=== fginther|away is now known as fginther [01:18] Saviq, I got silo 2 for your split-greeter landing. [01:18] Saviq, even hit build for you too: https://ci-train.ubuntu.com/job/landing-002-1-build/29/console [01:37] robru: mind getting me a silo for 58? [01:37] robru: need it via a silo as I want to build for all archs [02:04] === trainguard: IMAGE 281 building (started: 20140408 02:05) === [03:24] === trainguard: IMAGE 281 DONE (finished: 20140408 03:25) === [03:24] === changelog: http://people.canonical.com/~ogra/touch-image-stats/281.changes === [04:15] rsalveti: assigning, might be late.. [04:15] Mirv: still fine [04:16] rsalveti: landing-004 after it refreshes [04:16] Mirv: great, thanks! [05:18] rsalveti, sorry I was out at a movie. looks like Mirv got ya... next time! [05:29] robru: I don't think there's anything wrong in spending an evening at a movie! :D [05:29] ;-) [06:21] morning didrocks [06:22] hey Mirv [06:26] Mirv: unity-webapps-qml seems to be stuck in proposed (robru seems to have used the override) for hours, mind looking why? [06:30] * Mirv looks [06:31] why overriding :( [06:34] hmm, why on earth does unity-webapps-qml build against qtwebkit and runtime depend on both qtwebkit and oxide.. [06:36] I think I may need osomon online to check some facts before contacting release team [06:37] Mirv: right "why" :/ [06:37] Mirv: fowarding you the context [06:40] thanks for the email [06:40] imgbot, status 281 manta [06:41] Image 281 test results on manta - Total: 655, Pass: 622, Crashes: 3, Rate: 93.5% [06:43] didrocks, Mirv: are we in TRAINCON0 already? ;) [06:44] sil2100: I guess we've sort of TRAINCON status override or such... [06:45] sil2100: FYI I'm looking at the unity-webapps-qml migration problems, but now waiting for osomon to arrive online to ask some questions about the qtwebkit/oxide divide [06:46] u-w-q has very "funny" b-d/deps [06:48] sil2100: not yet, it's under discussion :p [06:48] Mirv: ^ [06:48] so right now, we continue pushing changes [06:48] Mirv: "funny"? I like fun! :p [06:55] Mirv: ACK [07:02] good morning [07:02] hey dbarth! [07:03] didrocks: hi [07:03] all good with oxide? [07:03] it got approved late last night [07:03] dbarth: webapps-qml isn't out of -proposed though [07:03] (see the emails) [07:03] dbarth: I'm unsure about the impact and Mirv needs osomon to be around to understand some details [07:03] however i need help for a reconfig on silo 8 cause i had forgotten a branch [07:03] uh [07:03] ok checking [07:04] dbarth: just read emails, if you can assess the impact there, it will be nice :) [07:06] didrocks: ok done [07:06] didrocks: the impact is simple: desktop webapps don't start at least, because they want the newer 1.0 qml layer which is still 0.1 until the new webapps-qml is promoted [07:07] dbarth: ok, so desktop only? [07:07] nothing with touch? [07:07] i suppose touch would be impacted, but in a lesser way [07:08] so anyway, this unity-webapps-qml promotion needs to be solved [07:08] dbarth: right, I'm trying to understand if we need to respawn then a Touch image, if you can grab more info on the Touch one [07:09] dbarth: for solving it, Mirv is on that once osomon is around [07:09] dbarth: the problem comes from the fact that webapps-qml builds against qtwebkit and depends from the binary package to both qtwebkit and oxide [07:09] Mirv: we have some archs it's not available for? [07:09] dbarth: if you happen to know the reason why it doesn't build against oxide instead, please tell (I thought oxide was about replacing qtwebkit) [07:09] do you have a build log? [07:09] didrocks: yes. it builds for all archs since it builds against qtwebkit, while it has a binary dependency against oide [07:10] the latest unity-webapps-qml builds against oxide 1.0 [07:10] dbarth: webapps-qml builds against everything, since it uses qtwebkit instead of oxide for building [07:10] it may have been build while oxide 1.0 was not yet in the archive [07:10] ie, in unapproved [07:10] dbarth: no, https://launchpadlibrarian.net/172041103/unity-webapps-qml_0.1%2B14.04.20140407-0ubuntu1.diff.gz - the one in proposed builds against qtwebkit [07:10] besides, u-webapps-qml has code to fallback on qtwebkit if oxide is not there [07:11] Mirv: ok checking [07:11] Mirv: oxide is in main now, so it needs to build-deps on both to at least restrict the archs I think [07:11] (in that diff it's possible to search for "control") [07:11] right, so that looks like the correct release [07:11] didrocks: yes, if it would build-dep on oxide then it would solve the problem of building but not being able to install [07:11] Mirv: not fully, as then: [07:11] unity-webapps-qml | 0.1+14.04.20140401.1-0ubuntu1 | trusty | source, amd64, arm64, armhf, i386, powerpc, ppc64el [07:11] so we need to check if it can be removed on other archs [07:12] didrocks: yes, then it would still need removal for the other archs, but as it is it builds but is uninstallable which is worse [07:12] Mirv: if I'm reading reverse-depends right, it should be fine [07:12] Mirv: nothing on the other archs are depending on it [07:13] so yeah, the build-dep trick should be enough [07:13] ok, so the problem is limited to webapps-qml at least [07:13] right ;) [07:13] dbarth: Mirv: can you handle a MP with that fix and release it? [07:13] (oxide being in main, there is no issue) [07:14] um, there is no oxide "dev" package in the first place [07:14] to build against [07:15] Mirv, didrocks: cjwatson had advised not to add arch-specific tricks though [07:15] dbarth: hence the build-dep [07:15] dbarth: it does avoid setting arch-specific metadata [07:15] hmm ok, fine then [07:15] Mirv: I would add the build-dep against liboxideqt-qmlplugin personnally [07:15] so, it's really just a dependency issue that had it blocked? [07:15] Mirv: with a comment on why [07:15] cause i think we had it uploaded before [07:16] the only thing that changed was the oxide 1.0 transition really [07:16] dbarth: it's the fact that "it built on all archs, but you won't be able to install it on all archs" [07:16] because of this new dep [07:16] ok [07:16] dbarth: then, now, because previous version was available on all archs, I'll need to remove some binary packages as well [07:16] ok [07:16] dbarth: however, please, can you really assess before 10am the consequence for the current touch image? [07:17] dbarth: that would really speed us up [07:17] i can review an MP, just not going to be fast at that [07:17] and i can check the impact on phone in the meantime [07:17] dbarth: I'm handling that with Mirv, just assess the impact :) [07:17] ie, image #281 [07:17] ok [07:17] yep [07:17] thanks! [07:17] cool [07:17] Mirv: yeah, so the -qmlplugin seems the best to me, wdyt? [07:18] didrocks: or the lib itself https://code.launchpad.net/~timo-jyrinki/unity-webapps-qml/build_dep_on_oxide/+merge/214673 [07:18] Mirv: as you feel more confortable with :) [07:19] Mirv: please, add a comment :) [07:19] or it will be dropped with a cleanswap without us remembering about it [07:19] didrocks: I'm not totally comfortable with either, but either would work [07:19] didrocks: is the commit message not enough? [07:19] "to limit archs to be built to the set for which webapps-qml is installable for." [07:19] Mirv: I would add it directly to debian/control [07:20] right, doing [07:20] Mirv: the same one than your commit message :) [07:20] pushed [07:21] Mirv: approved [07:21] Mirv: you file the landing ticket and handle it? [07:21] Mirv: just keep me posted one published so that I remove the binary packages on other archs [07:21] once* [07:21] yep, landing [07:22] thanks ;) [07:23] dbarth: this separate landing would mean you'd need to rebuild landing-008 after this has landed, is that ok? [07:30] Mirv: I would say it's ok, let's leave a comment on the other silo with unity-webapps-qml and assign this one [07:30] To get it landed quick [07:30] Mirv: sil2100: May I have a Silo for the line 60? [07:31] bzoltan: sure! We're allowed to land stuff as goes so let me assign a silo for you ;) [07:32] Mirv: thanks Timo \o/ [07:34] sil2100: yeah, I figured out it's probably ok and fired a build [07:34] sil2100: Thank you! [07:35] yw o/ [07:44] Mirv: it's ok, i was prepared to rebuild 008 anyway [07:47] fginther: can you shed any light on https://bugs.launchpad.net/ubuntu-ci-services-itself/+bug/1252933 ? [07:47] Ubuntu bug 1252933 in Ubuntu CI Services "Mir performance testing as a merge requirement" [Undecided,New] === psivaa changed the topic of #ubuntu-ci-eng to: Ubuntu CI Engineering Team | Vanguard: psivaa | CI Train support - US: robru, cyphermox, rsalveti - EU: sil2100, Mirv, didrocks | CITrain support no answer: use mup bot after 30 minutes, but choose right timezone | Known issues: - [08:04] didrocks, Mirv: i tested image #281 and webapps and html5 apps are not affected luckily [08:04] dbarth: thanks for confirming :) [08:05] Mirv: do you have a branch i need to +1 to re-upload? [08:05] dbarth: https://code.launchpad.net/~timo-jyrinki/unity-webapps-qml/build_dep_on_oxide/+merge/214673 [08:07] dbarth: there's a branch [08:13] ok [08:13] i can take this one in silo 008 since i need a valid unity-webapps-qml there as well [08:14] Mirv ^^ [08:14] dbarth: no, we want to solve the situation ASAP, let's do 2 landings [08:15] ok, so i'll remove the other branch from the silo and will build once you tell me the other one is landed [08:17] it's building [08:17] Mirv: it will deadlock on building btw [08:17] Mirv: as it was available on more archs [08:17] Mirv: so, actually, it's ready is seems :) [08:20] imgbot, status 281 [08:20] Image 281 test results on mako - Total: 676, Pass: 661, Crashes: 5, Rate: 98.4% [08:20] hmm, quite a few failures [08:23] didrocks: yes, I'll smoke-test and publish afterwards [08:23] Mirv: good! [08:23] Mirv: binary removed FYI [08:24] (after rechecking the rdepends) [08:24] ok, great [08:25] stopping the build job [08:36] tested, and published as there was already a packaging ack from a certain core dev on the change [08:37] 'A certain core dev' ;) [08:51] Is that a build-dep on a runtime library package? [08:54] Tolerable enough for this sort of purpose [08:54] Though I'd have used liboxideqt-qmlplugin to avoid the soname dependency [08:55] I get artifically causing dep waits, but there must be a better way than the library package [08:55] That's why I worded it like that. [08:58] popey: python-statgrab - interface to the libstatgrab library for Python [08:58] thanks davmor2 [09:04] ev: does whoopsie have smarts not to upload .crash files over 3g? [09:04] popey: yeah, so long as network-manager is around [09:04] it uses that API to make the determination [09:04] ok [09:06] davmor2: http://popey.com/~alan/phablet/device-2014-04-08-100638.png [09:06] scopes look screwy there for me. [09:07] popey: the images are missing I see them here [09:07] i am on 3g and have run out of credit [09:08] i dont think our UI should degrade badly when we have no internet connection [09:08] popey: that'll be why then 3g seems to only display data [09:08] the same would happen if i was out of range [09:09] popey: need to reboot my server back in a tick [09:13] also davmor2 http://popey.com/~alan/phablet/device-2014-04-08-101338.png [09:15] cjwatson: ah, I was with you on that one ;) (but I guess the qmlplugin one will have version soon thinking about it) [09:15] psivaa: could you give me a poke once the re-run of tests finishes? [09:16] sil2100: sure [09:16] bzoltan: so, we might have a small problem ;/ [09:18] sil2100: what is that? [09:18] popey: they will tell you "waiting on design input" [09:19] popey: what I got yesterday with the issue we had with the store being down [09:20] i dont think caching icons locally is a design issue [09:21] mhr3_: ^ [09:21] the scope should simply use a cache and only download diffs [09:21] bzoltan: so, qtcreator-plugin-ubuntu is blocked in -proposed because I missed that unity-scope-tool is not available for ppc-like platforms [09:21] bzoltan: because it's generated out of unity8 source, which has mir dependencies ;/ [09:21] didrocks: uh. [09:21] (so that the UI can be decoupled from the server a bit) [09:22] didrocks, scope has nothing to do with icons [09:22] didrocks, also, that didn't change one bit compared to old scopes [09:22] it displays them, no ? [09:22] no [09:23] mhr3_: the scopes scope isn't the one giving the url to unity8, [09:23] ?* [09:23] so what drwas that rsater with icons in it ? [09:23] *raster [09:23] didrocks, right, and unity can handle caching globally [09:23] at some point [09:23] thgis is clearly a caching issue http://popey.com/~alan/phablet/device-2014-04-08-101338.png ... whatever part of the system is responsible for it [09:23] sil2100: ohh... do not worry. Feel free to drop it now. [09:24] i did scroll through all of those icons while on wifi [09:24] then dropped wifi and see squares [09:24] well, and even on wifi scrolling through them is awful depending on your wlan speed [09:24] bzoltan: the package you mean? [09:25] bzoltan: or support for ppc, ppc64 and amd64? [09:25] sil2100: either of them [09:26] also... http://popey.com/~alan/phablet/device-2014-04-08-102620.png === vrruiz_ is now known as rvr [09:27] also http://popey.com/~alan/phablet/device-2014-04-08-102643.png [09:32] mhr3_: do we have a bug for this already? [09:32] i have a vauge recollection of one but cant find it [09:32] popey, the issue is there for over 6 months, i'm pretty sure we do [09:33] and the "not seeing anything" issue is already listed as blocker [09:35] didrocks: or oxideqmlscene ... [09:35] popey, https://bugs.launchpad.net/unity8/+bug/1224998 [09:35] Ubuntu bug 1224998 in Unity 8 "Cache preview and thumbnail images" [Medium,Triaged] [09:35] popey, https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1271963 [09:35] Ubuntu bug 1271963 in unity8 (Ubuntu) "Blank icons in "More suggestions" when no internet connection available" [Undecided,Confirmed] [09:35] thanks [09:36] didrocks: bug #1304292 is reported for telepathy-ofono crash occurred today [09:36] bug 1304292 in telepathy-ofono (Ubuntu) "telepathy-ofono crashed with SIGSEGV in QVariant::QVariant()" [Undecided,New] https://launchpad.net/bugs/1304292 [09:36] cjwatson: I was less thrilled by that one as I'm unsure unity-webapps-qml is going to use it directly, but it's posible [09:36] psivaa: thanks! [09:36] sil2100: you could make the unity-scope-tool dependency architecture-specific, perhaps [09:36] popey, also, you opened the latter :) [09:36] \o/ [09:37] Hm. I think this click change has exposed a pre-existing content-hub crash [09:38] Let's see if I can make it more obvious ... [10:02] popey, but it might take some time [10:02] popey, you can just Ctrl+C [10:03] popey, and upload the .crash somewhere for us - whoopsie will upload nevertheless, and drop a corresponding, root-owned .uploaded file, which means the next time it crashes, it will overwrite the whole thing [10:08] Saviq: http://chinstrap.canonical.com/~alan/_usr_bin_unity8.32011.crash [10:09] awesome, finally a non corrupt one [10:09] :) [10:09] bug 1304315 is the bug to go with it Saviq [10:09] Error: Launchpad bug 1304315 could not be found [10:09] didrocks: sil2100: the rerun is now complete, with gallery failing the same tests, sudoku passing, and telepathy-ofono not crashing during messaging-app test [10:10] popey, thanks! [10:10] np [10:10] psivaa: thanks! [10:10] yw :) [10:10] sil2100: trying the sdk revert on gallery? ^ [10:10] sil2100: didrocks: what is worng with the SDK? [10:11] bzoltan: we are trying to understand the gallery issue on the dashboard [10:11] bzoltan: trying to point what could have impacted that [10:11] doe anyone know why silo-06 was flushed ? [10:12] (that corresponds with line 12 on the pending sheet) [10:13] didrocks, ^^^any idea ? [10:13] huh? silo-006 is mine [10:13] cjwatson, since when ? [10:13] yesterday [10:13] well, it was the media-hub one before ... [10:13] ah, I guess line 12 was the previous occupier [10:14] right [10:14] it's line 51 now :) [10:14] didrocks, sil2100 gallery-app tests were already failing for me locally for a while because of broken imports [10:14] and there is no explanation who flushed it and if that was agreed with the devs [10:14] didrocks: I don't know why it used to work on jenkins. But elopio proposed a fix https://code.launchpad.net/~elopio/gallery-app/override_toolbar/+merge/213703 [10:14] (and i cant find anything in backlogs) [10:14] didrocks: I don't know what changed where.. maybe a different python version? [10:15] didrocks: reverting [10:15] sil2100: or try elopio's fix [10:15] sil2100: https://code.launchpad.net/~elopio/gallery-app/override_toolbar/+merge/213703 [10:15] ogra_: see what I told on the other channel, it's all info I have [10:15] k, will try this and then a revert [10:15] t1mp: that looks like it was relying on relative imports, which you don't have in Python 3 (and which were madness anyway ...) [10:15] didrocks, which other channel ? [10:15] ogra_: #phablet? [10:15] I wrote there [10:16] oh, probably before i re-joined [10:16] about the comment which was set, it's all the info I had [10:16] ok [10:16] would really help if people put their name on it if making such comments [10:16] t1mp: well, rather, you do have them, you just have to spell them differently (leading ".") [10:16] cjwatson: did the image recently switch to python 3? [10:16] t1mp: I don't know the timing details [10:16] ogra_: +1 [10:17] but there was a big push to get autopilot onto python 3 yes [10:17] the weird thing is that I have had tests failing on that for a while now, while they didn't seem to fail on jenkins. Jenkins caught up now ;) [10:17] hmm.. maybe one of the packages that I installed for testing upgraded python on my device [10:18] rather, upgraded autopilot [10:18] (you don't upgrade from python 2 to 3 as such; they're installed in different paths, /usr/bin/python vs. /usr/bin/python3) [10:19] dbarth: your unity-webapps-qml silo could now be rebuilt, the other one landed & merged to trunk [10:19] I tried to run the tests the same way jenkins does it, but apparently there still were differences [10:20] sil2100, silo for 54 pls [10:21] mhr3_: I can has low on silos, but let me see what we can do about that [10:21] Morning all [10:22] sil2100, it's blocker fix [10:22] Mirv: right, i just saw that [10:22] Mirv: will proceed now [10:22] didrocks: bug of the day Time set itself to 1 of Jan 1970 00:00 oops [10:22] sil2100: morning...could you reconfig silo 005? [10:23] kgunn: morning! Let me reconfig [10:23] kgunn: btw. I'm publishing the other one you set to ready, thanks [10:29] kgunn: ok, so, we need to wait for platform-api to appear in the archive from the other landing [10:30] kgunn: I will ask you then to rebuild everything, ok? SInce I see that landing 005 has platform-api [10:30] popey, can you please subscribe me to bug #1304315 [10:31] Saviq: done [10:31] popey, o/ [10:32] sil2100: ack [10:32] popey, can I make public? [10:32] Saviq: if its got nothing sensitive in it, sure [10:32] popey, core is gone already [10:32] dunno if my Google passwords and stuff would be in the crashdump [10:32] ok [10:32] I've lost my control instance - is anyone else seeing content-hub crashes in a current-ish proposed image? [10:33] feel free [10:33] tedg, bug #1304315 [10:33] tedg, looks like libUAL crashed in some string quoting [10:34] popey, any idea which app you were launching? [10:34] Saviq: yeah, G+ webapp, says in the bug [10:34] net.launchpad.click-webapps.googleplus [10:34] thats the last thing to be installed [10:38] popey, can't repro, but trace is rather good, should point us at what happened [10:38] yay [10:40] bzoltan: what do you think about this? https://code.launchpad.net/~sil2100/qtcreator-plugin-ubuntu/arch_specific_scope_dep/+merge/214710 [10:40] bzoltan: can I add it to your landing, rebuild and try re-publishing? [10:40] sil2100: ohh.. that is an elegant solution [10:44] sigh [10:44] my update manager UI vanished again during OTA [10:44] popey, davmor2 ^^^^did i ask you about that before ? i'm seeing that since a few days [10:45] nope, not heard that one [10:45] the whole progressbar UI part just vanishes [10:45] did you rotate? [10:45] sometimes i can trigger it easily by scrolling upwards [10:45] i have seen odd things happen in that app if you rotate it [10:45] sometimes it just happens on its own (like it did now) [10:45] nope [10:45] it lies flat on its back [10:46] it also gives you the restart screen after a while ... but until it does the screen is completely empty .. i only have the header [10:47] hmm, still greyed out scopes on boot [10:47] * ogra_ has to swipe once from the right to get the screen even active) [10:48] oh, g* now gets the scaling of the welcome screen right [10:48] *G+ === MacSlow is now known as MacSlow|lunch [10:53] ogra_: yeap and now when you follow a link that is scaled correctly too [10:53] i dont want that to be scaled at all [10:53] i want that to open in the browser :P [10:54] (like the webapp-container commandline from teh .desktop file implies ... but that doesnt seem to work) [10:56] i think i may have found my memory leak [10:56] ps_2014-04-08-104122.txt: 18.3 MiB + 609.5 KiB = 18.9 MiB ubuntu-location-serviced [10:56] * ogra_ hands popey a bucket [10:56] ps_2014-04-08-104726.txt: 21.6 MiB + 483.5 KiB = 22.1 MiB ubuntu-location-serviced [10:56] ps_2014-04-08-105332.txt: 24.9 MiB + 486.0 KiB = 25.4 MiB ubuntu-location-serviced [10:56] * popey passes the bucket to tvoss ☻ [10:56] what are oyu measuring there ? [10:56] RES ? [10:56] using a random python script i found online [10:57] lol [10:57] https://raw.github.com/pixelb/ps_mem/master/ps_mem.py [10:57] so it randomly generates numbers it spits out ? [10:57] heh [10:57] it generates nice sane output [10:58] * ogra_ would just have used top [10:58] sure, this is pretty tho ☻ [10:58] pretty is a feature [11:01] sil2100: didrocks, Mirv: the silo15 is good to go [11:03] bzoltan: \o/ [11:03] bzoltan: taking care of that one then [11:05] sil2100: thank you [11:05] didrocks, t1mp: so, I ran gallery-app with Leo's changes and got 12 failures [11:05] didrocks, there's a loose bot on the run! it's automerging approved branch for gsettings-qt! [11:05] didrocks, kill it now! [11:06] mhr3_: uh oh! [11:06] sil2100: can you try reverting sdk then and see? [11:06] mhr3_: for what project is that? [11:06] mhr3_: this is for cihelp ^ [11:06] mhr3_: I guess we need to disable it from cu2d [11:06] mhr3_: nice approach! :D [11:07] didrocks, indeed, help the poor projects, ci people! [11:08] sil2100, try reading the last word ;) [11:08] in that sentence [11:09] mhr3_: let me see, would help if you have a jenkins link by any chance [11:09] i do not [11:11] mhr3_: hah, right ;) [11:11] psivaa: nothing in cupstream2distro-config from what I see [11:12] sil2100: you are trying the sdk revert first? [11:12] didrocks: yes, just reverted it, running the tests again now [11:12] sil2100: otherwise, maybe check with psivaa if we run the python2 or 3 flavors and if anything changed [11:12] * sil2100 does a few things at once [11:12] ok [11:18] sil2100: it's already disabled from cu2d (like everything else for that matter), it's the automergers that may be still lurking [11:18] for some of the rarer updated projects [11:20] but automerger is using cu2d-config and I don't find it there [11:20] Was it redeployed? [11:21] didrocks: would https://code.launchpad.net/~ken-vandine/cupstream2distro-config/gsettings-qt/+merge/171634 have anything to do with mhr3_ 's issue? [11:22] psivaa: that's very very old, before CITrain was invented [11:22] gsettings-qt was in cu2d, but Mirv says it got disabled [11:22] psivaa: if you look at the config file [11:22] it's set to gsettings-qt: [11:22] daily_release: False [11:22] I would say it wasn't redeployed with the new config [11:22] autolanding_template: False [11:22] use_stack_ppa: False [11:22] which suppose to disabling upstream mergerd [11:22] sil2100: the deployement is automated [11:22] for upstream merger [11:22] didrocks: ah, ok [11:24] didrocks, bzoltan, t1mp: sadly, reverting the SDK helped :< [11:24]