming up in some hours [10:27] IRC logs but i'm not sure in which channel [10:27] might have been #phablet [10:28] ./unity8/qml/HudClient/libHudClientQml.so [10:28] that's what unity8 uses [10:29] lool, all i know is that there was quite some fuss to get it gone [10:29] ogra_: what I know is that dropping-letters depends on it :-) [10:29] lool, well, then fix dropping letters [10:29] ogra_: it's not mine [10:29] hud1.0 is deprecated [10:29] so leave it to upstream then :) [10:30] lets just wait for ted to clearify [10:31] talking to upstream is exactly why I want the details; I dont know what we're moving to [10:31] I guess this might be via unity or something [10:31] and it also looks like we should drop a source [10:33] I wrote to ted [10:34] good [10:34] 97 is good on maguro btw [10:35] * ogra_ doesnt like the new ringtone [10:35] ogra_: if you need help, I think sil2100 can help you to test the unity8, unity-notifications and ubuntu-touch-session [10:35] * sil2100 runs unity8 tests on 97 mako + new unity8 [10:35] sil2100: can you ensure already that everything we wanted has landed? ^ (check with Saviq) and that it's built in the ppa? [10:35] sil2100: ah, so the binaries are ok? [10:35] sil2100: I think you need ubuntu-touch-session from ogra_ [10:36] sil2100, if you use the latest unity8 you also want the ubuntu-touch-session change [10:36] ogra_: ok, where can I get that from? [10:36] let me roll you a package, one moment [10:36] didrocks, others: the release team just applied a britney block to all source and we'll need to use that branch cjwatson created to provide overrides for individual uploads [10:37] * didrocks needs to recheck syntax === ev changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: - Type: "@ci" for help | Britney is set to block all uploads, use lp:~ubuntu-touch-release/britney/hints-ubuntu-touch to override | Known issues: - [10:39] sil2100, http://people.canonical.com/~ogra/ubuntu-touch/ubuntu-touch-session_0.83_all.deb [10:39] * ogra_ is afk for a moment+ [10:39] ogra_: thanks! [10:39] * asac reminds folks how risky ubuntu-touch-session uploads were in the past :) [10:40] sil2100: please don't start unity8 stack now [10:41] ogra_: want me to push location-service? [10:41] ogra_: oh, there is the seed change, right [10:41] I'll let you do that :) [10:41] think about the hint then ;) [10:42] Mirv: ok ;) [10:43] Mirv: tell me once we can do that, as there was a commit landed a few minutes ago that would be nice to have as well [10:44] sil2100: I'll ping you then [10:47] Mirv: ok, anyway all is landed already so just this one commit I guess we'll fetch [10:48] 12:46 < tvoss> asac, [10:48] http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/97:20131015:20131015/4730/notes-app-autopilot/ reports 4 crashes in the overview, although nothing crashed [10:48] 12:46 < tvoss> asac, seems like crash "detection" is a wildcard *crash* [10:48] ev: plars: doanac: ^^ :() [10:48] lets fix that [10:50] looking into it [10:51] asac: tvoss is correct. I'll arrange a fix. [10:51] (correct as far as the code is concerned) [10:53] thx [10:57] published mir, unity-system-compositor, platform-api, unity-mir [11:01] sil2100: I think you can rebuild unity8 now ^ [11:03] Mirv: ^ ? Can I? [11:04] sil2100: yes you can [11:04] sil2100: "ping" :) [11:04] ;) [11:08] lool, ogra_ https://bugs.launchpad.net/dropping-letters/+bug/1221344 [11:08] Ubuntu bug 1221344 in Dropping Letters "remove usage of qtdeclarative5-hud1.0 " [Undecided,New] [11:14] sergiusens, thx ! [11:17] lool, in that light ^^^ can we please remove it again before rolling an image [11:18] ogra_, also, I have a 99% assurance that this is why most tests fail on maguro ;-) terminate called after throwing an instance of 'boost::exception_detail::clone_impl >' [11:18] what(): buffer allocation failed [11:18] boost ? [11:18] ogra_, buffer allocation failed [11:18] ogra_, that's unity8; after that it restarts [11:20] sergiusens, did you point Saviq to it yet ? [11:20] ogra_, no, saw that late last night; just recapping today [11:20] telling Saviq, is easy ^ [11:21] heh [11:21] * sergiusens goes to #ubuntu-unity [11:25] ogra_: can we get the dropping letters fix quickly? [11:25] lool, no idea, but that lib cleanup was done a while ago, i dont want to regress [11:26] (i didnt notice you had already uploaded) [11:27] apps are not bound to release, the image is ... they can hapily fix it asyncronous [11:27] hum, sil2100, when upgrading from the ppa, you didn't get any issue with unity8, [11:29] didrocks: what issues? [11:30] I had no issues, I'm running unity8 from daily-build - not the latest, since that one is still building [11:30] sil2100: like package file override [11:30] hm, no [11:30] sil2100: do you have http://bazaar.launchpad.net/~unity-team/unity8/trunk/revision/461 ? [11:31] didrocks: http://paste.ubuntu.com/6240141/ [11:31] didrocks: this was my log from installing [11:32] sil2100: with that commit included? [11:32] oh Replacing files in old package ubuntu-touch-session ... [11:33] Yes, with this commit, as this unity8 was built 8 hours ago I guess [11:34] didrocks, sil2100, hmm, looking at the MP i think there ios something missing in the unity8 merge [11:34] unity8 was supposed to call SIGSTOP once it is done setting up sockets ... to make upstart pick that event up there should be an "expect stop" in the upstart job [11:34] i dont see it in the MP [11:35] i pinged Saviq in -touch already about it [11:35] ogra_: ok, keep us posted :) [11:35] I think that doesn't stop sil2100 from testing :) [11:35] well he wont test the whole fix [11:35] keep an eye open for maliit crashes when restarting the shell [11:36] ogra_: right, but at least, we'll just have one more commit to test :p [11:36] indeed [11:36] * ogra_ needs to reboot [11:37] I filed bug 1239815 yesterday, probably the font should be seeded for the release or the language removed from the list [11:37] bug 1239815 in ubuntu-system-settings (Ubuntu) "Text not displayed properly in Chinese" [High,New] https://launchpad.net/bugs/1239815 [11:38] :| [11:40] lool, ogra_ https://code.launchpad.net/~sergiusens/dropping-letters/nohud/+merge/191164 [11:41] ogra_: hum, not sure about chinese, do you know what we need to seed? [11:44] sergiusens: approved [11:48] didrocks, i thought we have everything [11:49] * ogra_ expected the langpack-base package to depend on the right fonts [11:50] dpm: do you know this? ^ [11:50] (hey!) [11:51] we have language-pack-zh-hans and language-pack-zh-hans-base as well as the gnome equivalents for both [11:51] hi didrocks, I haven't been involved in langpacks this cycle, perhaps pitti or seb128 might know. For Chinese on the phone I thought we need fonts-droid and that it was already seeded? [11:52] ogra_: weird… [11:52] seb is travelling [11:52] didrocks, or happyaron would definitely know [11:52] ok, thanks dpm [11:52] root@ubuntu-phablet:/# dpkg -l |grep fonts-droid [11:52] ii fonts-droid 1:4.2.r1-2 all handheld device font with extensive style and language support [11:52] it is definitely there [11:53] ogra_: language-selector uses /usr/share/language-selector/data/pkg_depends at least on desktop to determine various fonts to install [11:54] well, looking at that file on my precise desktop there are no droid fonts listed [11:54] fn:zh-hans::ttf-wqy-zenhei [11:54] fn:zh-hans::fonts-arphic-uming [11:54] fn:zh-hans::ttf-arphic-ukai [11:54] thats what i see [11:55] indeed [11:55] can someone check on a saucy desktop ? [11:55] i only have raring and precise around [11:55] ogra_: I'm on saucy, same here, those fonts ^ and no droid in the file [11:56] right, so i suspect we want to seed these [11:56] possibly yeah [11:57] oot@ubuntu-phablet:/# apt-cache show fonts-arphic-ukai|grep ^Size [11:57] Size: 10274088 [11:57] root@ubuntu-phablet:/# apt-cache show fonts-arphic-uming|grep ^Size [11:57] Size: 7930652 [11:57] root@ubuntu-phablet:/# apt-cache show ttf-wqy-zenhei|grep ^Size [11:57] Size: 9184588 [11:58] not actually small [11:59] ogra_: well, maybe worth it for the demos [12:00] i'd like to be sure that we need them [12:00] * didrocks tries what pitti is suggesting [12:00] when the oom killer will be kind to not kill my apt-get [12:00] grrr, let's reboot [12:02] ok, not that one, next now :) [12:04] not ttf-wqy-zenhei [12:04] next [12:04] \o/ [12:04] fonts-arphic-ukai is the right one [12:05] great, lets seed it [12:05] is CI for the location service change done already ? [12:05] * ogra_ curses pulseaudio ... [12:05] ... another reboot === alan_g is now known as alan_g|lunch [12:07] sergiusens: Can you ping me when dropping-letters is updated so that I remove the seeding? [12:08] ogra_, didrocks: 2 failed, 22 success on unity8 [12:08] ogra_, didrocks: and from the looks of it, those are some flacky test I guess [12:08] sil2100: can you get Saviq's feedback on the 2 failing? [12:08] didrocks: sure [12:08] sil2100: well, flacky teste need to be fixed ;) [12:08] thanks! [12:08] ogra_: you are handling the location service, I think it's building right now though [12:10] new Mir components in release pocket now [12:11] didrocks, great, thanks [12:11] (checked with rmadison) [12:13] Mirv: I hinted them! excellent that they migrated, thanks! [12:14] ogra_: location-service is built in the ppa, you can test it [12:15] Do you guys want this new system-image? Looks like its autopkgtests have never passed [12:15] * ogra_ isnt really sure what to test or how ... since nothing uses the location service ... i guess as long as the indicator still shows up it is fine === gatox is now known as gatox_bbl [12:16] Which actually looks like it might just be a missing test dependency on python3-psutil ... [12:19] cjwatson: it's hinted (blocked) by lool [12:19] cjwatson: we want barry to test it with the ui first [12:20] ogra_: yeah, that was my (very limited) test before [12:20] ogra_: and looking at /var/crash :p [12:20] indeed :) [12:20] ogra_: tell me if/when I can publish it [12:23] cjwatson: so we should reupload with a dep on this in the autopkgtests I guess [12:23] cjwatson: we're pushing testing of system-image towards EOD due to other big things landing today (unity8 and mir main) [12:23] lool: Yeah, I was going to run the tests locally first [12:23] We can force it past the tests, but I'd rather fix them [12:23] cjwatson: +1 [12:24] cjwatson: also, barry ran them often in trunk and I think they are run as part of packge build, but would want to keep autopkgtests working on this critical package [12:25] Yeah, I expect they work fine as unit tests [12:26] sil2100, which one? I don't think we have flaky tests, we might have crashes on startup, though [12:26] sil2100, which tests and what output please? [12:31] * didrocks out for a run, back in an hour [12:54] Saviq: I pasted the tests on ubuntu-touch, I'll paste the output in a moment [12:55] I've uploaded system-image with the autopkgtest fix [12:56] cjwatson: thanks === alan_g|lunch is now known as alan_g [12:59] Saviq: I fetched it from the yaml file, but should be readable: http://paste.ubuntu.com/6240421/ [13:01] sil2100, yeah, crashes on startup [13:01] sil2100, I couldn't track those down :/ [13:01] sil2100, backtrace is just ?? ?? ?? [13:01] :| [13:02] Bleh, well, those don't seem to be related to the particular testcases, as I had different failures before (also 2) [13:09] morning [13:10] mornin' [13:12] sil2100, Saviq, so any outcome ? should i upload the ubuntu-touch-session change ? [13:12] ogra_, yes [13:12] k [13:12] ogra_: here it looks ok [13:12] * ogra_ does so [13:12] ogra_, anything we need next will be unity8-only [13:12] The tests are generally good, dogfooding still but don't see any issue [13:12] Saviq, yeah [13:12] Saviq, didrocks: you guys fine if I publish unity8 as well? [13:13] is there some special trick to re-trigger a CI build for somethig that was already built ? [13:13] https://code.launchpad.net/~thomas-voss/location-service/refactor-packaging/+merge/189878 would neeed a new build [13:13] lool, latest dletters is in http://people.canonical.com/~ubuntu-archive/click_packages/click_list [13:14] lool, test it if you want before removing the seed entry [13:14] sil2100, +10000000 [13:14] sergiusens: could you test it? [13:14] I'm busy with the setcap stuff [13:14] * lool was trying to take this day easy [13:14] seems release decided otherwise :-) [13:14] lool, I was looking for a 3rd person.... let me ask popey! [13:15] 3rd guy's a charm [13:15] hmm? [13:15] see [13:15] asac: lool: anyone looked at the bump in failures after music-app went to a click package? [13:15] ogra_, why do you need it to rebuild? and anyway it's building right now, and will be built once more during -autolanding [13:15] and good morning :) [13:15] plars: go for sergiusens [13:15] plars: not that I know of; sergiusens tested it before the switch though [13:15] ogra_, new commits in an MP == new build [13:15] Saviq, because it was approved, then issues were found and fixed ... [13:15] ah, cool [13:15] lool, asac: maybe just a missing dependency or something [13:15] plars: he knows the click transition details and would be nice to have a good story on these (if not already) [13:15] plars: morning; also filemanager switched to click [13:15] i didnt know thats automatic [13:15] plars, I'm looking at that actually [13:15] ogra_, just top-approve again [13:16] i did [13:16] sergiusens: http://people.canonical.com/~ubuntu-archive/click_packages/com.ubuntu.dropping-letters_0.1.2.2.42_all.click that one? [13:16] plars, the error on jenkins is really strange [13:16] perfect [13:16] so just waiting then [13:16] popey, yes [13:16] ogra_, yup [13:16] * ogra_ gets coffee [13:16] sergiusens: ok [13:16] file:///build/buildd/unity8-7.82+13.10.20131011.2/Shell.qml: File not found ^M [13:16] ohoh [13:17] lool, that's a mock-only issue [13:17] sergiusens: "IOError: [Errno 2] No such file or directory: '/home/phablet/music_app/content/1.ogg'" ? [13:17] lool, nothing new, and minor [13:17] plars, yeah, and look at the log line above :-/ [13:17] Saviq: it's not coming up for some reason [13:17] plars, it's missing the 'autopilot' thing in there [13:17] Saviq: I see /run/unity8-setcap/unity8 in process list [13:18] Saviq: trying without setcap [13:18] lool, me tries [13:19] didrocks, Saviq: I will publish in a few moments [13:20] lool, right, it seems to hang after creating the surface... [13:20] Saviq: even without setcap [13:20] Saviq: I guess it's looking for some resources or something [13:20] * Saviq straces [13:21] lool, fook [13:21] is that polish? [13:21] ;) [13:21] lool, yeah, we're expecting /usr [13:21] I think I can read a bit of polish now [13:21] lool, otherwise we're looking for stuff in $BUILDDIR [13:22] sergiusens: possibly music_app/tests/__init__.py:_create_music_library() is looking in the wrong spot? [13:22] I guess I could fool argv0 [13:22] lool, [13:22] static bool installed = (QCoreApplication::applicationDirPath() == [13:22] QDir(("@CMAKE_INSTALL_PREFIX@/@CMAKE_INSTALL_BINDIR@")).canonicalPath()); [13:22] plars, well give me a second, I'll run locally [13:22] lool, you'd need to fool that [13:23] lool, we're basically comparing that we're run from where we were installed to [13:24] ogra_, ETA 13 mins on location service [13:24] good [13:24] jeez all that setcap stuff is such a hack... [13:24] Saviq: how about mount --bind /run/unity8-setcap/unity8 /usr/bin/unity8? [13:24] well, then seed change and meta uplaod [13:25] this thing is getting uglier minute by minute [13:25] heh, yeah [13:25] lool, should work I think [13:25] lool, indeed [13:25] I wonder if it works ro [13:26] mounting it ro after copying ? [13:26] *re-mounting [13:26] bind mounting over a file on a ro mount [13:26] plars, works fine for me http://paste.ubuntu.com/6240529/ [13:26] ah [13:26] sergiusens: http://popey.com/~alan/device-2013-10-15-142644.png looks good [13:27] popey, only 15 points? [13:27] haha [13:27] popey, you need practice :-) [13:27] lool, that seems to be working, yeah [13:27] lool, dropping letters is good [13:27] lool, getcap shows the cap on /usr/bin/unity8 [13:27] sergiusens: Cool [13:27] doing meta dance [13:27] yay for evil hacks [13:27] ogra_: you're done with meta for location? [13:28] lool, location had issues, rebuilds atm [13:28] ogra_: actually would you unseed the hud stuff again and upload meta? [13:28] just go ahead [13:28] i'll go afterwards [13:28] or leave it to me as you like [13:28] (but i'll wait for location sertvice to be ready) [13:29] Saviq: just testing on r/o now [13:29] sergiusens: hey, I noticed that the terminal app has a number of apparmor denials because it is now a click package [13:30] isn't it meant to be unconfined? [13:30] jdstrand, yes, not just the terminal, all unconfined apps seem to want to access the system_properties [13:30] right [13:30] jdstrand, and it's blocked [13:30] I need to adjust the unconfined template [13:30] jdstrand, I'm guessing it just need supressing? [13:30] ah so unconfined is not really unconfined [13:30] lool, never! [13:30] :-P [13:31] maybe we need a really-unconfined--hehe-no-just-kidding profile [13:31] lool: it is meant to be unconfined-- but unconfined click apps still get a profile-- it is just a wide open one based on the unconfiend template [13:31] Saviq: /dev/loop0 on / type ext2 (ro,relatime,errors=continue) [13:31] from an app perspective it is unconfined [13:31] lool: that is what the unconfined template is meant to do. it just needs a coupel of tweaks [13:31] /usr/bin/unity8 = cap_sys_resource+ep [13:31] Saviq: so worked [13:32] didrocks: Feel like yelling at unity people today? [13:32] didrocks: elmo's in the process of filing a bug that multimonitor regressed today. [13:33] plars, any comment? [13:33] plars, where's the code for that on utah again? [13:33] lool: to be clear-- every non-kernel process on the system is mediated by apparmor [13:33] sergiusens: sorry, give me a min.. I'm flipping back and forth between this, and iso install testing, and working with someone on a bug I'm seeing with that [13:34] lool: and 'unconfined' to apparmor is actually a profile [13:34] jdstrand: so that profiles doesn't block anything, but logs things it would have denied? [13:34] lool: we aren't using the 'unconfined' profile for these things-- we implemented an 'unconfined' template [13:35] lool: no, that is different. if a process is running with the unconfined apparmor label-- there are no restrictions or logging [13:35] I'm so confused now [13:35] I actually preferred 5 minutes ago [13:35] when I thought I understood what this meant [13:36] lool: the click unconfined template is supposed to do the same thing. but there is a bug that I need to fix so that click apps that use the unconfined template have no restrictions and logging [13:36] sergiusens: probably doanac or me will need to set it up locally and try, I'm also installing my phone now [13:37] lool: perhaps this will shed some light on it: http://paste.ubuntu.com/6240571/ [13:39] it was an active decission to not include '*mount' in the unconfined template, because we didn't think it was needed. of course, in the case of the terminal app, if you want to remount,rw it is [13:39] the attach_disconnected is needed because of the recent AF_UNIX change. at the time of that change, nothing on the image was using the unconfined template (at least, that I was aware of) [13:40] didrocks: https://bugs.launchpad.net/ubuntu/+source/unity/+bug/1240074 [13:40] Ubuntu bug 1240074 in unity (Ubuntu) "multi-monitor regression" [Undecided,Confirmed] [13:40] so that didn't come up until things like the file manager and terminal app were shipped as click on the image [13:40] lool: anyhoo-- I've got it under control. is this something I can upload after I finish testing? [13:41] jdstrand: is this an apparmor-easyprof-ubuntu upload? [13:41] lool: yes [13:42] infinity: ^ this is in supported seed, do you care? [13:42] I can answer that [13:42] it is in the supported seed, but it was whitelisted in some manner [13:42] jdstrand: all uploads are scrutinized when they aren't either unseeded or only in touch seed; all uploads now blocked in proposed too [13:43] supported, but not on an image, so I can let it through. [13:43] ok [13:43] it might be whitelisted [13:43] stgraber whitelisted it last week along with click, etc due to a conversation with slangasek [13:44] multi-monitor> erk. /me puts unity on hold [13:45] Packaging ACK needed! [13:45] sil2100: url? [13:45] https://jenkins.qa.ubuntu.com/view/cu2d/view/Saucy/view/Unity8/job/cu2d-unity8-saucy-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_unity8_7.83+13.10.20131015.4-0ubuntu1.diff <- lool [13:46] cjwatson: And by "on hold", you mean upgrading to see if you can confirm and provide more info? :P [13:46] sil2100: yes, this change is ok, but needs to go at the same time as ubuntu-touch-session upload [13:47] ogra_: did you do the ubuntu-touch-session upload already? [13:47] sil2100, yes [13:48] sergiusens: I've tested the filemanager and the terminal app. what else is unconfined? [13:48] jdstrand, music [13:48] ogra_: didn't add a breaks though [13:48] it had one before [13:49] ogra_: actually I dont see the upload in bzr? [13:49] infinity: ok, I got one issue and a reboot worked, let me get people on it [13:49] lool, ? [13:49] ogra_, lool: then I press publish! [13:49] ogra_: ubuntu-touch-session [13:49] infinity: I have 2 monitors here and they work now [13:49] ...or should I wait? [13:49] lool, oh, didnt puch sorry [13:49] bregma: around? [13:49] ogra_: I dont see lp:~phablet-team/session-manager-touch/trunk in bzr [13:50] sil2100: hold on just a sec [13:50] didrocks: Broke on several setups for IS here in Bluefin, so it's not just one person being goofy. [13:50] didrocks: elmo and Spads can be more helpful, I imagine. [13:50] infinity: yeah, I can believe you… [13:50] lool, done [13:50] sil2100, go [13:50] sil2100: you're pushing 7.83? [13:50] yes [13:50] sil2100: shoot [13:50] was all tested already [13:50] no worries [13:51] sergiusens: on what image is music-app a click? on 96 it is still a deb that doesn't ship a profile [13:51] double-checking actual versions [13:51] cause sometimes we rebump them [13:51] jdstrand: http://people.canonical.com/~ogra/touch-image-stats/20131015.changes [13:51] infinity: following up with trevhino on #ubuntu-unity [13:51] jdstrand: that would be latest image [13:52] Published [13:53] sil2100: unity8/notifications/sessoin? [13:53] didrocks: so...i hear the image cut today will be it ?...so mir/unity8/unity-mir will all need to branch like we discussed last week? [13:53] didrocks: Can you make sure they take this "fix/revert today, if possible" levels of seriously> [13:53] kgunn: it has everything in yeah [13:53] didrocks: just looking to keep people working and not stalling out as they move to t-focused work [13:54] infinity: they are already aware, etrying to get elmo or spads in the channel [13:54] lool: so, will I need an ask or after I test these three apps on the latest release may I upload? [13:54] didrocks: so when we branch....do we just tell you ? [13:54] or whom do we need to inform ? [13:54] lool: note-- only the unconfined template is being updated, so only unconfined click apps are affected by the change [13:54] kgunn: please not now, we have critical issues for release [13:56] didrocks: unity8 and unity-notifications [13:56] didrocks: ogra_ did session [13:56] infinity: if I get a chance - couple of other things I want to do before I tear down my session [13:57] cjwatson: Yeah. I reboot so rarely these days that I don't see bugs like this for weeks/months after they're introduced. :/ [13:58] Suspend working on Linux completely ruined my ability to be useful QA. [14:02] jdstrand: I've added a slot already [14:02] jdstrand: I think you can go with apparmor-easyprof-ubuntu [14:05] infinity: FYI, elmo can't reproduce it anymore (even upgrading unity) [14:05] so same case than I [14:05] there is clearly something [14:05] ... [14:05] but the commits don't seem to be related… [14:05] That's not actually comforting. [14:05] sergiusens: are you aware of: ERROR: Could not parse click manifest. Skipping 'com.ubuntu.developer.dinko-metalac.sudoku-app_sudoku-app_1.0.133.json' [14:06] yeah [14:06] I don't see any other upload [14:06] infinity: I just find weird that the screen is off, unity doesn't manipulate Xorg [14:07] it's just rendering on a root window (but at least, you would have a cursor) [14:07] jdstrand, I asked popey and mhal about that. To be honest I just want to unseed it now [14:07] which was not the case for the 3 of us [14:07] didrocks: Does multimonitor start 2 X servers, or 1 server with two heads? [14:08] infinity: 1 server [14:08] sergiusens: the problem is /var/lib/apparmor/clicks/com.ubuntu.developer.dinko-metalac.sudoku-app_sudoku-app_1.0.133.json is a symlink to /usr/share/click/preinstalled/com.ubuntu.developer.dinko-metalac.sudoku-app/1.0.133/sudoku-app.json, but that doesn't exist [14:08] didrocks: Hrm, then there goes my theory that it could break without a reboot if you had two copies of X running at different versions. [14:08] yeah, sorry ;) [14:09] jdstrand: I thought we fixed all that sort of thing ages ago [14:09] sergiusens: looks like it should point to /usr/share/click/preinstalled/com.ubuntu.developer.dinko-metalac.sudoku-app/1.0.133/apparmor/sudoku-app.json [14:09] cjwatson: I think it is just the hook manifest entry is wrong. [14:09] oh right [14:09] infinity: rereading the code from last upload, there is clearly no (obvious) link [14:09] yeah, that would be easy to fix in the app [14:09] one is in the launcher code [14:09] the other, only activated when you spread the windows [14:10] jdstrand, cjwatson the app developer messed up the manifest I uploaded for him [14:10] it must be sudoku-app.json now, but it should be "apparmor/sudoku-app.json" based on the layout of the unpacked package [14:10] didrocks: Yeah, I couldn't see anything obviously related either. [14:10] * jdstrand nods [14:10] sergiusens: certainly an easy fix :) [14:10] infinity: trevhino is still looking at it, I ask spads to keep his broken setup in case this can help [14:10] though I', surprised the review scripts didn't catch it [14:10] jdstrand, yes, but I need the reasoning for such a change from the dev [14:10] didrocks: I hate heisenbugs. :/ [14:11] infinity: maybe there's still time to switch to xmir? [14:11] jdstrand, I'm going to back out to the latest workng sudoku [14:11] :-) [14:11] sergiusens: well, right now the app is busted. I would think that would be motivating [14:11] infinity: so do I :/ as I just suspended/resumed for days, I thought my situation was screwed this morning (it's rare I plug my laptop on the external monitor after while I suspend) [14:12] but now that we are 3, I'm scared [14:12] * didrocks tries to unplug/plug again [14:12] sergiusens: anyhoo-- whatever you think is best. I jsut noticed the issue and am passing it along [14:12] jdstrand, thanks :-) [14:12] 5 times… nothing [14:12] I played with xrandr to start the second screen this morning… nothing [14:13] I was really thinking my Xorg was screwed… [14:13] jdstrand, I was sort of mad when I saw this last night :-) [14:13] lool: with its well-known excellent multi-monitor support [14:13] cjwatson: at least "unknown bug fixed" that way ;) [14:13] cjwatson: exactly :-) [14:13] or hidden rather :p [14:14] didrocks: Did it just take a reboot to fix you? [14:15] yeah [14:15] but I upgraded without rebooting first [14:15] and compiz is known to not like changing .so files under its feet (it can segfault) [14:15] plars: looks like I need to update utah to launch the music app properly? [14:16] doanac: oh? I didn't see that... what changed in utah? [14:16] sergiusens: yeah, I bet :\ [14:17] doanac: I'm not showing that there were any changes since september [14:17] plars: actually - it shouldn't be utah. [14:17] plars: keep in mind the app is now a click-app, so its launched differently. [14:17] doanac: yeah, saw that... I'm seeing some inconsistency with it when running locally though that isn't obvious why at the moment [14:18] well, this time was a device not found issue [14:18] adb choked I guess [14:18] plars: i see the issue sergiusens hit. [14:19] doanac: sergiusens was able to make it pass: see http://paste.ubuntu.com/6240529/ [14:19] phablet-tools is doing a "cd /home/phablet/autopilot" and running the test. we are just setting PYTHONPATH=/home/phablet/autopilot [14:19] infinity: at least, with Spads's xrandr output, we confirm that the screen is off from a Xorg perspective [14:20] plars: http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/touch/view/head:/utils/host/autopilot-run [14:20] that's what needs to change [14:21] didrocks: Yeah, but what turned it off? Weird. [14:22] yeah, if g-s-d or an upgrade Xorg was in the mix, that would make sense… [14:22] let me look at last upgrade… [14:22] xorg got upgraded for a security fix, but it was a tiny patch. [14:22] yeah, doesn't seem related… [14:23] didrocks: Laney's upower upload, maybe? [14:23] infinity: hum, not impossible [14:24] +If the device state is unknown, don't guess based on the laptop's [14:24] Nah, doesn't look related. [14:24] I have an unknown screen [14:24] (the internal one) [14:24] but fishy :p [14:24] That's just battery state. [14:24] oh right [14:24] and client->priv->lid_is_closed = prop_val; [14:24] If your monitor has a battery... [14:25] maybe it thinks the lid is closed? [14:25] Actually, THAT fix could be it. [14:25] And it might be that it was broken before and fixed now, but the transition from A to B breaks without a reboot. [14:25] if only we had a reliable way to reproduce [14:25] possibly, yeah [14:25] plars, doanac ok, so we can probably fix that on the test side too [14:26] didrocks: Seems like a tough thing to prove reliably. [14:26] sergiusens: yeah - i'd say both sides are wrong. the test shouldn't care, but CI shouldn't be causing this [14:26] didrocks: But I could see screen's powering on/off according to dock states being (in)correctly assumed. [14:26] s/screen's/screens/ [14:26] Internet is ruining me. [14:26] Well, those properties were only broken for a few days [14:27] infinity: it seems to me way more likely than Unity TBH [14:27] doanac, what is confusing is that the log message points to the right location but the copy doesn't :-/ I'll take a look at that [14:27] didrocks: Yeah, but blaming unity is more fu. [14:27] But stuff could have gotten in a confused state due to it [14:27] n [14:27] didrocks: there's a keyboard fix in asks [14:27] infinity: you are more (rightly) used to it ;) [14:27] Laney: It seems like a plausible explanation. And nothing we can do anything about, really. [14:27] didrocks: are we building an image with latest mir? [14:28] Yep [14:28] lool: we will once ogra_ get location-service in (soon) [14:28] lool: is it ready? [14:28] didrocks: I guess the only useful debugging info here would be to see if affected people are all fixed after a reboot. [14:28] just rebooting with the new location packages [14:28] didrocks: do I need to ping on -release to get unity8 and unity-notifications out of proposed? As I see "Not touching package due to block request by freeze (contact #ubuntu-release if update is needed) " [14:28] didrocks: And if we really want to be sure, downgrade upower, reboot, upgrade just upower, see if shit breaks, reboot, see if it fixes? [14:28] Laney: infinity: I just want to retry once scenario: starting without the external monitor plugged in, starting the session and plugging it in [14:29] sil2100: I'm doing it [14:29] ACK, thanks! [14:29] :) [14:29] I have no idea how we would fix it if it is that [14:29] Laney: maybe what infinity told worth a try? [14:30] Which is? [14:30] "downgrade upower, reboot, upgrade just upower, see if shit breaks [14:30] That's not a fix [14:30] reboot, see if it fixes?" [14:30] Just a way to see if it is that [14:30] Laney: If the above steps prove the issue, I'm not sure trying to fix it is worth it. As you say, the broken/confused state only existed for a short while, so it wouldn't affect precise/raring->saucy upgrades. [14:30] Laney: well, that's at least a way to confirm that's the issue [14:30] Right [14:30] Then maybe put a little release note in [14:30] No. [14:30] didrocks: the phone stack stuff also looks interesting (dialer-app / telepathy-ofono) in terms of polish for the phone part of the product [14:30] sergiusens, doanac: it's a simple change for us to make to that script, but I don't think asac will like it. We should still probably do it for the short run and also make sure the test is neutral as well [14:31] hmm, so for the location seed change .. [14:31] lool: let's keep on the safe side right now [14:31] Laney: Release notes for bugs during a devel cycle aren't appropriate. If it's what we suspect, it doesn't affect upgrades from previous releases, nor new installs. [14:31] we dont seed libubuntu-location-service-dev (it was a dependency before) [14:31] does anyone have an indea where it should go ? [14:31] sdk-libs-dev ? [14:31] Spads reboot fixed it, he's doing one more test [14:31] didrocks: we'll see tonight; if they aren't going in, we should reject them [14:31] lool: yep [14:32] didrocks: Anyhow, if you can verify that it is what we think, then I'm happy to close the bug and ignore it. [14:32] infinity: Fine, if that's all that you care about for those [14:32] Laney: A message to ubuntu-devel about "you might lose your monitor before a reboot, lolz" might be appropriate, if we prove this. [14:32] infinity: can't right now (as you can see, a lots going on), I'll try that later, or if Laney can beat me to it… [14:32] lool, didrocks ^^^ any opinion ? [14:33] ogra_: sdk-libs-dev makes sense to me [14:34] sil2100: ogra_: unity8 & co unblocked [14:34] cjwatson: do we need to retrigger the autopkgtests for system-image manually? seem they didn't run against ubuntu2 [14:34] ogra_: not sure, ask tvoss whether it should be part of sdk-libs-dev [14:34] ogra_: but we might not need to [14:34] ogra_: in any case it can wait [14:35] lool: they're running at the moment, it seems [14:35] lool, they are running [14:35] cjwatson: ah right, public jenkins == not seeing in progress runs [14:35] cjwatson: sorry [14:35] lool, you mean i should just leave it unseeded for now ? [14:35] jibel: but apparently breaking for some odd reason, dbus falling over? [14:35] lool, we cant change the sdk easily after release ... i'd like to have all dev bits in that are needed [14:35] ogra_: I dont think it's important either way; we dont have location plugins [14:36] ogra_: using location service from apps is via platform api [14:36] ah, more sensible messages from i386, but still a failure [14:36] lool, ah k [14:36] ok, barry will need to look at these [14:36] lool, right, then i'll leave it out [14:36] * ogra_ updates the seedand re-rolls meta [14:37] cjwatson, it fails on a timeout on self.downloader.get_files(downloads, pausable=True), does the test tries to download files from outside of the lab? [14:37] dunno [14:37] lool, ^ [14:37] ? [14:37] I've asked barry to look [14:37] didrocks: excellent, thanks [14:37] cjwatson, jibel: Have poked barry on #ubntu-touch [14:38] I already poked him on #ubuntu-devel [14:38] poor barry [14:38] full og holes [14:38] jibel: it should not try to download outside the lab; barry knows it should not be done, it's rather a download-manager timeout handling problem [14:38] *of === psivaa is now known as psivaa-afk [14:38] but I pinged him _first_ so we get to use #ubuntu-touch [14:38] nah [14:40] didrocks, did we have any final word on the chinese font front ? [14:40] before i'm done with meta :) [14:40] ogra_: yeah, just install the last package I listed [14:40] ogra_: there are some bugs in the settings [14:40] but it's not related [14:41] didrocks, well sounds like a seed change nontheless [14:41] yep [14:41] want me to change it? [14:41] if so, what package ... and should i add it now [14:41] or yeah. feel free to change the seeds and my meta update will pick it up [14:43] ogra_: committed [14:43] meta update running [14:43] how is our "MIR on steroids landing" going [14:44] asac, landed already [14:44] afaik [14:44] ogra_: which image? [14:45] no image [14:45] asac: we didn't start building the image yet [14:45] in archive [14:45] whats ETA for image spin for me to try? [14:45] ogra_: btw, your location service is ready :) [14:46] asac: ask lool, he wants to sneak another change [14:46] sneaky creepy changes [14:46] did I? [14:46] I think we should cut the image as soon as location service is ready [14:46] as planned [14:46] didrocks, already tested, please publish [14:46] well, he surely earned a bunch of CI coins :) [14:46] so you set the price :) [14:46] didrocks, i need it in the archive for this meta rebuild [14:46] 16:27:31 lool | didrocks: there's a keyboard fix in asks [14:46] lool: you didn't want those? ^ [14:46] didrocks: I just wanted to ping you about important things in asks :-) [14:46] oh ok, I got that you wanted to land that now [14:46] ogra_: ok, publishing [14:47] asac, the next meta needs to go in (which waits for location service) [14:47] didrocks: I thought it was worth assessing quickly; perhaps for just before or just after this image [14:47] ok, so it's the last one [14:47] asac, after that we''re ready for a build [14:47] lool: yeah, as you saw, high pinging time, I didn't look at the sheet yet [14:48] xnox, any news about udevd ? [14:48] ogra_: working on it. [14:48] okie [14:48] ogra_: finish writting it. [14:48] ogra_: but found a bug. will be testing it in vm. [14:48] yeah, i dont mean to be pushy, just asking [14:48] ogra_: if it doesn't crash horibly i'll ask you to test on a device =) [14:48] \o/ [14:49] we need to discuss the final release logistics too, such as parties [14:49] with beer ? [14:49] I'm afraid most people around where I live dont have the faintest clue of what it is I'm doing for work [14:49] haha [14:49] ogra_: http://paste.ubuntu.com/6240849/ [14:49] makes it hard to celebrate with them [14:49] same here [14:49] ogra_: while I'm changing the hint, do you have the version of meta that you are going to upload? [14:50] ogra_: NB! work in progress, don't try at home =) [14:50] "Yeah so it's a phone? like Android? Ok." [14:50] didrocks, that will be 1.094 [14:50] didrocks, but needs the location packages publiushed in the archive first, else germinate cant pick them up [14:50] ogra_: it's copying right now! [14:50] ogra_: I'm just updating the hint [14:51] right, still a publisher run is needed [14:51] (into archive, not proposed) [14:51] hum, it's in UNAPPROVED [14:52] not sure why location-service is seeded [14:52] cjwatson: infinity: ^ [14:52] *shrug* check germinate output [14:52] cjwatson: I'm not asking why (I'm looking at it), just if you can review it [14:53] looking [14:54] didrocks: this is bad, missing Replaces [14:54] wasn't what ogra_ added? [14:54] * didrocks rechecks the commits [14:54] oh wait yes [14:54] * cjwatson looks harder [14:55] didrocks: however, what actually *is* bad is the incorrect claim that this is Multi-Arch: same [14:55] didrocks, it was renamed, not added ... [14:55] foreign, maybe ... [14:56] yeah, for a -bin, arch: same is not really plausible [14:56] rejected, sorry, try again [14:57] ogra_: want me to fix it or will you? [14:57] cjwatson: why being so rude and agressive? even when I'm asking just for reviewing. You say I'm taking bad critism on cu2d directly, but you are taking any request from me at least badly… [14:57] didrocks, oh, if you want [14:58] anyway, let's fix it and move on [14:58] didrocks, i was just starting to hunt tvoss down [14:58] ogra_: let's not wait on that [14:58] * ogra_ has a meeting in 2, else i would have done it [14:58] ogra_: I'll ping you for a review, I'm checking the .debs first [14:58] didrocks, ok [15:02] didrocks: mm? I think you misunderstood me, I certainly didn't intend to be rude or aggressive [15:02] read "try again" with a smile on my face [15:03] sorry if it came across that way [15:03] I was just in a rush to go and get belated lunch [15:03] didnt come across that way to me [15:03] ok, no worry, sorry for having misunderstood :) (or maybe frenchie-contextual ;)) [15:03] but now I have turkey soup and all is well === psivaa-afk is now known as psivaa [15:06] :) [15:41] plars, doanac can you think of a better way of doing this? https://code.launchpad.net/~sergiusens/music-app/test_asset_location/+merge/191228 [15:41] doanac, plars if it works for you please approve [15:42] storing the assets in te python package is the wrong way to do it imo though [15:51] sergiusens: i'm +1, i use that same trick all the time [15:52] doanac, I feel dirty though :-) [15:52] doanac, can you +1 in the MR so I can forward to devs? or balloons -> https://code.launchpad.net/~sergiusens/music-app/test_asset_location/+merge/191228 [15:53] sergiusens: did you feel dirty before, or just once you heard that your approach was something *I* like to do? :) [15:53] sergiusens, doanac: could we just use some of the ogg files from ubuntu-sounds? [15:53] doanac, LOL.... before, don't worry [15:53] sergiusens: its +1'd [15:53] sergiusens, ohh, nicey [15:53] plars, I'm not the test author or plan to maintain these in the long run, that's more a question for balloons or the music-app-devs [15:54] is ubunut-sounds a package or website? [15:54] balloons, that's somthing we would want to propagate over to other apps as well [15:54] doanac: package [15:54] ogg files from sounds? I actually migrated to no using a package [15:54] sergiusens, yea all the harcode paths in __init__ need cleaned up [15:54] balloons: ogg files from a package that is already installed and puts them on the system [15:54] heck all of __init__ does.. it's been a process [15:55] balloons, yup :-) [15:55] plars, they used example-content and I switched it to using sounds we ship. [15:55] I much prefer it that way [15:56] it keeps the app safer from breakage if something changes in ubuntu-sounds [15:57] yes, we need control over things, and in this case we needed ogg files specifically.. pulling dependencies is silly as I can craft my examples [16:01] ogra_: coming? [16:01] robru: ? [16:01] didrocks, yep [16:02] oops, yeah [16:02] sergiusens: i'm seeing something odd trying to run music_app at home. I *think* its because python can't import "mock" [16:02] does that make sense, or do i have something off? [16:06] sergiusens: I think I've confirmed this is a real issue now. music_app uses python-mock, but click-setup doesn't grab it. [16:07] and its not in the base image [16:14] can somebody checking what's wrong here? some "ssh: connect to host bazaar.launchpad.net port 22: Connection refused" and other errors https://jenkins.qa.ubuntu.com/job/generic-mediumtests-builder-saucy-armhf/2947/console [16:16] cyphermox: https://code.launchpad.net/~attente/ubuntu-system-settings/lp1234407 [16:16] kalikiana: someone needs to put that cable back where it belongs! [16:17] yeah that's what is looks like. the cleaning lady/ dude stepped on it [16:17] I'm still waiting for the day where that is treated as a clearly internal error instead of a failure [16:18] I'm still waiting for the day where we can merge anything new into UITK trunk [16:19] kalikiana: I check cyclops-node09 after seeing that failure but couldn't reproduce :-/ Probably a lp transient failure but too long for the retries that are already in place [16:19] doanac: it isn't python3? [16:20] vila: so should I wait for now if it posts new results? [16:21] cjwatson: i'm guessing no. I did an "apt-get install python-mock" and things work now [16:22] kalikiana: I just fired a rebuild [16:22] okay, thanks a bunch [16:26] doanac, I have an MR that no one dares to approve :-? [16:26] sergiusens: which one? [16:27] doanac, https://code.launchpad.net/~sergiusens/phablet-tools/click_testing_python_modules/+merge/189744 [16:27] doanac, I've been using it all along [16:28] cjwatson, autopilot 1.3 isn't python3 [16:28] sucks to be you :-) [16:28] cjwatson, autopilot 1.4 I think is, but that's T; and would also mean all the tests would need to be ported [16:28] cjwatson, actually, it sucks to be balloons :-) [16:29] wow I didn't want to look at that change [16:29] sergiusens: that is quite the hack. [16:29] ogra_: didrocks: maybe I missed it, but is there an ETA on when 98 will be kicked off? [16:29] doanac, yeah [16:29] plars: when location-service is in with meta update [16:29] doanac, it says so on the MR [16:29] lool: ack [16:30] plars, waiting for location-service to make it to the archive ... then i can rebuild -meta, upload, have that go in and then i'll build [16:30] plars, 1-2h i'd say [16:30] ogra_: cool... sounds like we're getting a respin in a bit on the isos, so I may have some time to shift focus. Still trying to fill some gaps on the iso testing though [16:30] sergiusens: any suggestions on how CI should manage this for now? [16:31] should we just install python-mock manually after we provision? [16:31] doanac, well plars was preinstalling them [16:31] plars, ah, k, i'll make sure you dont run out of work then :) [16:31] doanac: I thought sergiusens already took care of the python-mock dep in phablet-click-test-setup [16:32] plars, well no one approves the MR ;-) [16:32] plars, sergiusens: looks like a mis-communication [16:32] plars: i think we should just manually install it for now [16:32] plars, a gave it to you for review too :-) [16:32] doanac, sergiusens: but yes, if there are other deps we need, we can just add them in the prepare-autopilot-test step [16:32] plars, and you said but doanac said just now [16:32] sergiusens: sorry, let me take a look [16:32] plars, doanac the only bad thing about that route is that you don't get read only mode === gatox_bbl is now known as gatox [16:33] Ok; going afk, hopefully for the evening, but I might pop back to check if anything urgent is needed [16:33] doanac, tbh, as click, music app does not need python mock [16:33] doanac, we could conditionally import [16:33] sergiusens: should we try and patch music_app then? [16:33] ogra_: you're kicking the image I guess? [16:34] lool, yep, enjoy the evening [16:34] doanac, one sec for that [16:34] Ok [16:34] sergiusens: are you talking about https://code.launchpad.net/~sergiusens/phablet-tools/click_testing_python_modules/+merge/189744 [16:34] plars, yes [16:34] sergiusens: I don't seem to be on the list [16:34] plars, oh, I asked you on irc [16:35] sergiusens: I guess I didn't see it, let me take a look now [16:35] * sergiusens needs to afk for a bit [16:39] sergiusens, doanac: I don't see how that hack helps keep from being in rw mode... it's still going to need to be writable for extracting all that to the appropriate python paths, so is there really a big difference between that, and just installing the packages? [16:39] didrocks: can we try getting HUD into one of next landings? [16:39] plars: i think that hack puts the python code under /home/phablet/autopilot [16:40] plars: maybe we should save this and discuss at the end of our next meeting? [16:40] sort of a related topic [16:42] doanac: ah, I think I missed where it moves it to target_dir [16:50] jdstrand: hm, got the following when updating: [16:50] ERROR: Could not parse click manifest. Skipping 'com.ubuntu.developer.dinko-metalac.sudoku-app_sudoku-app_1.0.133.json' [16:50] after updating apparmor-easyprof-ubuntu [16:50] known issue? [16:51] rsalveti, iirc the app naming is broken [16:52] rsalveti: it is. sergiusens is working on it. basically the dev changed something in the click manifest that broke it [16:53] right, cool [16:54] thostr_: the hud seems like a minimal fix right? Can't we want post V1? [16:54] like Friday [16:55] can you guys unblock stuff? [16:59] if so: ubuntu-touch-sounds would be nice [16:59] it just adds a new sound [17:02] ev, doanac: are we having this call? or postponing until after the release? [17:02] jumping on now [17:02] plars: sorry, got distracted reading the meeting notes :) [17:08] lool: do we have a doc already for the -touch release notes? [17:08] it might be good to start putting the known issues in there [17:08] Laney: done [17:08] wesome [17:08] ta [17:09] Laney: cyphermox and robru are around if you need a new system-settings kick [17:09] yw [17:09] hello [17:09] would be nice [17:09] ok [17:10] === meta 1.094 uploaded === [17:11] Laney, ok, building now [17:12] ty [17:12] * ogra_ is sad that we dont have a single acrually alarming ringtone ... none of the once we ship qualify as tones i would use (i want to be able to hear my phone in the next room if it rings, not spheric cuddle music) [17:12] off for the evening now [17:12] see you [17:13] spheric cuddle music hahahaha [17:13] :) [17:13] i mean, the sounds are nice and all ... but not as ringtones === bfiller is now known as bfiller_afk === alan_g is now known as alan_g|EOD [17:21] plars, try my branch in read only mode [17:21] jdstrand, I reverted [17:21] jdstrand, next build will have the fix [17:21] plars, it does solve the problem of not going to rw [17:22] * jdstrand nods [17:22] sergiusens: we are still rw for now, even if we don't really change much or anything - I'm not sure we want to risk changing that at the last minute [17:25] sergiusens: doanac and I were just taking about it, and it seems the least invasive change for now would be to just require python-mock (on our side) before running those tests that need it. This is already done for some earlier click packages that needed other dependencies iirc [17:26] sergiusens: we should definitely still revisit going fully ro once everything goes to click, and we don't risk screwing up the results [17:29] plars, well unity8 will never go to click, so we need to revise on that too [17:29] plars, I think that branch I have does allow for testing unity8 in ro though [17:42] plars: FYI - the branch was merged: https://code.launchpad.net/~sergiusens/phablet-tools/click_testing_python_modules/+merge/189744 [17:44] doanac, so let's get that in a landing asks [17:49] fginther, what is the status of upstream merger? is it enabled at least for lp:cordova-ubuntu-tests? [17:49] robru, yes, it should be [17:49] fginther, ok, thanks [17:56] == image 98 building == === bfiller_afk is now known as bfiller [18:10] lool: added a landing ask, row 214, it's in jfunk's bad bugs list for today [18:22] https://launchpadlibrarian.net/153817043/buildlog_ubuntu-raring-i386.qtorganizer5-eds_0.1.1%2B13.10.20131011-0~16~ubuntu13.04.1_FAILEDTOBUILD.txt.gz this build failure is caused by the build environment being *raring* and not having new enough packages to satisfy dependencies. what would cause this? this package has built successfully before, implying that it was previously built on saucy [18:22] anybody have any ideas? ^^ fginther ogra_ lool [18:23] did someone do s/saucy/raring/ in debian/changelog ?? [18:24] robru, what PPA did the build? [18:24] fginther, ubuntu-unity/daily-build [18:25] fginther, this was from a daily_release jenkins job i kicked off earlier today [18:30] robru, I don't get it, was raring somehow introduced into the daily_release job? [18:30] fginther, i have no idea. [18:33] == image 98 done == [18:37] ev, plars, sergiusens: https://code.launchpad.net/~doanac/phablet-tools/unlock_screen/+merge/191268 [18:37] that should serve as a starting point we can discuss. [18:37] * doanac heads to lunch [18:41] rsalveti: dont think so [18:43] robru: where is this failure *from*? [18:43] robru: I dont see any raring build records in ubuntu-unity/daily-build PPA [18:44] lool, https://launchpad.net/~ubuntu-touch-coreapps-drivers/+archive/daily/+build/5105118 [18:44] brb [18:44] rsalveti: feel free to dump some notes in a google doc and send it to didrocks + asac + me? [18:45] doanac, added comments [18:47] I'm going afk again, but the qtorganizer robru pointed at doesn't seem to be from cu2d: [18:47] https://launchpad.net/~ubuntu-touch-coreapps-drivers/+archive/daily/+recipebuild/559985 [18:47] Recipe qtorganizer5-eds-daily for Ubuntu Core Apps Drivers [18:47] don't think we use recipes [18:47] ogra_: cool, /me upgrades [18:48] ogra_: ah but not imported yet [18:48] will try it in a few then [18:48] lool, oh, weird. i got that failure in an email shortly after kicking off a build in jenkins, got confused i guess [18:49] robru: http://10.97.0.1:8080/view/cu2d/view/Saucy/view/All/job/cu2d-misc-saucy/ shows it prepared and built fine, and is ready to publish [18:50] * lool & [18:50] lool, ok, still testing it before publishing. thx [19:00] new mir is really good! [19:00] got some keyboard issues at first with new image, now seems ok [19:01] lool: 97 or 98? [19:02] popey: 98 or whatever is latest [19:02] popey: just came up minutes ago [19:02] * popey checks for updates [19:02] yay [19:02] yeah 98 [19:02] * popey installs [19:04] blimey, thats way faster [19:08] lool: try installing something big like Sleepy Time... Power saving kicks in, then if you go back to install it again, it takes ages to restart.. [19:08] popey: eeeerk, insanely faster [19:17] vila: lool seeing bug 1239708 at all? [19:17] bug 1239708 in unity8 (Ubuntu) "Welcome screen clock missing" [Undecided,Confirmed] https://launchpad.net/bugs/1239708 [19:21] popey: comment added, seen on 97 [19:46] lool: https://code.launchpad.net/~renatofilho/qtorganizer5-eds/fix-match-all/+merge/191080 can that be added to landing asks please? [19:49] fginther: I see you aborted the last CI run here https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/ci-test/+merge/190906 [19:49] fginther: do you have any idea why the other runs (not aborted) fail? We cannot find the problem. [19:51] t1mp, I don't know the full story about why things are failing so often. (and sorry, I meant to restart that after image 98 was flashed)... [19:51] t1mp, there appears to have been two problems... [19:52] t1mp, turning on the screen has been unreliable for me. I think this is timing related. If the screen does not come on before the tests start, they all fail [19:53] t1mp, the other problem I was seeing is that unity8 was crashing or restarting during the tests, this was also causing failures. [19:55] t1mp, I have been working on changes to the test runner to work around the timing issues, but I've had bad luck. Testing with image 98 has been better, if I continue to see good results, I'll update the default testing [19:55] fginther: did something of that get fixed in image 98? [19:56] fginther: we cannot have any changes in ui-toolkit for some time now because the tests always fail. see this bug https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1239646 [19:56] Ubuntu bug 1239646 in Ubuntu UI Toolkit "CI fails most tests on UITK trunk" [Critical,In progress] [19:56] rsalveti: as lool isn't around, could you please add https://code.launchpad.net/~renatofilho/qtorganizer5-eds/fix-match-all/+merge/191080 to asks? [19:57] t1mp, unity8 has been running better the last few images, so I don't think that's a big issue anymore [19:57] welcome [19:57] i just asked ☻ [19:57] \o/ [19:57] t1mp, have you had good test results when running the uitk tests by hand? [19:58] channel 106 [19:58] will follow up with didrocks in the morning [19:58] thanks popey [19:59] fginther: I think they were working fine, but let me run it on a device myself now. I just installed the latest image and ubuntu-ui-toolkit-autopilot [19:59] fginther: what was the command to run the autopilot tests that are executed by CI? [20:00] t1mp, "autopilot run -v -o /tmp/${autopilot_results} -f xml $suite 2>&1 | tee /tmp/autopilot.log" [20:00] fginther: as user phablet? [20:00] t1mp, suite would be ubuntuuitoolkit [20:00] t1mp, yes [20:00] t1mp, and make sure the display is on first [20:00] fginther: how is that ensured on jenkins? [20:01] t1mp, heh, it's a best effort: "powerd-cli active &" and then "powerd-cli display on bright &" as root [20:03] fginther: running. so far it seems good. [20:05] fginther: the screen switches off during the tests. is that bad? [20:05] t1mp, yes [20:05] t1mp, I think the tests will all fail after that [20:07] t1mp, I'll be interested in the results. My last test run with uitk and the runner modifications are here: http://10.97.0.26:8080/job/generic-mediumtests-runner-mako-fjg/52/console [20:08] t1mp, looks like two tests failed (ubuntuuitoolkit.tests.gallery.test_gallery.GenericTests.test_textfield_password and ubuntuuitoolkit.tests.test_emulators.TabsTestCase.test_switch_to_next_tab_from_last) [20:09] t1mp, but then the device got hung up during package revert and I had to kill it [20:09] what the hell happened to calendar-app-autopilot? it's not in the PPA, not in distro, either. but Touch/Testing page references it and I have memories of running it last week. [20:10] t1mp, gotta go out for about 2 hours, I'll give an update later as I keep working on this. [20:10] fginther: ok also fails. see my last comment at https://code.launchpad.net/~ubuntu-sdk-team/ubuntu-ui-toolkit/ci-test/+merge/190906 [20:10] I didn't analyze what goes wrong yet [20:33] t1mp: 'ProcessSearchError: Process exited with exit code: -11' IIUC, *something* is SIGSEVing, the log gives no clue... (http://pad.lv/1239427 is about giving a bit more context) but if you can reproduce running a single test, you may be able to write a simpler test to reproduce ? [20:33] Launchpad bug 1239427 in Autopilot "ProcessSearchError gives poor error message" [High,Triaged] [20:34] t1mp: this sounds like an integration test you can add to your own suite to better detect regressions in the future too [20:35] vila: actually the errors occur in setup, before the actual tests are run, so even an empty test would fail [20:40] t1mp: whatever code is in setUp can be put in a different test that is not using that setUp [20:41] and would get closer to the actual failure [20:42] vila: I'm actually fixing that right now [20:42] thomi: \o/ [20:42] vila: the application under test is segfaulting [20:42] that's the *something* [20:42] thomi: in 1.3 right ? [20:42] veebers: yeah, but it'll likely need an SRU [20:42] thomi: ? [20:43] veebers: ? [20:43] veebers: sorry [20:43] vila: ^^ [20:43] veebers: he meant vila I think ;) [20:43] thomi: no worries ^_^ [20:43] I can communicate with, at most 26 people... one for each letter of the alphabet [20:43] an SRU so we can make tests pass, said tests needed to pass to land.. err wait [20:43] you two are just 'v' to me [20:44] thomi: you and I can communicate, but other people cannot communicate with both of us at the same time, right? ;) [20:44] thomi: some irc clients handle that based on the latest speaker ;) [20:44] vila: well, it won't make any more tests pass, it'll just make the exception more informative [20:44] thomi: https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1239646/comments/8 [20:44] Ubuntu bug 1239646 in Ubuntu UI Toolkit "CI fails most tests on UITK trunk" [Critical,In progress] [20:45] thomi: right but t1mp could then dig faster in the right direction, either the app is his or one it depends on or ... something ;) [20:45] thomi: I don't know if that helps. doesn't say much to me [20:45] vila: t1mp: qmlscene is segfaulting [20:45] *that's* the problem [20:45] maybe run it under gdb or something and see what the problem is? [20:45] vila: it is not an app, but ubuntu-ui-toolkit. since I came back from my holidays all jenkins tests fail so we cannot merge anything [20:46] t1mp: your tests used to pass using/providing some resources either your code or your tests needs to be updated to take changes in one of these resources into account [20:47] vila: we're trying to figure out what changed [20:47] t1mp: switch to mir ? [20:47] vila: yeah that's a big change :) to be more specific, we are trying to figure out how that change affects our tests [20:48] t1mp: they crash :) [20:48] vila: good. you know as much as we do now :) [20:49] t1mp: joke aside, the simplest test that leads to the scratch will help everybody, since yours are crashing in the setUp, you're already close [20:49] *crash [20:49] where the hell is that freudian slip coming from ? [20:50] wow. some times I unlock the phone, I see a big white square on the welcome screen and after that the device works really slow [20:50] +when [20:50] vila: perhaps you are itchy ;) [20:51] t1mp: I'm EOD, passing around and don't have a reproducing setup :-/ [20:52] vila: I'm way past EOD ;) [20:53] hmm.. $ qmlscene ubuntu-ui-toolkit-gallery.qml segfaults on my device [20:53] how come the apps that use the UITK work :s [20:53] t1mp: here we go ! That should allow you to re-assign bug #1239646 no ? [20:53] bug 1239646 in Ubuntu UI Toolkit "CI fails most tests on UITK trunk" [Critical,In progress] https://launchpad.net/bugs/1239646 [20:54] vila: re-assign? to who? [20:54] t1mp: well, re-assign or at least discuss with whoever is responsible of qmlscene and can help diagnose the crash ? [20:55] lool: we got a big regression with 98, phone is not up after boot [20:55] lool: we're fixing that and would need to trigger a new build after that [20:55] so we can have more people testing the faster mir image [20:55] vila: but yeah it brings us closer at least :) [20:56] vila: I have no idea who is responsible for qmlscene that I could assign the bug to [20:57] t1mp: yeah, that's the idea, reduce the scope, it may be that your tests triggers a side-effect in qmlscene that is not encountered elsewhere [20:57] t1mp: dunno who is responsible for qmlscene either [20:57] vila: yes some apps are running on the device apparently. so qmlscene does not *always* crash [20:58] fginther, lool, ogra_ ^ [20:58] grr [20:59] fginther, lool, ogra_ : who is responsible for diagnosing a crash in qmlscene ? t1mp can crash it with $ qmlscene ubuntu-ui-toolkit-gallery.qml [20:59] wow, qmlscene without parameters also segfaults for me [20:59] * vila really off [20:59] vila: have a nice evenign [21:00] I can help a bit if I can do it very soon. 11pm here, getting sleepy [21:00] t1mp: thanks, happy hunting ! [21:00] vila: thank you :) [21:05] @ci [21:06] nik90: how can we help? [21:08] asac, lool: I'm rerunning unity8 on mako now, it only had 1 failure this run bug maguro (!) had no failures except systemsettle, which isn't too surprising on this build [21:08] so... hopefully [21:08] seen a few other cases of new things going to 0 failures [21:09] mediaplayer seems to have gained a fail, on both maguro and mako [21:09] plars: hm, let me check that [21:09] plars: phone is broken with 98 [21:09] rsalveti: :( [21:09] rsalveti: what broke it? I haven't installed locally yet [21:10] plars: ubuntu-touch-session changes [21:10] https://code.launchpad.net/~rsalveti/session-manager-touch/ofono-setup-started-dbus/+merge/191295 [21:10] we're testing that now [21:10] doanac: I think we need to just add the dep for python-mock on music-app for now, at least until the phablet-tools changes propagate. You ok with that? [21:10] josepht: sry, I thought @ci would list the commands or something.. [21:11] plars: yeah. makes sense [21:11] nik90: no worries [21:13] plars, we'll do an emergency respin after that landed so people have working phones over night [21:15] ogra_: ack [21:15] doanac: https://code.launchpad.net/~pwlars/ubuntu-test-cases/add-python-mock/+merge/191296 [21:16] plars: +1'd [21:16] doanac: ok, I'm going to try giving back music-app then [21:17] on mako at least... see how it goes [21:21] bah [21:21] maliit crashes everywhere ... [21:25] ogra_: ? [21:25] ? [21:25] cyphermox, yes ? [21:25] maliit> ubuntu-keyboard (0.99.trunk.phablet2+13.10.20131015.2-0ubuntu1) saucy; urgency=low ? [21:27] cyphermox, its an upstart issue ... see landing 242 [21:27] maliit currently starts to early [21:27] ah [21:28] well I'm having enough problems with ubuntu-keyboard that I'm not confident with releasing it [21:28] for instance, when you swipe it away a few times (or perhaps just one) it doesn't want to come back again [21:29] oh, fun [21:29] yeah. [21:29] trading 4 quarters for a dollar :) [21:34] haha [23:04] triggering new build, with the phone fix