[00:07] <sergiusens> rsalveti, can you monkey press here: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0AuDk72Lpx8U5dFlCc1VzeVZzWmdBZS11WERjdVc3dmc&usp=drive_web#gid=27 ?
[00:10] <cyphermox> kgunn: ping
[00:11] <cyphermox> kgunn: curious about the current state of testing Mir, and an ETA if possible. I was thinking about going to bed early since I'm a bit sick
[00:14] <robru> cyphermox, he went to dinner, I tried testing it a little bit but the phone won't boot with the new mir installed. not looking great
[00:28] <robru> sergiusens, just published silo 9, you can merge in an hour or so
[00:34] <sergiusens> thanks
[00:52] <kgunn> robru: stupid me!...i have an idea....
[00:52] <kgunn> its the packages...you don't just "dist-upgrade"
[00:52] <robru> kgunn, yeah, I installed the new packages manually... that's what wasn't working.
[00:52] <robru> kgunn, actually I tried it once with and once without dist-upgrading
[00:52] <robru> kgunn, did you get it working on your device?
[00:53] <kgunn> robru: i had issues...i just reflashed...
[00:53] <kgunn> let you know in a moment
[00:53] <robru> ok
[01:01] <cjohnston> /8
[01:15] <kgunn> robru: sorry...on a call with my australian guys but yeah...seeing what you're seeing
[01:15] <robru> kgunn, any hope for a quick fix? (i have no idea)
[01:17] <kgunn> robru: let's try something simple...can you reconfig ? i've pared down the mp list...but will only need to rebuild usc & unity-mir
[01:17] <robru> kgunn, sure thing
[01:17] <kgunn> robru: its gotta be unity-mir
[01:18] <kgunn> its the only thing different
[01:18] <robru> kgunn, wow you really stripped down that list... it's like 1/3 as long ;-)
[01:18] <robru> kgunn, ok, reconned, please build
[01:18] <kgunn> robru: thank you sir
[01:19] <robru> kgunn, you're welcome
[01:28] <robru> kgunn, oh man, in 30 minutes cron will kick a new image build. I was hoping to have mir landed before then, I guess we won't make it ;-)
[01:29] <kgunn> no worries
[01:29] <kgunn> btw, i say it must be unity-mir, because those are the only MP's we farted with
[01:29] <kgunn> and we had built packages that booted and ran
[01:29] <kgunn> with the exact same MP list
[01:34] <robru> kgunn, i see
[01:45] <robru> kgunn, oh, just noticed the build is done. installing on my mako
[01:45] <rsalveti> sergiusens: hey, was having dinner, but it seems you're all set already :-)
[01:48] <sergiusens> rsalveti, yeah, you cn rid of goldfish soonish-now
[01:49] <rsalveti> great
[01:49] <kgunn> robru: hooray...mine booted....yours ?
[01:49] <robru> kgunn, ok, freshly flashed mako, silo 2 installed, rebooted... stuck on Google screen
[01:50] <robru> kgunn, oddly I can ssh into it...
[01:51] <kgunn> hmmm....
[01:51] <kgunn> mine is definitely updated to the ppa
[01:52] <robru> kgunn, trying a dist-upgrade on it
[01:52] <kgunn> robru: any chance you pulled the ppa too early (before copy was done?)
[01:52] <kgunn> robru: note...i did not dist-upgrade at all....
[01:52] <kgunn> only sudo apt-get install libmirplatformgraphics-android libmirclientplatform-android libmirplatform
[01:52] <robru> kgunn, dunno, it was just now.... if it was premature then dist-upgrade should help
[01:53] <robru> kgunn, k, will reflash and try again
[01:54] <kgunn> robru: dist upgrade will pull in libmirplatformgraphics-mesa & libmirclientplatform-mesa...which you don't want
[01:55] <robru> kgunn, if that's the case then how are those not included in the images?
[01:55] <kgunn> robru: its a note in the landing spreadsheet...in this mir we've changed/added some new packages....splitting out android vs mesa...
[01:56] <kgunn> the 2 subsystems we use...today mesa on desktop and android on phone/tablet
[01:56] <robru> kgunn, yeah but like... if a dist-upgrade pulls in bad packages, then what is stopping those packages from getting pulled in during the image build? I'm terrified we'll publish this an then the next image build will have those packages and not boot
[01:57] <kgunn> robru: rsalveti knows how to seed packages appearntly...
[01:57] <robru> kgunn, oh, I see, they need to be seeded.
[01:57] <kgunn> which will "magicaly" take care of this
[01:57] <robru> yeah, I'm not very familiar with the seeds.
[01:57] <rsalveti> yeah, I can update the seeds once you land mir
[01:57] <robru> kgunn, well we have to make sure that the seed happens around the same time as the publish (at least before the next image build)
[01:57] <robru> rsalveti, great, thanks
[01:57] <rsalveti> just let me know
[02:00] <kgunn> robru: ok..manual testing seems ok. just gotta run a select few AP tests...
[02:01] <robru> kgunn, make sure whatever aborted the last landing is fixed now ;-)
[02:01] <kgunn> robru: totally...i'm the one who aborted remember :)
[02:01] <robru> kgunn, ;-)
[02:02] <sergiusens> rsalveti, add the seed/meta to the ppa
[02:02] <sergiusens> that's what I did
[02:02] <sergiusens> easier to test as well
[02:02] <sergiusens> and with that comment I go to sleep
[02:02] <sergiusens> cheers
[02:02] <rsalveti> sergiusens: yeah, but guess I'll just dput it directly once kgunn is done with it
[02:10] <robru> kgunn, oh ho ho! it boots ;-)
[02:12] <kgunn> robru: now i wonder if those mp's were ok...i didn't do it clean the first time...i did an dist-upgrade, followed by a remove mesa/install android
[02:13] <kgunn> either way...we'll land this first....if my next ap run passes
[02:13] <robru> kgunn, yeah, I have a little helper script that grabs & installs *all* binary packages from a PPA, so it was pulling in that mesa stuff, no wondering it was so broken.
[02:13] <robru> kgunn, yeah, just running unity8 AP myself.
[02:13] <kgunn> damn i bet those mps are ok..oh well
[02:24] <robru> kgunn, damn, it's great that AP test coverage is increasing, but holy crap it takes forever to run them
[02:24] <kgunn> no kidding...i keep leaving the room....but damn its still running
[02:28] <robru> kgunn, 1 failure in unity8 AP!! abort the whole thing!!! ;-)
[02:29] <robru> kgunn, seems fine to me I guess. let me know when you're satisfied and I'll publish
[02:34] <rsalveti> robru: kgunn: guess you guys are still trying to land this today, so I'll disable cron so we don't get a new image in 26min
[02:35] <rsalveti> done
[02:35] <robru> rsalveti, oh, ok. should be landable soon, just waiting for kgunn's say-so
[02:36] <robru> thanks
[02:36] <rsalveti> no worries, just to avoid any in-flight issue
[02:37] <kgunn> robru: rsalveti ...let's do it...
[02:37] <kgunn> all good
[02:37] <rsalveti> \o/
[02:38] <robru> kgunn, alright, just need a core dev to ack all those packaging changes.
[02:40] <robru> rsalveti, interestingly it seems you are a core dev. can you sanity check the four diffs at http://162.213.34.102/job/landing-002-2-publish/ ?
[02:41] <rsalveti> sure
[02:41] <robru> thanks
[02:43] <rsalveti> robru: looks good
[02:43] <robru> rsalveti, thanks
[02:44] <robru> rsalveti, ok, i'm publishing. can you update the seed and then kick an image build once everything is through -proposed
[02:44] <robru> ?
[02:44] <rsalveti> robru: sure
[02:44] <robru> rsalveti, thanks!
[03:26] <robru> rsalveti, kgunn: mir packages made it to distro, so I'm merging & cleaning, so I can assign the next silo
[03:26] <rsalveti> cool, just saw everything got promoted
[03:26] <rsalveti> nice, let me push the meta-package changes
[03:26] <robru> thanks
[03:34] <robru> kgunn, alright, you got silo 2 again, please build ;-)
[03:36] <kgunn> woohoo
[03:36] <robru> kgunn, I'm EOD 3 hours ago however, so no rush to land this one ;-)
[03:36] <robru> good night!
[03:37] <kgunn> robru: thanks for all the help...and of course, sir...good night
[03:37] <robru> kgunn, you're welcome!
[04:37] <rsalveti> [05:29] <Mirv> landing-006 would need more space, 10GB full :(
[05:35] <rsalveti> [05:36] <rsalveti> but for some reason it got both libmirclientplatform-android and libmirclientplatform-mesa, and that's not good
[05:37] <robru> ugh, I knew this was going to happen.
[05:38] <rsalveti> not sure why yet
[05:39] <rsalveti> apt-get update/dist-upgrade worked fine after changing the meta package
[05:40] <rsalveti> flashing cdimage and will check
[05:40] <robru> rsalveti, thanks.
[05:40] <rsalveti> meta is fine
[05:41] <rsalveti> libmirclientplatform-android, libmirplatformgraphics-android
[05:42] <rsalveti> http://people.canonical.com/~ubuntu-archive/livefs-build-logs/trusty/ubuntu-touch/20140311/
[05:44] <rsalveti> afaik both will have the same alternatives priority
[05:45] <rsalveti> Setting up libmirclientplatform-android:armhf (0.1.6+14.04.20140310-0ubuntu1) ...
[05:45] <rsalveti> update-alternatives: using /usr/lib/arm-linux-gnueabihf/mir/clientplatform/android/ld.so.conf to provide /etc/ld.so.conf.d/arm-linux-gnueabihf_mirclientplatform.conf (arm-linux-gnueabihf_mirclientplatform_conf) in auto mode\
[05:45] <Mirv> robru: since the new unity-mir/u-s-c is also now set to tested, is it supposed to be published as well? I'm mostly interested since Qt 5.2 testing is broken until I can rebuild unity-mir
[05:45] <rsalveti> it might still work, let's see :-)
[05:46] <robru> Mirv, oh, is kgunn still around to answer that question? I didn't think he'd test it so quickly since I only assigned that silo very recently (like 2 hours ago). I guess it can be published, sure.
[05:47] <Mirv> robru: yeah, it's interesting but it seems he tested it really quickly :)
[05:47] <Mirv> "manual and AP tests look good" it says
[05:49] <robru> Mirv, publish away then! ;-)
[05:49] <robru> although if rsalveti can't resolve this seed issue the whole thing might get reverted.
[05:50] <Mirv> robru: done. yeah, true, if there's a problem with the whole it might get reverted.
[05:51] <Mirv> robru: oh, actually, pkging changes
[05:51] <Mirv> robru: so I can't publish
[05:52] <Mirv> unless cyphermox is also awake in the middle of the night http://162.213.34.102/job/landing-002-2-publish/47/artifact/packaging_changes_unity-mir_0.3+14.04.20140311.1-0ubuntu1.diff + http://162.213.34.102/job/landing-002-2-publish/47/artifact/packaging_changes_unity-system-compositor_0.0.2+14.04.20140311.1-0ubuntu1.diff
[05:53] <robru> Mirv, rsalveti is a core dev... 3AM for him but I'm sure he'd love to review packaging diffs since he's here right now ;-)
[05:55] <Mirv> let's see about that :) otherwise I'll wait until didrocks wakes up in about 1.5h
[05:56] <robru> oh yeah, that's probably a better idea ;-)
[05:58] <rsalveti> lrwxrwxrwx 1 root root 60 Mar 11 05:54 /etc/ld.so.conf.d/arm-linux-gnueabihf_mirclientplatform.conf -> /etc/alternatives/arm-linux-gnueabihf_mirclientplatform_conf
[05:58] <rsalveti> lrwxrwxrwx 1 root root 66 Mar 11 05:54 /etc/alternatives/arm-linux-gnueabihf_mirclientplatform_conf -> /usr/lib/arm-linux-gnueabihf/mir/clientplatform/android/ld.so.conf
[05:58] <rsalveti> so we're still good
[05:58] <rsalveti> robru: both packages got installed, but alternatives is using the one provided by the android backend
[05:59] <rsalveti> but, ideally the mesa packages shouldn't be installed at all
[05:59] <Mirv> rsalveti: nice!
[06:00] <Mirv> the reverts are a hassle anyway
[06:00] <rsalveti> now the diffs
[06:00] <robru> rsalveti, oh ok. weird.
[06:00] <rsalveti> I just don't know why
[06:00] <rsalveti> +++ unity-mir-0.3+14.04.20140311.1/debian/com.canonical.Unity.conf was removed
[06:00] <robru> rsalveti, haha, i was joking about the diffs. go to bed!
[06:00] <rsalveti> but the rest looks fine
[06:00] <rsalveti> :-)
[06:01] <robru> I guess we'd have to ask mterry why that file got dropped, his name is in the changelog. i have no idea.
[06:01] <robru> or kgunn of course ^^
[06:01] <rsalveti> yeah
[06:02] <robru> both of those guys are US TZ so that's not looking good for Mirv's shift today...
[06:02] <Mirv> https://code.launchpad.net/~mterry/unity-mir/drop-dbusscreen/+merge/202236
[06:02] <rsalveti> robru: so, we don't have any packages bringing mir-mesa* explicitly
[06:02] <Mirv> greyback approved it
[06:02] <rsalveti> I believe it's just apt/livecd including them when doing debootstrap
[06:03] <Mirv> aha, "Drop dbusscreen code because it has moved to unity-system-compositor."
[06:03] <robru> Mirv, that makes sense I guess. publish?
[06:03] <rsalveti> yeah
[06:04] <robru> rsalveti, just flashing the image now, just to be sure it works. when I was testing it, the phone wouldn't boot when both were installed
[06:04] <rsalveti> at least 230 is not busted
[06:04] <rsalveti> robru: I believe that's because the mesa ones got installed first
[06:04] <robru> rsalveti, ahhhh, ok
[06:05] <Mirv> robru: does not look that simple, https://code.launchpad.net/~thomas-voss/unity-mir/qt-no-keywords/+merge/209678 states that process-cpp would need updating too https://code.launchpad.net/~thomas-voss/process-cpp/add_death_observer_for_child_processes/+merge/204629
[06:05] <Mirv> on the other hand of course, it's said to be tested
[06:06] <robru> Mirv, huh I thought tvoss did that landing already
[06:06] <robru> oh, i see it's blocked by qt. sigh
[06:07] <Mirv> eh, it's not really blocked, like Mir wasn't
[06:07] <Mirv> so instead it should be assigned a silo and built
[06:08] <Mirv> well, I'll check with didrocks once again, I'm not sure then if that was meant to be landed or not
[06:08] <robru> Mirv, ok
[06:08] <Mirv> now that we started on Mir but maybe did not realize process-cpp would be needed to for the further fixes
[06:08] <robru> rsalveti, ok, phone boots, I'm happy ;-)
[06:09] <Mirv> robru: ah ha, the same branch is in both tvoss' and kgunn's landings :D
[06:09] <robru> Mirv, but wait, if kgunn tested the silo, how did it work without that other stuff? maybe it's not really necessary?
[06:09] <robru> Mirv, oh, ok
[06:09] <robru> Mirv, that makes more sense
[06:10] <rsalveti> robru: :-)
[06:10] <Mirv> robru: well it still doesn't makes sense in that only unity-mir branch is in kgunn's landing, while both unity-mir and process-cpp are in tvoss's
[06:11] <Mirv> anyhow, this'll get sorted out sooner or later
[06:12] <robru> Mirv, ok. 11PM here, I'm going to bed ;-)
[06:12] <robru> good night!
[06:12] <Mirv> robru: good night!
[06:12] <rsalveti> same
[06:13] <rsalveti> Mirv: mind testing qtwebkit from https://launchpad.net/~rsalveti/+archive/qtwebkit-b ?
[06:13] <rsalveti> Mirv: build on top of the landing 6 ppa, trying to bring back the add_experimentalDevicePixelRatio patch
[06:13] <rsalveti> *built
[06:14] <rsalveti> just finished, still didn't try it
[06:14] <rsalveti> might not necessarily fix the issue, as I saw another class that I might need to change, but good to test it anyway
[06:14] <Mirv> rsalveti: ok, although now landing-006 is not fully functional because of the Mir landing but doable. so that's the alternative option to Kaleo's qtubuntu/UITK fixes?
[06:15] <rsalveti> Mirv: yeah
[06:15] <Mirv> Kaleo's qtubuntu fails to build now, so it's useful there's something to test. I might give pope_y a poke too since he's familiar with the issues.
[06:15] <rsalveti> cool
[06:16] <rsalveti> in the end Kaleo said that there is no easy way to fix the remaining issues with his fix
[06:16] <rsalveti> so we decided to give that patch a try again
[06:16] <rsalveti> so we can unblock the landing
[06:16] <rsalveti> will get some sleep, let me know if something bad happens after installing that package :-)
[06:16] <Mirv> ok. we should have something to report by the meeting.
[06:17] <Mirv> yeah, do that :)
[06:19] <rsalveti> later
[07:39] <Mirv> didrocks: so, we eventually left silo 002 unpublished (further unity-mir/u-s-c fixes) with robru in the morning. it was said to be tested by kgunn, but one commit there has a comment which says the tvoss's process-cpp should land first - and actually the related unity-mir landing is in the both landings now.
[07:40] <Mirv> so it was left uncertain if it really works as is without the process-cpp update
[07:40] <didrocks> Mirv: ok, sound legit, can you had a comment about that and check with tvoss/kgunn once around?
[07:41] <Mirv> yeah mainly tvoss: ^ could check if it's possible https://code.launchpad.net/~thomas-voss/unity-mir/qt-no-keywords/+merge/209678 works also without the process-cpp update https://code.launchpad.net/~thomas-voss/process-cpp/add_death_observer_for_child_processes/+merge/204629
[07:41] <Mirv> since that's basically what kgunn's landing now says
[07:58] <tvoss> didrocks, Mirv yup, works without process-cpp changes
[07:58] <didrocks> Mirv: I guess you can land then (are retake the lock/rebuild ;))
[07:58]  * didrocks waits on image result to get AP in
[08:04] <Mirv> tvoss: didrocks: thanks!
[08:04] <Mirv> packaging changes were checked by rsalve_ti
[08:06] <didrocks> sweet :)
[08:06] <Mirv> and after that I'll indeed get landing-006 back to usable state since it's now not upgradeable to (unity-mir in archives newer than in PPA)
[08:06] <sil2100> Will that unity-mir landing land before landing AP?
[08:06] <sil2100> Ah, see them migrating now ;)
[08:07] <Mirv> yes :)
[08:07] <didrocks> sil2100: yeah, I just want if we can ensure that current image is good before publishing AP
[08:07] <Mirv> didrocks: I hit the 10GB limit in Qt PPA by the way, I didn't anticipate ppc64el + arm64 builds which we now both have..
[08:07] <didrocks> sil2100: like, let's wait to get the sdk results
[08:07] <Mirv> I had it bumped to 20GB
[08:07] <didrocks> Mirv: ah, done already, rocking :)
[08:07] <sil2100> didrocks: true, makes sense
[08:07] <didrocks> yeah, ppc64el and arm64…
[08:07] <didrocks> :)
[08:08] <didrocks> Mirv: how is arm64 rebuild going btw?
[08:10] <Mirv> didrocks: pretty good, Qt fixed to the extent it will be fixed at this point in time, and I'm progressing on rebuilding all the rest.
[08:10] <didrocks> sil2100: I'm sad that corey didn't answer though on the autopiot-qt :/
[08:10] <didrocks> Mirv: perfect! no blocker so far?
[08:11] <Mirv> didrocks: nope, similar symbols updates to various packages as with ppc64el. Colin solved the qtdeclarative FTBFS (although we didn't have that before either for arm64) and I forward-ported qtscript so Qt = good.
[08:11] <didrocks> great ;)
[08:11] <didrocks> the copy to archive will be a breaze I guess! :)
[08:11] <Mirv> so far also the ~20 other builds have been fine, I'm progressing over the day
[08:11] <didrocks> breeze*
[08:12] <Mirv> didrocks: hehe, one can wish for :) it's at least a whole lot easier than it would have been without these ppc64el + arm64 preparations
[08:12] <didrocks> Mirv: did you have any meeting in the end? Progress on the AP side?
[08:12] <didrocks> yeah
[08:12] <Mirv> didrocks: today at 1pm UTC, no meeting yesterday except for syncup. elopio continues to look at the tests - notes app mentioned to be non-blocker as it will be replaced. weather-app has a fix/workaround pending apparently.
[08:13] <didrocks> Mirv: yeah, but for notes-app he's going to workaround it as we discussed, right?
[08:13] <didrocks> (on Friday)
[08:13] <didrocks> elopio: mind confirming? (you are going to remove the "click on tab"?) ^
[08:13] <Mirv> didrocks: maybe, we didn't catch up well yesterday.
[08:14] <Mirv> entirely plausible
[08:14] <didrocks> ok
[08:14] <didrocks> we shouldn't ship broken tests
[08:14] <didrocks> either we remove them
[08:14] <didrocks> or we fix them (workaround)
[08:14] <didrocks> otherwise, it's confusing the report on the dashboard
[08:15] <thostr_1> can I get a silo for line 25?
[08:17] <sil2100> thostr_1: let me see
[08:18] <sil2100> thostr_1: sure
[08:18] <sil2100> thostr_1: btw. silo 004 has packages built since a while now
[08:20] <sil2100> thostr_1: assigned
[08:20] <thostr_1> sil2100: thanks. yes, 4 is under testing
[08:50] <tvoss> sil2100, didrocks can I get a silo for line 19?
[08:53] <sil2100> tvoss: let me see that one
[08:53] <sil2100> tvoss: you mean the process-cpp one, yes?
[08:54] <tvoss> sil2100, yup, the unity-mir stuff has already landed
[09:03] <Mirv> restealing unity-mir to Qt 5.2
[09:03] <sil2100> I'm assigning a silo for process-cpp then ;)
[09:07] <didrocks> Mirv: stealer! :p
[09:07] <tvoss> sil2100, \o/
[09:17] <mhr3> Mirv, sooo, when do we have 5.2 in archive?
[09:17] <Mirv> mhr3: when bug #1207270 is fixed and enough autopilot tests breakage has been fixed to satisfy everyone
[09:18] <didrocks> (and arm64 work done)
[09:18] <Mirv> arm64 is almost as good as done! :) granted, it'll take until EOD at least
[09:18] <sil2100> \o/
[09:19] <Mirv> I've to update that bug though, there's qtwebkit from rsalveti trying to address the problem (and seems to work for me)
[09:19] <didrocks> Mirv: perfect! yeah, I don't see that will be the blocker, but best to mention/not forget about :)
[09:20] <didrocks> Mirv: maybe we'll play the same "I steal you/you steal me" game with u-s-s today for seb128
[09:21] <seb128> ;-)
[09:22] <Mirv> didrocks: ok, just let me know. it breaks Qt 5.2 PPA usage after it has landed in the archives so it's useful to launch the Qt PPA rebuild as soon as possible
[09:22] <didrocks> Mirv: yeah, it's a leaf though (and more isolated than unity-mir/u-s-c), so should be easy and constrained
[09:22] <Mirv> luckily, only at that point so it's just merge+clean & rebuild that needs to be done at that point
[09:22] <didrocks> Mirv: why merge and clean?
[09:24] <Mirv> didrocks: I mean, when you land u-s-s from another silo, Qt 5.2 PPA is broken until the u-s-s silo is merge+cleaned so that it can be rebuild again in the Qt 5.2 PPA
[09:24] <Mirv> so that the rebuild has the newest source
[09:24] <didrocks> Mirv: oh yeah
[09:25] <didrocks> seb128: ubuntu-themes landed
[09:25] <didrocks> clock and weather back to… 100% \o/
[09:25] <didrocks> Mirv: so, Qt 5.2 should have 100% tests passing with the current results
[09:26] <didrocks> that's way easier to say "yes or no" :)
[09:26] <didrocks> \o/\o/\o/
[09:26] <didrocks> -> insert music here
[09:26] <sil2100> YEAAH
[09:26] <sil2100> ;)
[09:27] <Mirv> didrocks: yeah, I have to ask how the gallery-app tests will be run (manual?) since it needs a recompilation
[09:27] <Mirv> anyhow, \o/ for 100%
[09:28] <didrocks> Mirv: hum, not sure to understand about gallery-app, click package right?
[09:29] <didrocks> Mirv: I guess that was already discussed with sergio, wasn't it?
[09:29] <didrocks> (I was not into those)
[09:29] <Mirv> didrocks: yes, so it's a click package so a rebuild of it in the PPA won't help anything since it's not installed from archives
[09:29] <Mirv> we'll check that in the meeting in 3.5h
[09:29] <didrocks> Mirv: yeah
[09:30] <didrocks> Mirv: do you think I should come?
[09:30] <didrocks> hum, no email for the morning meeting, weird…
[09:30] <sil2100> didrocks: https://plus.google.com/hangouts/_/calendar/Y2Fub25pY2FsLmNvbV91cTRvNmQyMWJvNmJ0bm1mcW9xZWtsNTdnOEBncm91cC5jYWxlbmRhci5nb29nbGUuY29t.us2orfbhb8ssqjui2u15tajj3s if anything
[09:31] <didrocks> thanks :)
[09:31] <didrocks> too late :p
[09:31] <didrocks> ogra_: ^
[09:31] <didrocks> davmor2: ^
[09:31] <Mirv> didrocks: maybe it would be good to have everyone updated, but of course optional
[09:31] <didrocks> Mirv: yep
[09:48] <sil2100> tvoss: you can build in silo 002 for testing ;)
[09:48] <tvoss> sil2100, ack and thx
[09:53] <didrocks> Saviq: hey, we have an unity8 crash at least at every boot
[09:54] <didrocks> Saviq: as there a new mir, probably unity-mir?
[09:55] <didrocks> Saviq: http://jenkins.qa.ubuntu.com/job/trusty-touch-mako-smoke-daily/137/artifact/clientlogs/ubuntu_terminal_app/_usr_bin_unity8.32011.crash/*view*/
[09:55] <didrocks> for instance
[09:55] <didrocks> greyback: ^
[09:55] <didrocks> that might interests you as well
[09:55] <greyback> looking...
[09:56] <ogra_> didrocks, bug 1290771
[09:56] <didrocks> ogra_: thanks!
[10:00] <Mirv> hum, seb128 lost internets. I'll ping him when he's back.
[10:02] <Saviq> didrocks, hmm not processed :/, image #230?
[10:02] <didrocks> Saviq: yeah, I just upgraded and confirm I have the crash
[10:02] <Saviq> didrocks, /me flashing as well
[10:02]  * didrocks reboots to see if it's on every boot
[10:02] <didrocks> Saviq: my guess is on boot as we have it at least once per testsuite
[10:03] <Saviq> didrocks, sounds like it, yeah, probably screen off or some such, so respawning works
[10:03] <Saviq> Mirv, seb128's back ↑↑
[10:03] <seb128> Saviq, ?
[10:03] <Saviq> seb128, <-- seb128 has quit (Quit: Ex-Chat)
 hum, seb128 lost internets. I'll ping him when he's back.
[10:03] <didrocks> seb128: u-s-s is unlocked
[10:03] <Mirv> seb128: so ubuntu-system-settings is now deconfigured from Qt 5.2 silo temporarily, you an go ahead
[10:04] <didrocks> Saviq: hum, no crash after this reboot though
[10:04] <seb128> Saviq, didrocks, Mirv: thanks (and I didn't loose internet, I installed the unity ppa for screenlocking in unity, and restarted my box since we got a new lightdm as well I wanted to test)
[10:04] <Saviq> seb128, details
[10:04] <seb128> ;-)
[10:05] <seb128> did you need me for anything? or just the u-s-s slot info? ;-)
[10:05] <didrocks> seb128: just the u-s-s slot info
[10:05] <seb128> good
[10:06] <didrocks> Saviq: weird, no more crash here…
[10:07] <didrocks> but we do have crashes on every tests on both manta and mako
[10:07] <Saviq> didrocks, maybe first boot is longer due to some setup?
[10:07] <Saviq> didrocks, and the devices are reflashed for every suite, are they not?
[10:07] <didrocks> Saviq: no, they aren't… that's why I'm wondering
[10:07] <Saviq> mhm
[10:07] <didrocks> and they are in rw mode, like I am here
[10:08] <didrocks> maybe the faster will be to retrace locally
[10:08]  * didrocks reflashes cleanly first
[10:14] <sil2100> seb128: any news on the unity-gtk-module fix? Can we have a landing for that :D ?
[10:15] <seb128> sil2100, review rounds seem slot on the other change, we should go for your no change mr to do a landing ... what to propose a branch for that? ;-)
[10:16] <didrocks> Saviq: didn't get the crash on a fresh system…
[10:16] <sil2100> seb128: let me propose one then in a moment ;)
[10:16] <seb128> sil2100, thanks
[10:20] <didrocks> davmor2: do you reproduce that crash btw? seems I can't ^
[10:25] <didrocks> Saviq: what's this unity8 update btw?
[10:25] <didrocks> Saviq: seems it's quite late and in conflicts with all session I have to host or attend
[10:25]  * didrocks wonders who add to someone else track without telling
[10:30]  * didrocks reschedules
[10:30] <Saviq> didrocks, don't ask me, it's kgunn that's dealing with those
[10:31] <didrocks> kgunn: can you next time schedule them in advance (and not a week after the deadline), it's hard to schedule it and people are adding to one's track without checking with the track lead first (I'm in sessions on other tracks)
[10:32] <didrocks> kgunn: I've set it on Thursday, first session and moved some other to make it fit
[10:32] <Saviq> didrocks, I got a crash on `restart unity8` here
[10:33] <sil2100> seb128: https://code.launchpad.net/~sil2100/unity-gtk-module/null_flush_trunk/+merge/210374 <- flush merge for u-g-m if anything ;)
[10:34] <didrocks> Saviq: oh, so maybe only test related… (I don't get a crash on normal stop though)
[10:34] <Mirv> popey: davmor2: would either of you have time to test rsalveti's qtwebkit regarding the DPR fix from that angle? in addition to landing-006, you'd need to use ppa:canonical-qt5-edgers/qt5-beta1
[10:34] <seb128> sil2100, cool, just finishing what I'm doing and putting a landing ask for that one next
[10:34] <Mirv> popey: davmor2: we've the Qt 5.2 meeting in 2.5h
[10:34] <Saviq> didrocks, checking on stop
[10:35] <Saviq> didrocks, crashed here
[10:35] <Mirv> popey: davmor2: so, 5.2.1+dfsg-0ubuntu2 of dpkg -l | grep webkit gets installed from that PPA
[10:36] <didrocks> Saviq: confirmed on restart and stop
[10:37] <didrocks> Saviq: weird we don't have it on reboot though
[10:38] <ogra_> [10:38] <didrocks> ogra_: \o/
[10:38] <sil2100> Yeaaa
[10:39] <ogra_> (fell over again while doing it :( )
[10:39] <sil2100> AP still in proposed
[10:39] <Mirv> ogra_: \o/
[10:41] <didrocks> sil2100: autopkgtest for unity-scope-click 0.1+14.04.20140226.1-0ubuntu1: RUNNING (Jenkins: public, private)
[10:41] <didrocks> if passed btw, so should be at next britney's tick
[10:42] <sil2100> Yep :)
[10:50] <didrocks> Saviq: do you handle it from here, or do you need help?
[10:59] <davmor2> ogra_: WOOHOO!
[11:01] <davmor2> didrocks: 230 has locked up on me twice in the last hour I'm going to monitor it properly now that I'm at work testing :)
[11:01] <didrocks> davmor2: ok ;)
[11:03] <davmor2> morning all
[11:04] <didrocks> morning davmor2 ;)
[11:04] <davmor2> Mirv: I'll give it a go in about an hour I don't know if popey can beat that
[11:04] <davmor2> Mirv: actually does it need to be on mako if not I can install it on manta or flo
[11:05] <mhr3> didrocks, any chance to get a silo for the new-scopes ask? i'm aware that it can't land because of qt5.2, but would be nice being able to "pre-test" it
[11:06] <didrocks> mhr3: unity8 needs a lot of changes, so we can cheat, but I guess for that one, it's better to wait on 5.2 for yourself and now having to retest
[11:06] <didrocks> mhr3: especially as we may pend on an unity8 or unity-mir fix for now due to that crash ^
[11:08] <mhr3> didrocks, well as you can see it's a lot of branches, and the changes are pretty big, so i'd really like to have a ppa with it all asap
[11:09] <didrocks> mhr3: yeah, but anyway, you will have to rebuild and retest after the fact, right?
[11:09] <didrocks> mhr3: so, what that will bring you?
[11:09] <mhr3> indeed
[11:09] <didrocks> (it's the model of the airline, you will be able to work independently. The train has this tradeoff of locking meanwhile)
[11:09] <mhr3> i will know that it at least works now
[11:10] <mhr3> and any issues that might arise will be 5.2 specific
[11:11] <didrocks> mhr3: we have complex transition and regressions as of now, I'll see what we can do once those are solve (if 5.2 didn't land yet)
[11:11] <mhr3> didrocks, very well
[11:11] <didrocks> mhr3: just be aware that your work will be ditched and you will need to retry and retest
[11:11] <didrocks> (if we can do that)
[11:12] <ogra_> sil2100, seems AP has moved its butt ... should i kick an image ?
[11:13] <mhr3> didrocks, but i would also know that we need some extra changes if things blow up now
[11:14] <sil2100> ogra_: indeed! Yes, let's do it ;)
[11:14] <sil2100> ogra_: thanks!
[11:14] <ogra_> [11:14] <didrocks> mhr3: yeah, but when we made the train, there were some tradeoff which were advertized
[11:15] <didrocks> mhr3: and this is one of them, we can't have the airline without having the airline code :/
[11:15] <didrocks> mhr3: we are already doing some complex juggling to avoid blocking everyone, you should be aware of that
[11:15] <didrocks> but at some point, there are some that are way too risky though
[11:16] <Mirv> davmor2: ok. if the small fonts issue in webapps can be seen on flo/manta too, then yes before/after comparison would work there too
[11:16] <mhr3> didrocks, i understand, and if you ultimately say no, i'll shut up and crawl back into my cave :)
[11:17] <didrocks> mhr3: so, I'll just "try", just can't give you 100% assurance
[11:17] <didrocks> mhr3: we first need to be green anyway
[11:17] <mhr3> that's all i want :)
[11:25] <Saviq> didrocks, yeah, we'll look into it
[11:27] <didrocks> Saviq: thanks! keep us posted, we can't keep it longer than today though as it will hide some of the Qt 5.2 other potential crashers
[11:27] <didrocks> Saviq: so, we'll have to revert Mir otherwise at EOD
[11:27] <Saviq> didrocks, k
[11:27] <didrocks> kgunn: FYI (on the rationale) ^
[11:32] <ogra_> didrocks, urgh, have you see the test results ?
[11:32] <ogra_> seems the unity8 crash isnt actually manta specific
[11:33] <didrocks> ogra_: backlog, it's been an hour that we discuss that with Saviq ^
[11:33] <ogra_> oh, ok
[11:33] <didrocks> it's on "stop unity8"
[11:33] <ogra_> sorry
[11:33] <didrocks> (but doesn't happen on rebooting the phone though)
[11:33] <didrocks> no user impact then
[11:34] <didrocks> just gave before EOD for a fix or we'll have to revert Mir (don't want that we hide potential Qt 5.2 bug and crashers due to that)
[11:37] <ogra_> stop unity8 actually hangs here
[11:37]  * ogra_ assumes thats an upstart job issue 
[11:38] <Saviq> ogra_, no
[11:38] <Saviq> ogra_, it's crashing, and apport is collecting the crash
[11:38] <Saviq> ogra_, it's got a 30s kill timeout for that
[11:38] <ogra_> ah, k
[11:43] <Saviq> didrocks, the one in smoke is the d8f8 afaict
[11:44] <ogra_> funny, i dont get a .crash here
[11:44] <ogra_> but i get
[11:44] <ogra_> phablet@ubuntu-phablet:~$ ls /var/crash/
[11:44] <ogra_> _usr_lib_arm-linux-gnueabihf_upstart-app-launch_desktop-hook.32011.crash
[11:44] <ogra_> thats a device that was freshly bootstrapped (with broken touchscreen, so no app was ever started)
[11:45] <ogra_> all i did since flashing is to restart unity8 about ten times via adb
[11:50] <cjwatson> ogra_: that's probably https://bugs.launchpad.net/notes-app/+bug/1267882
[11:51] <cjwatson> I suspect if you look at the .crash file you'll find it's actually a RecoverableProblem not a crash as such
[11:54] <ogra_> yeah
[11:55] <ogra_> you are right, its the same error
[11:57] <Saviq> didrocks, ogra_, fyi: got symbols http://pastebin.ubuntu.com/7073103/ indeed mir/unity-mir somewhere, digging
[12:03] <sil2100> didrocks: we're not in the complete degraded mode right now, right? Maybe we could update the topic? ^
[12:03] <davmor2> didrocks: okay so everything is looking okay bah the lock on 230 which seems to be happening far more frequently that on 226 or 229
[12:10] <ogra_> what about the complaints on the ML
[12:11] <ogra_> i cant really reproduce the issue ...
[12:14] <davmor2> ogra_: which ones in particular
[12:14] <ogra_> the two mails about not being able to start some apps
[12:15] <ogra_> (replies to the landing mail )
[12:23] <ogra_> [12:32] <cyphermox> yo yo yo
[12:44] <sil2100> cyphermox: morning! ;)
[12:44] <sil2100> cyphermox: since didrocks seems to be awayish, could you take a look at a packaging diff and tell me what you think?
[12:44] <sil2100> cyphermox: http://162.213.34.102/job/landing-004-2-publish/31/artifact/packaging_changes_thumbnailer_1.0+14.04.20140307-0ubuntu1.diff
[12:46] <pmcgowan> ogra_, davmor2 in 231 calculator will never launch for me, just white screen
[12:46] <pmcgowan> clock and calendar seem to work ok though
[12:47] <ogra_> pmcgowan, yeah, i saw the mails on the ML ... but i cant reproduce it here
[12:47] <pmcgowan> I did not see updates from the store howver
[12:47] <davmor2> pmcgowan: still testing 230 here
[12:47] <pmcgowan> 230 is so earlier today
[12:48] <davmor2> pmcgowan: and also the version I was asked to test :D
[12:49] <davmor2> pmcgowan: opens fine in 230
[12:50] <pmcgowan> I have calculator 0.1.3.124
[12:50] <cyphermox> sil2100: sure, won't be long
[12:53] <pmcgowan> davmor2, what version of calculator do you have installed?
[12:53] <Mirv> cjwatson: didrocks: arm64 done AFAICS
[12:54] <cyphermox> sil2100: yeah looks fine, but there is no FFE for thumbnailer and you're not closing a bug
[12:54] <davmor2> pmcgowan: com.ubuntu.calculator	0.1.3.224
[12:54] <pmcgowan> qmlscene is complaining it cannot open the main qml file
[12:54] <cjwatson> Mirv: cool
[12:55] <didrocks> Mirv: waow, excellent!
[12:55] <sil2100> cyphermox: it's covered by the standing FFe anyway:
[12:55] <sil2100> cyphermox: https://bugs.launchpad.net/ubuntu/+bug/1208989/comments/17
[12:55] <tvoss> sil2100, tsdgeos is helping out with the process-cpp testing. I think feedback today is realistic
[12:55] <cyphermox> er,, I searched and didn't catch it
[12:55] <sil2100> tvoss: excellent, thanks!
[12:55] <cyphermox> sil2100: fair enough, I see it
[12:56] <cjwatson> Mirv: a bunch of the failures are actually quite easily portable, so I've been sending in branches, but I haven't spotted any blockers yet
[12:56] <seb128> sil2100, is there any issue with u-s-s (wondering why it's not getting a slot)?
[12:56]  * cjwatson is generally a big believer in build-it-everywhere-because-you-might-need-it-one-day, although I know this is unfashionable :)
[12:58] <sil2100> seb128: ah, right! didrocks detached it from the Qt landing already, right?
[12:58] <didrocks> sil2100: yeah, as told this morning :)
[12:58] <Mirv> cjwatson: ok
[12:58] <sil2100> seb128: it's being preparrred, sorry for the delay ;)
[12:59] <didrocks> pmcgowan: calculator opens here on 230, let me switch to 231
[12:59] <seb128> sil2100, right, Mirv&didrocks said it was ok to do an u-s-s landing and detached it from 5.2
[12:59] <seb128> sil2100, thanks
[13:00] <seb128> didrocks, pmcgowan: #ubuntu-app-devel discussed launch issues with some clicks earlier
[13:00] <didrocks> seb128: but on 230 for some
[13:00] <didrocks> and on ML
[13:00] <didrocks> I wonder if it's just the qmlscene crash or something else
[13:00] <pmcgowan> I dont have a crash, qmlscene says it cannot open the qml file
 dpm -> qmlscene: could not exec '/usr/lib/arm-linux-gnueabihf/qt4/bin/qmlscene': No such file or directory
[13:01] <seb128> the qt4 there seemed weird
[13:01] <didrocks> yeah
[13:01] <seb128> but maybe that error was him trying to run it by hand and is something else
[13:01] <pmcgowan> seb128, that means the qt_select is wrong
[13:01] <pmcgowan> could be as the wrong user
[13:01] <didrocks> I see no change on that side though
[13:02] <pmcgowan> didrocks, do you want to attend our qt landing disucssion?
[13:02] <pmcgowan> which is now
[13:02] <didrocks> pmcgowan: calculator opens here
[13:02] <didrocks> pmcgowan: sure, can come
[13:03] <sil2100> seb128: assigned!
[13:03] <seb128> sil2100, thanks
[13:03] <didrocks> pmcgowan: I'm reflashing with bootstrap to see if I can reproduce
[13:07] <kgunn> Saviq: greyback fyi...line 226 of the archive tab in the landing sheet will show you i also landed unity-mir/usc changes seperately...just in case, i split it out from the mir landing...so if we do have to revert
[13:07] <kgunn> i'd revert that landing first
[13:07] <kgunn> hope we don't have to
[13:10] <greyback> me too
[13:12] <didrocks> kgunn: the crash is in the previous image
[13:12] <didrocks> #230
[13:12] <didrocks> kgunn: the one containing the Mir landing
[13:14] <didrocks> pmcgowan: calculator works here on a wiped image (bootstrapped)
[13:14] <didrocks> as well
[13:14] <didrocks> I'm puzzled
[13:15] <didrocks> davmor2: will need your excellent debugging skills here ^
[13:16] <davmor2> didrocks: what for in particular
[13:16] <didrocks> davmor2: see all the discussions about the "I can't start app X or Y"
[13:17] <davmor2> didrocks: and this is all on 231 right
[13:19] <didrocks> davmor2: some are mentionning 230
[13:19] <didrocks> davmor2: see the phone ML, latest couple of messages
[13:23] <didrocks> sil2100: can you reproduce this as well? ^
[13:24] <sil2100> didrocks: let me upgrade to latest of the latest and check
[13:24] <didrocks> thx
[13:24] <didrocks> pmcgowan: can you paste the error you are seeing?
[13:25] <didrocks> really can't open the main qml file? (do you konw which qmlscene binary?)
[13:25] <Mirv> elopio: the only thing you need to be aware of if you plan on running AP tests today is that the landing team is planning to land ubuntu-system-settings. when/if it hits the archives, from that moment on until it has been rebuilt in the landing-006 the dist-upgrade will not work.
[13:25] <pmcgowan> didrocks, qmlscene: failed to check version of file 'ubuntu-calculator-app.qml', could not open...
[13:26] <didrocks> pmcgowan: you don't have any /usr/lib/arm-linux-gnueabihf/qt4/bin/qmlscene on the device, right?
[13:26] <didrocks> oh
[13:26] <didrocks> qmlscene --version
[13:26] <didrocks> qmlscene: could not exec '/usr/lib/arm-linux-gnueabihf/qt4/bin/qmlscene': No such file or directory
[13:27] <cjwatson> didrocks: Despite the test rebuild in progress, I think we should probably just turn on arm64 for all the other ci-train PPAs, completing its architecture roster; it seems to be working fine for landing-006.  Would that be OK to do nowish?
[13:27] <didrocks> seb128: are you on an older image? can you try that?
[13:27] <Mirv> didrocks: if the QT_SELECT=qt5 from /etc/environment is not in use, you need qmlscene -qt5 file.qml
[13:27] <didrocks> cjwatson: I'm fine with that
[13:27] <didrocks> Mirv: ah, so running qmlscene directly expects on the phone to select the Qt4 one, this is a wrong lead?
[13:28] <Mirv> didrocks: might be, I'm not sure if that's something new but maybe related to adb shell not reading /etc/environment? if the user has not selected Qt version, qtchooser falls back to Qt4 for backwards compatibility
[13:28] <didrocks> davmor2: oh good catch!
[13:28] <didrocks> (reading your email)
[13:28] <davmor2> didrocks: replied to the mailing list it sounds like they are updating the app and then trying to open it known bug need to refresh first an then open it. Lets see if that fixes the issue in the meantime I'll see if there are any updates and install them
[13:29] <didrocks> yeah, probably the issue pmcgowan is having as well
[13:29] <didrocks> Mirv: yeah, I think you're right, it's not that regression but probably what davmor2 pointed at :)
[13:29]  * Mirv is not even sure how to run apps from command line nowadays
[13:29] <didrocks> pmcgowan: see https://lists.launchpad.net/ubuntu-phone/msg06854.html
[13:30] <davmor2> Mirv: you don't need to you just need to do a search for the app that refreshes the click scope then it opens
[13:30] <pmcgowan> davmor2, didrocks did not fix my calculator issue
[13:31] <davmor2> pmcgowan: you are on 231 though right?
[13:31] <pmcgowan> davmor2, yes
[13:31] <elopio> Mirv: got it.
[13:31] <pmcgowan> but same version of calculator app
[13:31] <pmcgowan> davmor2, you should no longer need to search to see the updated apps
[13:31] <sil2100> seb128: there seem to be some merge conflicts in the u-s-s branches
[13:31] <davmor2> right okay so it is working here on 230 I'll upgrade to 231 and see what happens
[13:32] <davmor2> pmcgowan: what does it say in /home/phablet/.cache/upstart/com.ubuntu.calculator.log or whatever the exact ending is for the app log
[13:32] <pmcgowan> davmor2, qmlscene: failed to check version of file 'ubuntu-calculator-app.qml', could not open...
[13:33] <seb128> sil2100, indeed, I've having a look, thanks
[13:33] <davmor2> didrocks: ^ sounds like qmlscene hates you and is infecting images left right and centre now because of it ;)
[13:33] <pmcgowan> app startup seems even slower now
[13:35] <davmor2> pmcgowan: I have just updated to 231 and calculator is still opening here, didrocks I'll do a fresh install and see if it opens there but I can't see an issue currently
[13:36] <pmcgowan> davmor2, so I had a wonky update of sorts, I can try to remove and reinstall the app
[13:37] <didrocks> davmor2: same for me
[13:38] <seb128> didrocks, davmor2, pmcgowan: there was an email that just hit the phone list that suggests the launcher config might point to an outdated version of the click
[13:38] <seb128> "If you are can you please try opening search on the apps scope and type
[13:38] <seb128> in the name of the app you want to open then click on it that should
[13:38] <seb128> refresh the scope and point the launcher at the right version."
[13:38] <davmor2> seb128: I wrote it
[13:38] <pmcgowan> seb128, yeah I tried that, no help
[13:39] <pmcgowan> I do not think that is needed any longer btw with latest scope
[13:39] <pmcgowan> as it gets invalidated after installs
[13:39] <seb128> davmor2, oh ok, sorry I misread the quoting in the reply you got :p
[13:39] <davmor2> seb128: however 2 replies (1 pmcgowan and 1 on the list) say it doesn't help
[13:40] <davmor2> 3 now
[13:40] <didrocks> pmcgowan: do you have /usr/share/click/preinstalled/com.ubuntu.calculator/0.1.3.224/ubuntu-calculator-app.qml installed?
[13:40] <pmcgowan> didrocks, I do
[13:40] <pmcgowan> oh wait
[13:40] <pmcgowan> wrong path
[13:40] <didrocks> ah?
[13:40] <didrocks> (this is my fresh image)
[13:40]  * sil2100 reboots his image
[13:40] <sil2100> s/image/device
[13:41] <didrocks> current -> 0.1.3.224 here
[13:41] <davmor2> sil2100: has images on the brain today it seems
[13:41] <cjwatson> It only has to be present in the path reported by "click pkgdir com.ubuntu.calculator"
[13:41] <pmcgowan> didrocks, yes its there
[13:41] <davmor2> right lunch while this is flashing back in 30
[13:42] <didrocks> cjwatson: ah, so the per user is done in /usr/share/click/preinstalled/.click/users/ for preinstalled apps
[13:42] <didrocks> $ click pkgdir com.ubuntu.calculator
[13:42] <didrocks> /usr/share/click/preinstalled/.click/users/@all/com.ubuntu.calculator
[13:42] <didrocks> pmcgowan: for you as well I guess? ^
[13:42] <didrocks> (which points to /usr/share/click/preinstalled/com.ubuntu.calculator/0.1.3.224)
[13:42] <pmcgowan> right
[13:42] <pmcgowan> same here
[13:42]  * didrocks is puzzled
[13:43] <cjwatson> didrocks: there are a bunch of symlinks, the exact details are unimportant for apps :)
[13:43] <sil2100> huh
[13:43] <didrocks> cjwatson: do you have any idea why qmlscene won't find the ubuntu-calculator-app.qml for them, any easy way to debug it's not due to a click update or a broken symlink anywhere?
[13:44] <cjwatson> like I say check "click pkgdir com.ubuntu.calculator" and see if that path exists
[13:44] <pmcgowan> cjwatson, it does
[13:44] <cjwatson> I'd be pretty seriously unimpressed if the launcher has embedded paths that point to a particular version *at all*
[13:44] <sil2100> didrocks: ok, so by accident I flashed 231, but I saw an upgrade of terminal-app in the update manager, upgraded and now I cannot launch terminal-app now
[13:44] <cjwatson> it shouldn't need invalidation
[13:44] <sil2100> didrocks: clicking on the icon in Installed does nothing, phone just vibrating but nothing appearing
[13:45] <cjwatson> didrocks: qmlscene should be finding the .qml file by way of its current directory
[13:45] <pmcgowan> my symptom is a white screen
[13:45] <cjwatson> it might be worth checking the generated .desktop files in ~/.local/share/applications
[13:45] <cjwatson> they should have some key or other that tells it which current directory to use
[13:45] <didrocks> sil2100: tried to "search" trick?
[13:46] <pmcgowan> cjwatson, looks ok to me, there is a path to what click reported
[13:47] <pmcgowan> how do I run the app the same way from the command line
[13:47] <sil2100> didrocks: it works when searching, so confirmed ;/
[13:47] <didrocks> sil2100: ok, so not the issue that pmcgowan is seeing
[13:47] <pmcgowan> so clearly my instance got hosed, probably an outlier
[13:48] <didrocks> tedg might help on the upstart-app-launch side ^
[13:48]  * tedg hides
[13:49] <didrocks> ;)
[13:49] <tedg> The Path key should be the directory of the click install.
[13:49] <didrocks> tedg: multiple persons complain about apps not starting (but not for everyone, not for the same apps), it seems that pmcgowan is seeing "qmlscene: can't find the qml file"
[13:49] <tedg> But we don't use that when launching the application.
[13:49] <cjwatson> tedg: why not?
[13:50] <pmcgowan> sure we do
[13:50] <cjwatson> I'm sure you used to
[13:50] <tedg> cjwatson, We only build the desktop files for other desktops, we don't use them ourselves.
[13:50] <cjwatson> Oh, true, it ultimately winds up as the same thing though
[13:50] <tedg> Yeah, we change directory, but don't read that file.
[13:51] <pmcgowan> tedg, how do I run a click app properly from the command line to mimic what the system does
[13:51] <tedg> pmcgowan, Are there any warnings in the application's log file?
[13:51] <cjwatson> upstart-app-launch com.ubuntu.calculator_calculator_0.1.3.244
[13:51] <tedg> Yes, that.
[13:51] <tedg> But it doesn't give you stdout.
[13:51] <tedg> stdout still goes to ~/.cache/upstart/application-click-com.ubuntu.calculator_calculator_0.1.3.244.log
[13:52] <pmcgowan> tedg, this is what it reports each time http://pastebin.ubuntu.com/7073650/
[13:53] <tedg> pmcgowan, Can you look in /var/log/syslog and see if apparmor is blocking it from being opened?
[13:53] <tedg> There might be a more specific deny there.
[13:54] <pmcgowan> tedg, it is denying
[13:54] <tedg> pmcgowan, Can you pastebin that deny?
[13:55] <pmcgowan> http://pastebin.ubuntu.com/7073664/
[13:56] <pmcgowan> tedg, I dont get anything from that
[13:56] <tedg> pmcgowan, I get that I should blame jdstrand ;-)
[13:56] <cjwatson> It might be worth checking whether /var/log/upstart/click-system-hooks.log exists (and has a recent mtime); failures from click-apparmor ought to wind up there
[13:57] <rsalveti> Mirv: should I test qt 5.2 with the landing 006 or qt5-beta2 ppa?
[13:57] <tedg> pmcgowan, It's probably the apparmor profile that's wrong then.
[13:57] <tedg> pmcgowan, Because it should really have access to that file.
[13:57] <pmcgowan> cjwatson, I do not see that log
[13:57] <cjwatson> OK, hopefully not my fault then :-)
[13:57] <tedg> jdstrand, Is there a way to dump a generated apparmor profile?
[13:57] <pmcgowan> I wonder how it got this way
[13:58] <didrocks> and why only some people on some apps…
[13:59] <Mirv> rsalveti: landing-006 always now (also reads at https://launchpad.net/~canonical-qt5-edgers/+archive/qt5-beta2) - it now also has the qtwebkit update
[13:59] <rsalveti> Mirv: great, thanks
[13:59] <cjwatson> What apps work?
[14:00] <cjwatson> 12:46 <pmcgowan> clock and calendar seem to work ok though
[14:00] <pmcgowan> only calculator seems to be failing, I can try some more....
 dpm, popey: We got a code red. Clock, Calculator and weather are not opening on #image 230. Music, Calendar, Shorts are fine.
[14:00] <cjwatson> calculator has "policy_groups": [], others are non-empty, wonder if that matters
[14:01] <seb128> ^ that was earlier on #ubuntu-app-devel
[14:01] <cjwatson> ... probably not if clock breaks for some people but not others though
[14:01] <cjwatson> it kind of sounds like partial apparmor profile generation or something
[14:01] <didrocks> yeah, sounds the more plausible
[14:01] <didrocks> most*
[14:01] <pmcgowan> right
[14:01] <pmcgowan> how do I see the profile
[14:01] <jdstrand> policy_groups is allowed to be empty
[14:02] <pmcgowan> jdstrand, my calculator is being denied access to the qml file http://pastebin.ubuntu.com/7073664/
[14:03] <jdstrand> the click security manifests are in /var/lib/apparmor/clicks, and the profiles in /var/lib/apparmor/profiles
[14:03] <jdstrand> interesting
[14:03] <cjwatson> http://paste.ubuntu.com/7073695/ <- that's my calculator profile, which works
[14:04] <cjwatson> certainly worth checking that /var/lib/apparmor/clicks/com.ubuntu.calculator_calculator_0.1.3.224.json exists
[14:04] <cjwatson> should be a symlink to /usr/share/click/preinstalled/com.ubuntu.calculator/0.1.3.224/apparmor.json
[14:04] <pmcgowan> cjwatson, it does
[14:05] <didrocks> jdstrand: pat isn't an isolated case, if you read on the phone ML, multiple persons complains about some apps in the last 2 images (seems to be on preinstalled apps however)
[14:05] <cjwatson> so then compare /var/lib/apparmor/profiles/click_com.ubuntu.calculator_calculator_0.1.3.224 against the one I pasted
[14:08] <pmcgowan> jdstrand, http://paste.ubuntu.com/7073709/
[14:08] <jdstrand> I can reproduce it with 'shorts'
[14:09] <jdstrand> 226
[14:09] <davmor2> didrocks, pmcgowan: fresh install and calculator is working fine
[14:09] <didrocks> jdstrand: waow, you are the first on getting it on shorts though, award ;)
[14:10] <pmcgowan> I think this is not related to the recent builds but from last week perhaps
[14:10] <didrocks> yep
[14:10] <jdstrand> huh, /usr/share/click/preinstalled is not in the profile
[14:10] <didrocks> weird that nobody complained beforehand and just today though
[14:11] <didrocks> jdstrand: everything should be a huge failure, no?
[14:11] <jdstrand> not everything, no
[14:11] <jdstrand> just preinstalled apps. if you install from the store, then it would work cause the paths are in /opt/com.ubuntu.click, not /usr/share/click/preinstalled
[14:11] <cjwatson> hmm, that sounds like it has something to do with having installed a user upgrade to the app at some point
[14:12] <didrocks> jdstrand: but the same apps should fail for everyone the same way
[14:12] <jdstrand> it depends
[14:12] <cjwatson> pmcgowan: could you pastebin "find /usr/share/click/preinstalled /opt/click.ubuntu.com -ls" please?
[14:12] <pmcgowan> jdstrand, so thats a big clue
[14:12] <jdstrand> these preinstalled apps are also in the store
[14:12] <pmcgowan> I bet we all upgraded these at some point from the store
[14:13] <jdstrand> apparmor-easyprof-ubuntu has never had /usr/share/click/preinstalled as part of its path
[14:13] <jdstrand> I think that is supposed to be set by CLICK_DIR in the profile
[14:13] <cjwatson> It is, yeah
[14:13] <pmcgowan> jdstrand, http://paste.ubuntu.com/7073736/
[14:13] <jdstrand> and shorts has @{CLICK_DIR}="/opt/click.ubuntu.com"
[14:14] <cjwatson> apparmor/click.py works it out on the fly
[14:14] <jdstrand> hmm
[14:14] <cjwatson> so, that's interesting, this app apparently has the same version in preinstalled and /opt/click.ubuntu.com
[14:15] <cjwatson> however, it has no user symlink in /opt/click.ubuntu.com
[14:15] <davmor2> didrocks, pmcgowan: Camera, calendar, calculator, gallery, browser, contacts, dialer ( rsalveti 's fix for loud speaker works), friends, settings, clock, amazon, facebook, gmail all open for me on mako on 231  I haven't got any further than that though
[14:15] <cjwatson> so the version there should be disregarded - it isn't registered
[14:15] <cjwatson> let me get coffee and then figure out what question to ask next :)
[14:15] <didrocks> davmor2: yeah, seems that there is a good lead here, continue focusing on the rest of dogfooding please :)
[14:15] <didrocks> davmor2: btw, #230 was +1?
[14:16] <davmor2> didrocks: worked for me but didn't look good in the auto tests right?
[14:16] <pmcgowan> coffee sounds good
[14:16] <didrocks> davmor2: #230? apart from the crash that doesn't impact the user experience, what else?
[14:16] <pmcgowan> davmor2, I assume clean installs will be good
[14:17] <jdstrand> cjwatson: right, so I have 0.2.179 of shorts in preinstalled and /opt/click.ubuntu.com
[14:17] <davmor2> pmcgowan: this is a fresh install after testing on an update both worked fine
[14:17] <jdstrand> cjwatson: and that is the highest version
[14:17] <davmor2> pmcgowan: I can install an older version and upgrade if that helps
[14:17] <pmcgowan> davmor2, seems to have to do with having preinstalled apps updated from the store
[14:17] <pmcgowan> davmor2, wait for now
[14:18] <davmor2> pmcgowan: okay
[14:18] <jdstrand> it seems that upstart-app-launch is preferring /usr/share and click-apparmor is preferring /opt/click in this situation
[14:18] <cjwatson> click-apparmor is doing a lot of logic itself that really ought to be done by asking click
[14:18] <tedg> cjwatson, Could this be a libclick issue? ^
[14:18] <cjwatson> it's probably doing disallowed things
[14:19] <cjwatson> tedg: I doubt it, click-apparmor isn't using libclick - in fact it isn't even using click directly
[14:19] <didrocks> cgoldberg: hey, we didn't get feedback from you on autopilot-qt yesterday…
[14:19] <jdstrand> cjwatson: that wouldn't surprise-- it was one of the first hooks and hasn't really been updated
[14:19] <cjwatson> now it's possible it *should* be, but that probably isn't a sensible way to fix a regression
[14:19] <jdstrand> I don't think this is a regression necessarily
[14:20] <tedg> cjwatson, I was more concerned whether your UAL patch to make it use libclick landed, and that could be the issue.
[14:20] <cjwatson> tedg: It hasn't landed
[14:20] <tedg> K
[14:20] <cjwatson> tedg: I was rather assuming you would have to be involved in landing it :-)
[14:20] <jdstrand> it seems this could have happened at any time, at least from click-apparmor's point of view
[14:20] <cjwatson> tedg: I don't think this is a UAL problem anyway
[14:21] <cjwatson> There's an outside chance that it's been exposed by click changes, perhaps
[14:21] <cjwatson>   * user: When setting the registered version of a package to the version in
[14:21] <cjwatson>     an underlay database (e.g. a preinstalled version vs. one in the
[14:21] <cjwatson>     user-installed area), remove the overlay link rather than setting a new
[14:21] <cjwatson>     one equal to the underlay; this was always the intended behaviour but
[14:21] <cjwatson>     didn't work that way due to a typo.
[14:21] <cjwatson> But even if so, I think we need to sort it out in click-apparmor
[14:21] <jdstrand> is enough of libclick in the archive for me to move to it?
[14:22] <cjwatson> pmcgowan: Could you pastebin "find /var/cache/apparmor /var/lib/apparmor -ls" as well for me?  Just need a bit more context
[14:22] <jdstrand> (for this bit)
[14:22] <cjwatson> jdstrand: Not quite, and as I say I suspect a quick fix should be simpler
[14:22] <cjwatson> jdstrand: The bits you probably need are being staged in landing-012 at the moment
[14:23] <jdstrand> the problem seems to be in get_package_manifest()
[14:23] <pmcgowan> cjwatson, http://paste.ubuntu.com/7073779/
[14:23] <cjwatson> 218241    0 lrwxrwxrwx   1 root     root           67 Mar  5 18:55 /var/lib/apparmor/clicks/com.ubuntu.calculator_calculator_0.1.3.224.json -> /opt/click.ubuntu.com/com.ubuntu.calculator/0.1.3.224/apparmor.json
[14:24] <cjwatson> Hm, isn't that directory populated by click?
[14:24] <cjwatson> Pattern: /var/lib/apparmor/clicks/${id}.json
[14:24] <cjwatson> should be
[14:24] <jdstrand> ok, so on my system /var/lib/apparmor/clicks/com.ubuntu.shorts_shorts_0.2.179.json points to /opt/click.ubuntu.com/com.ubuntu.shorts/0.2.179/apparmor.json
[14:25] <jdstrand> the way get_package_manifest() in click apparmor works, is it just walks up the directory until it finds .click
[14:25] <cjwatson> Yeah, I think this is actually a click bug then, somehow
[14:25] <jdstrand> which will be /opt/click.ubuntu.com/.click
[14:25] <jdstrand> so CLICK_DIR is set to that
[14:25] <cjwatson> It's removed the symlink in the topmost database but failed to update the hook symlinks
[14:25] <jdstrand> ah
[14:33] <cjwatson> trying to see if I can forcibly reproduce this
[14:38] <cjwatson> ok, this surely wasn't the original path as such, but this reproduces it:
[14:38] <cjwatson> wget http://people.canonical.com/~ubuntu-archive/click_packages/com.ubuntu.calculator_0.1.3.224_all.click
[14:38] <cjwatson> click install --user=phablet com.ubuntu.calculator_0.1.3.224_all.click
[14:38] <cjwatson> click register --user=phablet com.ubuntu.calculator 0.1.3.224
[14:38] <cjwatson> (all as root)
[14:39] <cjwatson> didrocks,pmcgowan: I have a click landing in progress that among other things lands supported-framework interfaces people have been waiting for; do you think it would be reasonable of me to just fold a fix into that?
[14:39] <cjwatson> I expect to be able to get it tested today
[14:39] <didrocks> cjwatson: +1 on folding the fix
[14:42] <sil2100> Would be awesome
[14:43] <pmcgowan> cjwatson, also +1
[14:43] <pmcgowan> cjwatson, do folks need to manually repair their app installs or will your fix handle that?
[14:43] <cjwatson> I think I'll need to figure something out to do that automatically
[14:44] <pmcgowan> vg
[14:44] <cjwatson> I'll have "click hook run-system" do it, probably
[14:44] <cjwatson> (which already runs at boot)
[14:45] <didrocks> sil2100: can you try on latest image to run the unity8 AP tests?
[14:45] <didrocks> sil2100: then, if you can reproduce the failures, look at the unicode error ones
[14:45] <didrocks> sil2100: try to revert AP and see if that fixes it
[14:46] <cjwatson> ... in fact it might even fall out of the fix automatically
[14:51] <sil2100> didrocks: hm, ok - by latest you mean the one with AP already?
[14:51] <didrocks> sil2100: right
[14:52] <sil2100> Upgrading and checking
[14:52] <didrocks> thanks ;)
[14:55] <cjwatson> jdstrand: Hmm, this is potentially somewhat deep
[14:55] <cjwatson> jdstrand: It seems to me that we have to make sure that the same version never exists in two different databases (preinstalled vs. the user-installed stuff in /opt/click.ubuntu.com) at the same time
[14:56] <sil2100> np :)
[14:56] <cjwatson> jdstrand: Let's say we have foo 1.0 in /usr/share/click/preinstalled/, and user cjwatson installs foo 1.0 in /opt/click.ubuntu.com/ as well
[14:56] <jdstrand> I had kinda figured we would do that, but I never really thought it through or realized the implications if both did
[14:56] <jdstrand> ah, yes
[14:56] <cjwatson> jdstrand: Where does /var/lib/apparmor/clicks/foo_foo_1.0.json point?
[14:57] <jdstrand> we have only one profile
[14:57] <jdstrand> yep
[14:57] <cjwatson> jdstrand: On a multi-user system, cjwatson would want it to point to /opt/click.ubuntu.com/ but jdstrand would want it to point to /usr/share/click/preinstalled/ ...
[14:57] <jdstrand> we could make a choice
[14:57] <cjwatson> I think we must not make a choice
[14:57] <cjwatson> We need to make sure the situation doesn't arise :)
[14:58] <cjwatson> It's invalid and should be disallowed
[14:58] <jdstrand> other than it being wrong from the pov of the user, is there another reason why we couldn't just choose?
[14:58] <cjwatson> Because there's no choice that can be correct for all users
[14:58]  * jdstrand is playing devil's advocate
[14:58] <cjwatson> Either choice will break something
[14:58] <jdstrand> I'm probably not thinking through the preinstalled case enough
[14:59] <jdstrand> (eg, jdstrand is point at preinstalled, then uninstalls, then installs fro the store)
[14:59] <cjwatson> That situation ought to flip jdstrand back to the preinstalled version
[15:00] <cjwatson> We can spot that
[15:00] <jdstrand> otoh, it seems making a choice would be quite brittle with lots of corner cases, so yeah, not allowing it at all seems to make sense
[15:00] <cjwatson> That's actually pretty much exactly the fix I'm proposing
[15:00] <cjwatson> (a) attempts to install the same version that exists in an underlay database shouldn't even unpack the new version, they should simply register the existing one
[15:01] <cjwatson> (b) click hook run-system should sanitise the world, to handle the situation where you upgrade to an image that now has preinstalled versions of something you'd previously installed from the store
[15:01] <jdstrand> 'a' seems very reasonable
[15:01] <cjwatson> I ran into this when I tried to follow through the logic that resulted in the bug in the reproduction case I gave above, and found that the fix would be (IMO) nonsensical
[15:01] <jdstrand> and I was going to ask about 'b'. also reasonable
[15:02] <cjwatson> (I was going to have to run system hooks after registering a different user version, and then wondered what the target ought to be)
[15:03] <jdstrand> what you propose sounds simple and robust. I like it :)
[15:04] <cjwatson> just as well, I don't have an alternative idea that would work :-)
[15:04] <jdstrand> hehe
[15:22] <tvoss> sil2100, didrocks silo 2 is good to go, tsdgeos and me did the testing
[15:22] <sil2100> tvoss: excellent! Thanks, let me publish
[15:22] <sil2100> didrocks: still running the tests
[15:24] <pmcgowan> mandel, you could possibly land ahead of qt5.2, we have been juggling things all week
[15:24] <mandel> pmcgowan, I would really appreciate that, do you know who I have to talk about it?
[15:25] <mandel> pmcgowan, the ideal would be to have two branches from my side and one from barry (although mine are not compulsory)
[15:25] <sil2100> tvoss: love the symbols map addition!
[15:26] <sil2100> cyphermox, didrocks: can anyone of you take a look at the packaging diff for http://162.213.34.102/job/landing-002-2-publish/49/artifact/packaging_changes_process-cpp_0.0.1+14.04.20140311-0ubuntu1.diff ? It's ok, all added dependencies are used in cmake, and there is an enhancement of the symbols
[15:26] <cyphermox> sure thing
[15:26] <sil2100> cyphermox: thanks a lot :)
[15:26] <pmcgowan> mandel, didrocks can help coordinate or see if its possible
[15:27] <didrocks> pmcgowan: mandel: out of context, but sure :)
[15:27] <mandel> pmcgowan, great, I'll follow it with him then
[15:27] <didrocks> depends on the branches, we only do the juggling for rebuilds (which is already quite some work)
[15:27] <mandel> didrocks, we want to create a silo with a branch from system updates and two branches from udm to land in the img to improve a race condition that we had in the manual check for updates
[15:28] <mandel> didrocks, exactly.. or case is not a rebuild, is new code
[15:28] <didrocks> mandel: do you have the new of branches?
[15:28] <sil2100> cyphermox: let me add that it's not in the standing FFe list, but it's only installed on touch - the only rdeps are touch specific packages
[15:28] <mandel> didrocks, if I have a list of the branches?
[15:29] <mandel> didrocks, I do have the ones from udm, I'll need to ask barry to refresh my memory
[15:29] <sil2100> cyphermox: like dbus-cpp and unity-mir
[15:29] <didrocks> mandel: please get everything in the spreadsheet first
[15:29] <cyphermox> sil2100: well, if it's not in the FFE list, it either should be or it needs its own
[15:29] <didrocks> mandel: then, we can see :)
[15:29] <mandel> didrocks, ok, I'll try to find someone to edit that thing :)
[15:30] <didrocks> mandel: great!
[15:30]  * didrocks checks udm meanwhile
[15:30] <sil2100> didrocks: what should I do in the case of process-cpp? It's not in the standing FFe list, dbus-cpp I don't see as well
[15:30] <didrocks> sil2100: get upstream requesting a FFe then
[15:31] <sil2100> didrocks: ok then, these are touch only, so I supposed they were in the FFe for touch, but it seems they're not
[15:31] <sil2100> tvoss: can you submit a FFe for process-cpp?
[15:32] <didrocks> sil2100: are they touch only?
[15:32] <didrocks> sil2100: if they are not on my list, there is a reason
[15:32] <didrocks> sil2100: that's why I always ask to check regarding the list :)
[15:32] <didrocks> sil2100: ok, only seeded in process-cpp, so we can shelve into another one
[15:32] <didrocks> the one for Mir
[15:33] <sil2100> didrocks: the only rdeps are dbus-cpp and unity-mir (with build-depping on platform-api and qtubuntu-sensors)
[15:33] <didrocks> one sec, looking for it
[15:33] <sil2100> All of those seem touch specific ;)
[15:33] <sil2100> Indeed, I just checked now after requesting the packaging ACK and noticed it gune
[15:33] <sil2100> I should have checked before assigning a silo, just this seems so touch to me that I thought it's in a FFe ;)
[15:33] <didrocks> hum
[15:34] <didrocks> process-cpp is in the list
[15:34] <didrocks> https://bugs.launchpad.net/ubuntu/+bug/1282590
[15:34] <sil2100> It is?
[15:34] <sil2100> Come on, I just was searching that!
[15:34] <sil2100> Could I have made a typo :| ?
[15:34] <didrocks> probably
[15:35] <sil2100> No, wait, I opened the wrong bug!
[15:35]  * sil2100 was looking in https://bugs.launchpad.net/ubuntu/+bug/1208989
[15:35] <cyphermox> sil2100: ack from me
[15:35] <sil2100> didrocks: I was looking at the old one ;p
[15:36] <sil2100> Someone should change the status of that one, it's confusing!
[15:36] <sil2100> The name doesn't say any version
[15:36] <sil2100> cyphermox: thanks
[15:36] <cyphermox> I get to wonder why we need to ship copies of gtest in everything when it's in the archive though
[15:37] <cyphermox> tvoss: ^^ ?
[15:37] <tvoss> cyphermox, where do you need the gtest copy?
[15:38] <cyphermox> tvoss: I thought it was in process-cpp?
[15:38] <didrocks> cyphermox: sil2100: added the links to the spreadshed
[15:38] <didrocks> for FFe
[15:38] <tvoss> cyphermox, nope, it pulls in the distro one
[15:38] <cyphermox> tvoss: nevermind, I misread
[15:38] <sil2100> didrocks: thanks, this bug is really confusing - the one you filled in for trusty is better as the name explicitly mentions the release name
[15:41] <didrocks> sil2100: no worry
[15:41] <didrocks> sil2100: any news on unity8?
[15:43] <sil2100> didrocks: ok, I see the tests finished, and I see some <unicode> errors, running again after downgrade
[15:44] <sil2100> didrocks: I see that actually at least 2 of the same tests fail locally here because of the unicode error, let's see autopilot reverted
[15:45] <didrocks> sil2100: ok ;)
[15:46] <didrocks> cgoldberg: did you see my ping?
[15:49] <cgoldberg> didrocks, hi...  i can forward you latest email .. or thomi can update?
[15:49] <thomi> hmmm? Sorry, what's the question?
[15:49] <cjwatson> jdstrand: fixed (a), working through (b) now
[15:49] <didrocks> cgoldberg: I was expecting getting some refresh directly before we land it :)
[15:50] <didrocks> thomi: we reverted both autopilot-qt and autopilot. You now have a new autopilot that we landed without the autopilot-qt revert, that's fine?
[15:50] <jdstrand> nice
[15:50] <cgoldberg> thomi, autopilot-qt status .. the unity8 test suite thing
[15:50] <thomi> didrocks: yes, they're not binary dependant on each other.
[15:51] <didrocks> thomi: cgoldberg: ok, however it seems that for unity8 has 2 new failures, not sure yet it's due to the new Autopilot but it's one of the few that entered
[15:51] <thomi> didrocks: the ap-qt issue is actually a problem in the unity8 test suite, or in Qt/qml, depending on how you look at it
[15:51] <didrocks> thomi: cgoldberg: unicode issue: http://ci.ubuntu.com/smokeng/trusty/touch/mako/231:20140311.1:20140304/7093/unity8/877293/
[15:51] <didrocks> (and another test with exactly the same issue)
[15:51] <didrocks> http://ci.ubuntu.com/smokeng/trusty/touch/mako/231:20140311.1:20140304/7093/unity8/877305/
[15:51] <didrocks> for the second one
[15:52] <thomi> didrocks: have those tests changed in the last image? Cost we didn't get those failures... let me find out job, one second
[15:52] <didrocks> thomi: no, unity8 itself (and so its autopilot test) didn't change
[15:53] <didrocks> thomi: the list of change is sparsed: http://people.canonical.com/~ogra/touch-image-stats/20140311.1.changes
[15:53] <didrocks> sil2100 is trying to downgrade on his phone autopilot just to confirm
[15:53] <thomi> didrocks: so, our test job runs for this release were http://q-jenkins:8080/job/autopilot-release-gatekeeper/67/ and http://q-jenkins:8080/job/autopilot-release-gatekeeper/68/
[15:54] <thomi> so, if this is an AP regression, I'm confused as to why we don't see it in those test results
[15:54] <didrocks> thomi: agreed, that doesn't make sense, you have the unity8 known failure + app crashers
[15:54] <didrocks> nothing else
[15:54] <thomi> yeah
[15:55] <didrocks> let's see sil2100's feedback
[15:55] <thomi> yeah
[15:55] <didrocks> thomi: but see the previous run: http://ci.ubuntu.com/smokeng/trusty/touch/mako/230:20140311:20140304/7082/unity8/
[15:55] <didrocks> only the same known issue
[15:55] <thomi> yeah - odd
[15:56] <thomi> so, that code *has* changed in AP this release, but then... how come our test job didn't pick that up?
[15:56] <thomi> :-/
[15:57] <didrocks> thomi: maybe something to check on the job, if it really installed the latest version?
[15:57] <robru> sil2100, I saw you merged my "disable all cu2d" branch, did you redeploy it too? or should I?
[15:58] <thomi> didrocks: already on it :)
[15:58] <sil2100> robru: ah, didn't redeploy it yet, you can do it if you have a moment :)
[15:58] <robru> sil2100, ok, on it ;-)
[15:59] <sergiusens> sil2100, can i get a silo for l37?
[15:59] <sil2100> sergiusens: let me see that one ;)
[16:01] <sil2100> sergiusens: is goget under the standing FFe? ;D
[16:01] <sergiusens> sil2100, it should be, I did a sync yesterday
[16:02] <sergiusens> sil2100, can't that question be automated ;)
[16:03] <sergiusens> sil2100, hmm, it's not!
[16:03] <didrocks> sergiusens: sil2100: see my coment please
[16:03] <sergiusens> sil2100, phablet-tools isn't either
[16:03] <didrocks> comment
[16:03] <sergiusens> didrocks, press enter or something :-)
[16:04] <didrocks> oh sorry
[16:04] <didrocks> not right line
[16:04] <didrocks> ignore me
[16:04] <sergiusens> didrocks, heh, np
[16:05] <sergiusens> didrocks, we do have an issue though; no tools entered the blanket FFe which sucks
[16:05] <didrocks> sergiusens: I think we should prepare another bug and get that acked quickly
[16:05] <sil2100> sergiusens: yep ;/ I think you need to try adding it to the standing FFe or submit a new one
[16:06] <sergiusens> sil2100, I need to submit a new one
[16:06] <sil2100> As the bug is a wishlist
[16:06] <sergiusens> but I don't feel like that now; I'll do it later
[16:06] <didrocks> sergiusens: prepare one for blanket FFe for all the tools like that
[16:06] <didrocks> phablet-tools and goget, what else?
[16:06] <didrocks> sergiusens: tell me, I'll link the bug in the spreadsheet
[16:07] <sergiusens> didrocks, I don't see qtcreator there either; bzoltan do you have one?
[16:07] <sil2100> didrocks: ok, confiming - reverting autopilot makes the errors go away
[16:07] <sil2100> thomi, cgoldberg: ^
[16:07] <didrocks> sergiusens: yeah, it's not as well, however, I will have a separate one compared to phablet-tools + goget
[16:07] <didrocks> thanks sil2100
[16:07] <sil2100> Now I only get the known random crashers
[16:08] <thomi> didrocks: sorry man - it appears that something is broken in our jenkins job :(
[16:08] <didrocks> thomi: cgoldberg: can you work on one MP for fixing that? I'm happy to not revert but we need to kick an image soon
[16:08] <didrocks> thomi: no worry, better to find it now that in a more critical situation :)
[16:08] <thomi> didrocks: sure. I'll do that now
[16:08] <sil2100> thomi: thanks :)
[16:08] <didrocks> thanks
[16:14] <didrocks> cyphermox: robru: sil2100: so update: we try to refrain getting things in until we have 3 fixes in + an image kick
[16:14] <robru> didrocks, ok, what fixes? are they in a silo?
[16:14] <sil2100> didrocks: ACK!
[16:14] <didrocks> 1. a click fix (l34) where cjwatson is including it to it
[16:15] <didrocks> 2. unity-mir (l35)
[16:15] <didrocks> 3. autopilot (thomi just started to work on it)
[16:15] <didrocks> click fix is going to fix applications not starting for some people
[16:15] <didrocks> unity-mir is going to fix all the unity8 crash that we see on the dashboard
[16:15] <didrocks> autopilot is going to fix 2 new AP tests failures on unity8 (the unicode)
[16:16] <didrocks> then we kick an image build (ogra_, cyphermox) ^
[16:17] <ogra_> just ping :)
[16:17] <didrocks> ogra_: I'll probably be away, so I won't, but I think robru and cyphermox are going to track :)
[16:18]  * ogra_ will definitely attend the last twi uds sessions ... s i'll be around for a while
[16:18] <ogra_> *two
[16:19]  * sil2100 needs to jump out for a while
[16:19] <robru> i'll be around
[16:26] <cyphermox> I'm around
[16:30] <didrocks> thomi: ok, same source of error for ubuntu-system-settings I guess?
[16:30] <didrocks> http://ci.ubuntu.com/smokeng/trusty/touch/mako/231:20140311.1:20140304/7093/ubuntu_system_settings/879028/
[16:30] <didrocks> http://ci.ubuntu.com/smokeng/trusty/touch/mako/231:20140311.1:20140304/7093/ubuntu_system_settings/879055/
[16:30] <didrocks> http://ci.ubuntu.com/smokeng/trusty/touch/mako/231:20140311.1:20140304/7093/ubuntu_system_settings/879060/
[16:31] <thomi> didrocks: looks like it. I'm in the middle of the fix now. BTW, we know what happened as well - some of the provisioning scripts we use were broken, apparently. doanac` knows more
[16:37] <didrocks> thomi: ah great! :)
[16:46] <sergiusens> didrocks, does that sound good? https://bugs.launchpad.net/ubuntu/+bug/1290944
[16:48] <didrocks> sergiusens: I would add "tool" in the title
[16:48] <didrocks> sergiusens: you can directly target the 2 packages btw in the bug
[16:48] <didrocks> (the ubuntu packages)
[16:49] <didrocks> and then, just subscribe the release team
[16:49] <didrocks> maybe a ping on #ubuntu-release can help
[16:49] <didrocks> sergiusens: just emphasize that they have no impact on the desktop or any flavor
[16:49] <sergiusens> didrocks, heh, I thought I added tools there
[16:54] <sergiusens> didrocks, ok, I updated and now subscribing the release team
[16:55] <didrocks> sergiusens: I renamed the title to include "tools"
[16:55] <didrocks> sergiusens: I don't see the release team subscribed?
[16:57] <sergiusens> didrocks, I said I was doing that now :); not done ;-)
[16:57] <didrocks> ah :p
[16:57] <sergiusens> didrocks, now it's done :-)
[16:58] <didrocks> sergiusens: perfect! keep us posted :)
[16:58] <sergiusens> didrocks, your subscribed for fun mail :-)
[16:58] <didrocks> sergiusens: sure sure, I never have enough emails :p
[17:00] <tvoss> sil2100, didrocks anything I need to do for process-cpp now?
[17:00] <didrocks> sil2100: I think it was going to be published, right? ^
[17:06] <didrocks> robru: it should start automatically in the right order, if not, you should get a case and ask on the webops channel to investigate
[17:07] <robru> didrocks, for what, the depwait issue?
[17:07] <didrocks> robru: yep
[17:07] <robru> ok
[17:08] <bregma> hey guys, could I get a reconfigure in silo landing-008 please?
[17:08]  * Mirv runs merge and clean for ubuntu-system-settings in order to get that back to Qt 5.2
[17:08] <robru> bregma, on it
[17:08] <seb128> Mirv, I was about to do it
[17:09] <Mirv> seb128: ah, ok, feel free to do it then
[17:09] <seb128> Mirv, should I not press the button? ;-)
[17:09] <Mirv> seb128: just press it! :)
[17:09] <seb128> Mirv, done
[17:09] <Mirv> thanks
[17:09] <seb128> yw
[17:14] <robru> bregma, humm, indicator-sound is already in silo 7...
[17:15] <ogra_> didrocks, planning a busy nightshift ?
[17:15] <robru> thostr_1, what's happening in silo 7? can I bump indicator-sound out of there for bregma?
[17:15] <didrocks> ogra_: not that much for me, still have one vUDS session to host
[17:15] <ogra_> didrocks, ... "and kick #332 with all desired fixes." ...
[17:15] <ogra_> thats 111 images !
[17:15] <bregma> oops, I missed a check
[17:15] <didrocks> ogra_: we can do it
[17:15] <didrocks> and I can fix it :p
[17:15] <ogra_> lol
[17:16] <ogra_> .oO ( why do i always only recognize the typos *after* sharing it)
[17:16] <didrocks> ogra_: thanks, fixed
[17:16] <ogra_> :)
[17:16] <didrocks> ogra_: it seems I have a problem with numbers :p
[17:17] <seb128> bregma, thostr_1, robru: silo 7 has packages not building, it doesn't seem ready to land (and most of the mps didn't get reviewed/approved yet)
[17:17] <robru> seb128, yes, i just noticed the most recent buld was 12+ hours ago, issues unresolved. ok, i'm turfing that silo for bregma
[17:17] <ogra_> didrocks, only your fingers ...
[17:17] <seb128> imho drop it so we can land the unity lock screen
[17:17] <seb128> robru, thanks
[17:17] <ogra_> didrocks, you should start practicing to write the G+ posts with your toes ... proabably that helps ;)
[17:18] <didrocks> ogra_: assign me a WI for that :p
[17:18] <ogra_> lol
[17:22] <robru> seb128, bregma: humm, silo 7 is taking a long time to clean... can't reconfig silo 8 until it's done.
[17:22] <seb128> robru, those usually takes 5 to 10 mins, wait a bit
[17:22] <bregma> robru, I am a patient man
[17:22] <seb128> I'm not :p
[17:23] <cjwatson> building click fixes (hopefully) in landing-012 now
[17:23] <seb128> bregma, that's fine, but the distro ffe you got granted is about to expire
[17:23] <cjwatson> but I'll test them myself before inflicting them on anyone else
[17:25] <tvoss> sil2100, didrocks can I clean the silo?
[17:25] <tvoss> oh, someone did according to status
[17:26] <Mirv> I restored ubuntu-system-settings to landing-006's config while keeping the unity-mir still away
[17:26] <Mirv> and rebuilding the new u-s-s trunk
[17:27] <Mirv> for the magic route there's 'config.backup' with original config in the silo directory, which can be copied over config after unity-mir has landed. or then just the usual reconfigure route.
[17:31] <sergiusens> didrocks, Mirv why is the Qt5.2 message green now? ar we finally all unblocked?
[17:31] <sergiusens> :-)
[17:32] <Mirv> sergiusens: :) well the landing is ready, bugs have been squashed, but a small handful of apps autopilot results need more tinkering still
[17:33] <Mirv> sergiusens: elopio's latest run's results are at http://q-jenkins:8080/job/autopilot-release-gatekeeper/69/#showFailuresLink - gallery-app doesn't apply since it needs a recompilation and AP tests run after that. notes app being resolved.
[17:33] <didrocks> sergiusens: green?
[17:33]  * didrocks looks
[17:33] <Mirv> didrocks: I guess he means the top text
[17:33] <didrocks> ah ok :)
[17:33] <Mirv> green is a promising color
[17:33] <didrocks> sergiusens: yeah, we have "normal landings"
[17:36] <Mirv> hehe, "Some packages (unity-mir) that are not in the MP and not in sources list have been detected in the ppa." I know! the main thing is that ubuntu-system-settings is building now, so PPA is back to functional after that (until unity-mir lands)
[17:41] <sergiusens> yeah, was the top text
[17:41] <bregma> robru, silo 7 looks squeaky clean now, howsabout silo 8?
[17:41] <sergiusens> Mirv, gallery and notes are click fwiw
[17:41] <sergiusens> Mirv, this might be a mess if it doesn't go from devel-proposed to devel in one swipe
[17:42] <didrocks> Mirv: checking my check system? :p
[17:42] <robru> bregma, ah yes, sorry. got distracted watching UDS vids.
[17:42] <sergiusens> as well as all the other clicks
[17:42] <Mirv> sergiusens: yes, that's why it's now not possible to run gallery-app AP tests with the Qt PPA anymore, since obviously the package isn't installed (click one is)
[17:42] <Mirv> didrocks: works, still!
[17:42] <didrocks> ;)
[17:43] <didrocks> thomi: you will ping robru for the autopilot landing? (are you close to land it?)
[17:43] <Mirv> sergiusens: notes should work without recompilation so that's why the tests for it were not all failing
[17:43] <didrocks> cjwatson: same for click I guess? (just ping robru for publication or ensure stéphane is around once the fix is tested)
[17:43] <sergiusens> Mirv, the link you gave me has all notes tests failing though
[17:44] <robru> yep, i'm here all day, just ping me whenever something needs to land
[17:44] <sergiusens> AssertionError: After 10.0 seconds test on NoteItem.height failed: 198.0 != dbus.Double(432.0, variant_level=1)
[17:44] <didrocks> kgunn: so, no more crash with silo 005 on unity8 stop?
[17:44] <robru> bregma, ok, recon complete, please build
[17:45] <Mirv> sergiusens: I count 7 failures, doesn't notes have like 27 of them? anyhow, elopio is up-to-date on the notes-app Qt 5.2 AP situation
[17:45] <didrocks> robru: I published silo5 FYI
[17:45] <robru> didrocks, ok thanks
[17:45] <sergiusens> Mirv, don't know; just assumed 'rebuilding' applied to all; no worries if it's on track
[17:46] <kgunn> didrocks: tested via, stop unity8/start unity8...no problems
[17:46] <didrocks> kgunn: excellent! One less, just 2 issues on the line before getting good again :)
[17:46] <didrocks> thanks
[17:46] <kgunn> i checked /var/crash/ before and after
[17:47] <didrocks> kgunn: I'll then even open the HO for your Mir session :p
[17:50] <cjwatson> didrocks: I learned how to do it myself the other day, so that'll be fine if need be
[17:51] <didrocks> cjwatson: thanks a lot, you don't need to wait for merge and clean before tomorrow if it's too late though
[17:52] <cjwatson> I'm testing it now, so hopefully won't be too long (assuming nothing goes wrong)
[17:53] <didrocks> great :)
[17:53] <mhr3> didrocks, so, will i get silo tomorrow? :)
[17:53] <didrocks> thomi: once you are done with an early breakfast, keep us posted please :)
[17:53] <didrocks> mhr3: yeah, do you have one line with everything?
[17:53] <didrocks> mhr3: as it's new deploy code, I would prefer push it tomorrow morning in case I've done a typo
[17:53] <thomi> didrocks: just pushed a branch, making an MP now
[17:54] <didrocks> thomi: \o/
[17:54] <mhr3> didrocks, minus unity8, but will get that from ppa
[17:54] <thomi> tested it with unity8 failing tests. tests passnow
[17:54] <elopio> sergiusens, Mirv: oh, actually, I'm ready for a review here: https://code.launchpad.net/~elopio/notes-app/dont_click_header/+merge/210349
[17:54] <elopio> Quick hack to remove most of the failures.
[17:54] <didrocks> thomi: excellent, so I assume it's the same for ubuntu-system-settings
[17:54] <elopio> who was working on notes? Now I don't remember.
[17:55] <didrocks> elopio: bfiller_afk's team
[17:56] <elopio> nerochiaro was the name I was looking for.
[18:08] <thomi> didrocks: mp is up: https://code.launchpad.net/~thomir/autopilot/trunk-fix-uncide-error/+merge/210464  have asked veebers to review
[18:08] <thomi> yurns out I can't spell branch names though :-/
[18:08] <thomi> *turns
[18:08] <didrocks> thomi: ahah, no worry ;)
[18:08] <didrocks> thomi: mind setting a line in the spreadsheet?
[18:08] <didrocks> then, I'll assign the silo for now
[18:08] <thomi> sure
[18:15] <thomi> didrocks: line 39. Not had a review yet though - veebers is in a UDS session
[18:15] <didrocks> thomi: landing 001 for you, you can start building still
[18:15] <didrocks> thomi: please set "ready" to yes ;)
[18:16] <didrocks> thomi: then, just put "tested" to yes once veebers ack it
[18:16] <didrocks> and robru will be your host for this landing (and then we'll have a new shiny image \o/)
[18:16] <thomi> didrocks: ack
[18:16] <didrocks> thanks again for the quick fix!
[18:24] <thomi> no worries
[18:25] <thomi> sorry for the regression - we'll have something in place to make sure that doanac` doesn't break us again :P
[18:25] <doanac`> thomi: its hard to put a plan in place that can withstand my incompetence :)
[18:26] <thomi> hah
[18:26] <didrocks> thomi: well, no worry, as long as we didn't really miss a step, all is fine (and we had 3 issues to fix today and could have avoid a revert, so all's fine!) :)
[18:27] <didrocks> double watching doanac` now :p
[18:29] <cjwatson> click 0.4.18.1 looks fine and fixes the problem above (at least as far as I can tell) - stgraber is releasing it
[18:36] <didrocks> perfect!
[18:38] <robru> man, you guys don't even need me ;-)
[18:40] <didrocks> robru: you'll do the AP one, and ask for kicking the image :)
[18:40] <didrocks> robru: I'm EOD now
[18:41] <robru> didrocks, goodnight!
[18:42] <didrocks> thanks, you too
[18:42] <seb128> cyphermox, you can do it!
[18:42] <cyphermox> seb128: oy
[18:42] <seb128> cyphermox, hey ;-)
[18:42] <seb128> cyphermox, I see your selection color over the theme landing ask for a bit ;-)
[18:42] <cyphermox> yes
[18:43] <seb128> cyphermox, googledoc spying on you ;-)
[18:43] <cyphermox> I was trying to test it
[18:43] <seb128> oh ok
[18:43] <seb128> I see you trust your teammates :p
[18:43] <cyphermox> I do
[18:43] <cyphermox> but I also want to see it for myself ;)
[18:43] <seb128> give a silo, test the ppa!
[18:43] <Mirv> rebuilding unity-mir to Qt 5.2.
[18:43] <seb128> ;-)
[18:44] <cyphermox> seb128: good point, I'm an idiot
[18:44] <robru> Mirv, great!
[18:44] <seb128> cyphermox, getting used to the futur can take some time ;-)
[18:46] <cyphermox> there
[18:48] <seb128> cyphermox, thanks
[18:50] <cyphermox> np
[18:50] <seb128> cyphermox, do you know what's going on silo 8?
[18:50] <seb128> cyphermox, the status is complaining about gnome-screensaver not being listed in the sources when it is
[18:51] <cyphermox> well, it was probably initially configured without
[18:51] <cyphermox> let me fix that
[18:51] <cyphermox> you *do* want a fixed gnome-screensaver to go with this?
[18:54] <robru> seb128, cyphermox: oh, I must have goofed that when I reconfigured it recently. sorry
[18:55] <cyphermox> robru: no worries
[18:55] <cyphermox> just triple-checking to not mess it further
[18:55] <seb128> cyphermox, yes, g-s is in the ppa already, it's part of that landing
[18:55] <robru> cyphermox, yeah, do you want me to reconfigure it again or are you doing it?
[18:55] <cyphermox> robru: about to do it
[18:55] <robru> ok
[18:55] <cyphermox> WTOOMANYTABS
[18:56] <cyphermox> I can hardly find jenkins between the ci train spreadsheet, tab for uds and all :)
[18:56] <robru> haha, yeah
[18:57] <sergiusens> cyphermox, don't use tabs ;-)
[19:00] <cyphermox> sergiusens: how would you do it otherwise?
[19:00] <cyphermox> ;)
[19:02] <sergiusens> cyphermox, windows with your window manager :-)
[19:02] <cyphermox> sergiusens: sometimes that's worse ;)
[19:02] <sergiusens> cyphermox, makes you also close the useless stuff :-)
[19:02] <cyphermox> ahah
[19:02] <sergiusens> cyphermox, fwiw I'm just messing with you ;-)
[19:02] <cyphermox> I know
[19:02] <cyphermox> still worth the questioning though
[19:03] <cyphermox> but my process usually works, just painful sometimes when there are lots going on at once
[19:09] <bschaefer_>  /nick bschaefer
[19:16] <davmor2> cyphermox: bluetooth is there any progress on this at all?
[19:18] <Mirv> sil2100: robru: new unity-mir fails to build against new process-cpp, so unity-mir failed to build (now a bit later than the previous unity-mir build to the archives) in Qt 5.2 PPA https://launchpadlibrarian.net/169119692/buildlog_ubuntu-trusty-amd64.unity-mir_0.3%2B14.04.20140311.3-0ubuntu1_FAILEDTOBUILD.txt.gz
[19:18] <rsalveti> sil2100: process-cpp is now causing another ftbfs
[19:18] <rsalveti> yey
[19:18] <rsalveti> the same issue
[19:18] <rsalveti> https://launchpadlibrarian.net/169072758/process-cpp_0.0.1%2B14.04.20140122-0ubuntu1_0.0.1%2B14.04.20140311-0ubuntu1.diff.gz
[19:18] <rsalveti> +#include <core/signal.h>
[19:18] <rsalveti> +
[19:18] <rsalveti> this is the problem
[19:18] <robru> Mirv, huh?
[19:18] <rsalveti> should probably be #include <core/posix/signal.h
[19:19] <rsalveti> Mirv: I'm currently trying to rebuild unity-mir and had the same issue
[19:19] <rsalveti> rebuilding the one in the archive should already be enough
[19:19] <rsalveti> latest process-cpp brought a regression
[19:21] <cyphermox> davmor2: rsalveti: have you had time to look at my email re: bluetooth yet?
[19:21] <Mirv> yeah, I need to go to sleep now but just FYI
[19:21] <rsalveti> cyphermox: sorry, not yet, next on my list
[19:21] <rsalveti> had to work on webkit
[19:21] <cyphermox> sure, np
[19:22] <rsalveti> server unavailable, wtf
[19:23] <rsalveti> http://bazaar.launchpad.net/~phablet-team/process-cpp/trunk/revision/36
[19:23] <rsalveti> tvoss: hey, unity-mir is now failing to build with latest process-cpp
[19:23] <davmor2> cyphermox: there's and email
[19:23] <rsalveti> https://launchpadlibrarian.net/169119692/buildlog_ubuntu-trusty-amd64.unity-mir_0.3%2B14.04.20140311.3-0ubuntu1_FAILEDTOBUILD.txt.gz
[19:23] <cjwatson> rsalveti: bazaar> being looked at in #webops
[19:24] <davmor2> cyphermox: where
[19:25] <Wellark> hmm.. I need someone to help me to get a project to universe
[19:25] <cyphermox> davmor2: that was directed to rsalveti
[19:25] <Wellark> new (unity8) indicator dependencies
[19:25] <rsalveti> cjwatson: thanks
[19:26] <davmor2> cyphermox: ah okay it threw me when my name was there too phew I thought I'd missed an email for a minute :)
[19:27] <Wellark> who should I bug? the package needs packaging review. it should be covered by the unity8 general FFe AFAIK
[19:29] <Wellark> I'm not sure how introducing new source packages relates to the ci train
[19:32] <Wellark> sil2100, robru: ^
[19:32] <Wellark> maybe cyphermox ^^
[19:32] <robru> uh
[19:32] <davmor2> kgunn: I'm still getting locks on 0.1.6 similar kind of issue as on 0.1.5, ie they are random and seem to happen more frequently on closing apps than anything else.
[19:32] <cyphermox> Wellark: link the pacakge you want to review?
[19:33] <cjwatson> robru:  click | 0.4.18.1 | trusty | source, amd64, arm64, armhf, i386, powerpc, ppc64el
[19:33] <robru> Wellark, typically very little config is required technically to set up a project in citrain. but we do need an officially designated lander, and a test plan so that we know what the acceptance criteria are for that package.
[19:33] <cjwatson> that should unblock that item
[19:33] <Wellark> robru: yeah, I'm just wondering what's the marching order here
[19:33] <robru> Wellark, well what's the package?
[19:33] <Wellark> I have the project and jenkins ci set up for it already
[19:33] <robru> cjwatson, huh?
[19:34] <davmor2> kgunn: 1 on manta on 231 and one on flo on 231
[19:34] <Wellark> robru, cyphermox: lp:connectivity-api
[19:34] <davmor2> kgunn: not had one on mako as of yet
[19:34] <cjwatson> robru: that's one of didrocks' blockers for building the next image, which I understood you to be driving
[19:34] <robru> Wellark, and so what, that's a component of unity8?
[19:34] <robru> cjwatson, oh, you're informing me that click is in distro now? thanks.
[19:35] <cjwatson> robru: right
[19:35] <Wellark> robru: it's a prerequisite of indicator-network
[19:35] <Wellark> and upcoming networking status QML api
[19:36] <Wellark> robru: thostr is already in set as Lander
[19:36] <Wellark> in CITrain rampup, owners, bootcamps document[C
[19:37] <rsalveti> robru: need a silo for 40, to fix the ftbfs regression introduced by process-cpp (same issue as reported by Mirv)
[19:37] <robru> Wellark, yeah I just saw that. ok looks fine. as far as I can tell it's fully ready for citrain. if you have an MP, get thostr to approve it and request a landing for it
[19:37] <robru> rsalveti, ok
[19:37] <Wellark> robru: no additional setup needed for getting the packages to universe?
[19:37] <robru> rsalveti, except there's no MP there
[19:37] <rsalveti> oh, sorry
[19:38] <robru> Wellark, well, you just have to *do a landing* to get the package in universe.
[19:38] <rsalveti> robru: there's now :-)
[19:38] <rsalveti> control+v didn't work
[19:38] <Wellark> robru: ok, if a MP through the train is all it takes then I'm happy
[19:38] <cyphermox> robru: did we have an image built before landing click?
[19:39] <rsalveti> ricmm: robru: mind doing a quick review as well? I believe it needs to be top approved as well: https://code.launchpad.net/~rsalveti/process-cpp/adding-missing-dev-dependency/+merge/210482
[19:39] <robru> cyphermox, no, i was told to get click in before kicking an image
[19:39] <Wellark> just need somebody to review the packaging. cyphermox: you got the time to do it?
[19:39] <robru> rsalveti, ok you got silo 5
[19:39] <rsalveti> robru: thanks
[19:39] <robru> rsalveti, i means ilo 4 oops
[19:39] <cyphermox> I guess yes then
[19:39] <robru> rsalveti, can't review that MP, i don't have a clue about that project
[19:40] <rsalveti> robru: np, let me ping sergiusens then
[19:40] <rsalveti> not sure if ricmm is around
[19:40] <sergiusens> rsalveti, what?
[19:40] <sergiusens> :)
[19:40] <robru> rsalveti, stylistically I approve. but I have no idea if that dep is necessary or correct or not ;-)
[19:40] <rsalveti> sergiusens: quick one https://code.launchpad.net/~rsalveti/process-cpp/adding-missing-dev-dependency/+merge/210482
[19:40] <rsalveti> robru: rev36 introduced a header dependency: https://launchpadlibrarian.net/169072758/process-cpp_0.0.1%2B14.04.20140122-0ubuntu1_0.0.1%2B14.04.20140311-0ubuntu1.diff.gz
[19:41] <rsalveti> +++ process-cpp-0.0.1+14.04.20140311/include/core/posix/child_process.h	2014-03-11 11:02:54.000000000 +0000
[19:41] <rsalveti> +#include <core/signal.h>
[19:41] <rsalveti> which is part of libproperties-cpp-dev
[19:41] <robru> rsalveti, that diff already has libproperties-cpp-dev in the debian/control diff?
[19:42] <rsalveti> robru: right, but that is the build-dep of that package itself
[19:42] <cyphermox> robru: so I guess we're just wiating for autopilot now to kick an image
[19:42] <seb128> cyphermox, how is the theme update looking to you?
[19:42] <robru> rsalveti, oh, I see. ok, thanks for explaining. I guess I know enough to approve now ;-)
[19:42] <robru> cyphermox, yeah I think so
[19:42] <rsalveti> np
[19:43] <robru> rsalveti, do you want to add that to the silo 4 I just gave you? are those related?
[19:43] <Wellark> cyphermox: we could use this as the first MP through ci train https://code.launchpad.net/~unity-api-team/connectivity-api/ci-testrun/+merge/210479
[19:43] <sergiusens> rsalveti, seems legit
[19:43] <Wellark> and I could do the necessary packaging changes through that branch at the same time
[19:43] <rsalveti> sergiusens: :-)
[19:44] <thomi> cyphermox: robru: wait no longer!
[19:44] <rsalveti> robru: sorry, add what exactly? that's basically silo 4
[19:44] <robru> thomi, excellent
[19:44] <rsalveti> building atm
[19:44] <davmor2> sergiusens: did something change recently with google sync?   http://sergiusens.github.io/posts/syncing-contacts-from-google-on-latest-touch-images.html  I followed this and it isn't working  load of Gkr: secret service operation failed: The name org.freedesktop.secrets was not provided by any .service files
[19:44] <cyphermox> thomi: ok, so ready to publish your stuff then?
[19:44] <cyphermox> we'd still need to wait a bit for it to make it into release
[19:44] <robru> rsalveti, oh sorry, misread it. i thought those were two different MPs
[19:44] <rsalveti> robru: right, np
[19:44] <sergiusens> davmor2, yeah, look at the emails and rsalveti askubuntu
[19:44] <cyphermox> Wellark: what are you talking about?>
[19:45] <rsalveti> davmor2: http://askubuntu.com/questions/360554/how-do-i-sync-google-contacts
[19:45] <robru> cyphermox, thomi: published
[19:45] <davmor2> sergiusens: ah right will do thanks first time I've tried it in a while :)
[19:45] <Wellark> cyphermox: that packaging review. I thought you volunteered ;)
[19:46] <robru> cyphermox, i'll do the packaging review.
[19:46] <robru> Wellark, ^
[19:46] <Wellark> robru: sweet.
[19:46] <Wellark> let's take this off the channel
[19:46] <cyphermox> Wellark: there is no packaging review from this if it's not in citrain?!?
[19:46] <sergiusens> rsalveti, I approved, but do I need to build?
[19:46] <robru> cyphermox, it needs a NEW review because it'll be the first upload to universe
[19:46] <rsalveti> sergiusens: already building
[19:46] <sergiusens> rsalveti, just did a scavenge review
[19:46] <cyphermox> ah
[19:47] <rsalveti> sergiusens: just need a top approval
[19:47] <robru> cyphermox, also it is in citrain. just first release
[19:47] <cyphermox> but where is it in citrain?
[19:47] <sergiusens> rsalveti, ci train doesn't care about that
[19:47] <sergiusens> rsalveti, but I'll do it
[19:47] <robru> cyphermox, well, it doesn't have a landing assigned yet. but it's in that *other* spreadsheet
[19:47] <cyphermox> in the rampup one?
[19:47] <rsalveti> sergiusens: oh, ok
[19:47] <robru> cyphermox, yeah
[19:47] <rsalveti> sergiusens: thought we'd need someone to top approve it first
[19:47] <sergiusens> rsalveti, discovered by accident myself :-)
[19:48] <cyphermox> robru: then you should be able to just assign a silo and we'll do the review
[19:48] <thomi> cyphermox: robru: awesome. Thanks
[19:49] <robru> cyphermox, yeah, I'm just going to do like a pre-pre-NEW review before even assigning the silo. then once I'm satisfied we'll build it in a silo and then you can do a more formal review
[19:49] <cyphermox> robru: so that then someone else can do another one? ;)
[19:49] <cyphermox> hehe
[19:49] <rsalveti> sergiusens: interesting :-)
[19:50] <davmor2> sergiusens, rsalveti: \o/ that worked thanks guys :)
[19:51] <rsalveti> davmor2: great
[19:54] <cyphermox> seb128: I haven't forgotten you, just need to logout to test now I guess
[19:56] <seb128> cyphermox, you can just switch theme to something else and back
[19:56] <cyphermox> riugfht, I could
[19:56] <cyphermox> fwiw, that was supposed to be "right"
[19:57] <cyphermox> hawt
[19:57] <seb128> cyphermox, that's why you shouldn't speak with food in your mouth...
[19:57] <cyphermox> that's why you shouldn't type with food in your keyboard too ;)
[19:59] <robru> *whew* vuds finally over for today, now I can get some real work done ;-)
[19:59] <cyphermox> seb128: looks very nice
[20:00] <cyphermox> I was kind of worried about usability to resize windows with the 1px border removed
[20:02] <seb128> cyphermox, yeah, resizing is fine, it looks a bit weird though
[20:02] <seb128> let's land and see the feedback
[20:02] <rsalveti> cjwatson: getting chroot problem when building process-cpp on ppc64el: https://launchpad.net/~ci-train-ppa-service/+archive/landing-004/+build/5802999
[20:02] <rsalveti> cjwatson: should I just retry or is there anything useful in there for debugging?
[20:02] <rsalveti> or if it's also a known issue
[20:03] <cjwatson> rsalveti: what, again?  we had that the other day but I thought that was resolved
[20:03] <cjwatson> rsalveti: let me see if I can summon infinity
[20:03] <rsalveti> sure
[20:05] <cyphermox> ah, I was about to mention that
[20:09] <cjwatson> I'll bulk-retry once infinity's happy for me to do so
[20:17] <cjwatson> rsalveti,cyphermox: OK, affected builders on manual, failures bulk-retried, infinity's looking into either a band-aid or a proper fix depending
[20:17] <cyphermox> ok
[20:17] <cjwatson> thanks for the heads-up
[20:18] <cyphermox> ah, we still don't have ust on arm64?
[20:18] <cjwatson> Nope
[20:18] <cjwatson> I sent a branch for UAL this morning to avoid caring
[20:19] <cjwatson> https://code.launchpad.net/~cjwatson/upstart-app-launch/porting/+merge/210385
[20:19] <cjwatson> So if you have something else similar then you could use the same idea
[20:21] <rsalveti> cjwatson: thanks
[20:24] <cyphermox> cjwatson: it just doesn't build there?
[20:24]  * cyphermox fires up schroot
[20:38] <bregma> robru, cyphermox, line 22/silo 008 is ready for a publish, even though it has a failed-to-build error (on ppc64el) -- is this a not unexpected condition?
[20:39] <robru> bregma, failing on ppc64el is only ok if you always fail on ppc64el... regressions are not allowed
[20:39] <bregma> robru, yes, seb128 said it's not a regression
[20:39] <robru> bregma, ok perfect
[20:40] <robru> cyphermox, oh, did you kick an image yet? we need that before landing anything else
[20:40] <robru> cyphermox, oh, AP is still in -proposed
[20:40] <robru> bregma, ok so we're still waiting for autopilot to land, then we need to kick an image, then we can publish your thing
[20:41] <bregma> robru, as long as it lands before the FFe expires, I'm good with that
[20:41] <robru> bregma, when does that expire? i'm thinking it should land within hours?
[20:41] <cyphermox> robru: correct
[20:41] <seb128> bregma, seems like some of thoses are real issues...
[20:41] <bregma> "tonight" for a given definition of "tonight"
[20:42] <seb128> bregma, unity failed with "dpkg-shlibdeps: error: no dependency information found for /usr/lib/powerpc64le-linux-gnu/libunity/libunity-protocol-private.so.0 (used by debian/unity/usr/lib/compiz/libunityshell.so)", that's a weird one
[20:42] <robru> bregma, well it's not even 2PM here, so if I define "tonight" as 6 hours from now, should be no problem ;-)
[20:43] <bregma> seb128, double weird because that was a rebuild of the same branch as before
[20:43] <tvoss> rsalveti, hey there
[20:44] <seb128> http://162.213.34.102/job/landing-008-1-build/48/console indicates indicator-session as well but that one is fine on the ppa
[20:44] <seb128> bregma, you can maybe get cjwatson or another buildd admin to retry https://launchpad.net/~ci-train-ppa-service/+archive/landing-008/+build/5802886
[20:46] <seb128> those warnings are a bit weird "Use of uninitialized value $v{"DEB_BUILD_GNU_TYPE"} in split at /usr/bin/dpkg-architecture line 188."
[20:46] <seb128> infinity, ^ is that a known issue?
[20:48] <robru> cyphermox, autopilot just landed, please kick an image ;-)
[20:49] <robru> thomi, please merge silo 1 ;-)
[20:49] <thomi>  \o/
[20:49] <infinity> seb128: That was the last build before that machine went insane, I don't think it has anything to do with reality.
[20:50] <bregma> I often feel that way
[20:50] <seb128> infinity, can you retry https://launchpad.net/~ci-train-ppa-service/+archive/landing-008/+build/5802886 for us?
[20:51] <cyphermox> zug zug
[20:51] <infinity> seb128: I will in a second, yeah.  Just saving logs and trying to make sense of WTF what wrong.
[20:51] <seb128> infinity, thanks
[20:51] <seb128> bregma, I would say "wait for the retry, if that works, publish by checking the box letting you ignore the state of the previous steps"
[20:52] <seb128> bregma, the CI is not smart enough to pick the status of a retry, but kicking a rebuild for all archs doesn't make sense, override are there for those cases ;-)
[20:52] <cyphermox> robru: done
[20:52] <cyphermox> robru: it's gonna take a bit
[20:53] <cyphermox> urgh, wait, that's wong
[20:53] <cyphermox> *wrong
[20:53] <cyphermox> robru: always wait after rmadison says it's in release.
[20:55] <robru> cyphermox, oh crap, i was just going for the spreadsheet
[20:55] <robru> cyphermox, usually rmadison says so very shortly after launchpad says so?
[20:56] <cyphermox> robru: nah, can take a little bit
[20:56] <rsalveti> tvoss: hey
[20:56] <cyphermox> not the end of the world, we'll see soon enough
[20:56] <tvoss> rsalveti, hey, unity-mir trunk builds fine locally with latest process-cpp
[20:56] <rsalveti> tvoss: nops
[20:57] <tvoss> rsalveti, did you see: http://bazaar.launchpad.net/~mir-team/unity-mir/trunk/revision/194
[20:57] <tvoss> ?
[20:57] <rsalveti> tvoss: you probably have libproperties-cpp-dev installed as well
[20:57] <rsalveti> tvoss: that was before
[20:57] <tvoss> rsalveti, what exactly fails for you then?
[20:57] <seb128> bregma, if you end your day without publishing unity-lock-screen and don't publish it for a reason please send me an email saying why (just as a fyi, otherwise I might end up trying to the publishing tomorrow when I start my day, if it didn't happen by then)
[20:57] <rsalveti> tvoss: this is the fix https://code.launchpad.net/~rsalveti/process-cpp/adding-missing-dev-dependency/+merge/210482
[20:58] <rsalveti> libprocess-cpp-dev is now depending on libproperties-cpp-dev, but that's not part of the deps list
[20:58] <tvoss> rsalveti, ah, right
[20:58] <rsalveti> tvoss: should be landing the fix in a few minutes
[20:58] <rsalveti> but it's all good already
[20:59] <tvoss> rsalveti, cool, thanks
[21:00] <tvoss> rsalveti, slipped through as we took unity-mir out of the process-cpp landing
[21:03] <rsalveti> tvoss: yup
[21:03] <bregma> seb128, ack
[21:04] <seb128> bregma, thanks
[21:05] <tvoss> rsalveti, my bad
[21:06] <rsalveti> no worries
[21:14] <robru> Wellark, cyphermox: ok I'm building connectivity-api in silo 1 now.
[21:57] <robru> cyphermox, so what happened with the image build? can we tell if it included autopilot or not?
[21:58] <robru> cyphermox, also please review the packaging of the package in silo 1.
[22:01]  * Wellark_ needs a smoke...
[22:01] <robru> Wellark_, cyphermox doesn't seem responsive (probably dinner time for him i guess) so don't hold your breath
[22:01] <Wellark_> robru: I've got all night...
[22:02] <Wellark_> robru: is there anyone else besides cyphermox that could do it?
[22:02] <Wellark_> I'm happy to bug anyone right now ;)
[22:03] <robru> Wellark_, any core dev can do it, see https://launchpad.net/~ubuntu-core-dev just that cyphermox is my go-to guy for this ;-)
[22:03] <Wellark_> ok.. sending email the the whole team then!
[22:03] <robru> hahaha
[22:03] <robru> Wellark_, try to narrow it down to people who are awake right now ;-)
[22:03] <Wellark_> or have we removed that feature from LP..
[22:06] <Wellark_> seb128: are you up for a NEW packaging review? :)
[22:06] <robru> Wellark_, probably that team disabled the permission to avoid spam from you ;-)
[22:07] <seb128> Wellark_, you can still give the link, I might have a look tonight or tomorrow
[22:07] <sergiusens> robru, hey l8 and l26 are the same thing
[22:07] <Wellark_> robru: didn't even have a chance to spam yet!
[22:07] <robru> hehe
[22:07]  * sergiusens just confirms comment in sheet
[22:07] <Wellark_> seb128: silo 1
[22:08] <robru> sergiusens, very similar, but line 8 has one extra MP from you. line 26 only has branches from xnox and doanac
[22:09] <Wellark_> seb128: meaning https://code.launchpad.net/~unity-api-team/connectivity-api/ci-testrun/+merge/210479 and the actual branch to review the packaging on would be lp:~unity-api-team/connectivity-api/ci-testrun
[22:10] <seb128> Wellark_, licensecheck states that src/platform/nmofono/link.h is GPL where debian/copyright lists only LGPL
[22:11] <Wellark_> robru: ugh.. what happened.
[22:11] <Wellark_> the MP turned into "Merged"
[22:11] <cyphermox> robru: it's done, it did contain the new autopilot after all
[22:11] <Wellark_> seb128: uuh, interesting. looking.
[22:11] <robru> cyphermox, great!
[22:11] <cyphermox> robru: I'm just confused by an extra other image
[22:12] <cyphermox> give me a minute
[22:12] <sergiusens> robru, yea, that first MR in there was added to a different silo once upon a time and makes it essentially the same thing
[22:12] <robru> Wellark_, yeah, so what happened was that I had to manually merge it, because citrain got confused by the trunk branch having a different source package name. the first try at building failed because of that. so I just merged it myself, and then rebuilt, and once the source package name all matched then the build succeeded
[22:13] <cyphermox> robru: I don't see test results for that image though
[22:13] <Wellark_> robru: oh, right
[22:13] <cyphermox> but it's basically 232 I guess
[22:13] <robru> cyphermox, maybe you kicked two builds by mistake at the same time? dunno
[22:13] <seb128> Wellark_, looks mostly fine otherwise to me from an eod easy review (would be nice to document why gcc 4. is enforced)
[22:13] <cyphermox> cjwatson: if you're still around, do you know if there's an image that gets kicked off at 2200 UTC? I thought it was only at 0200 UTC.
[22:14] <cyphermox> robru: I don't think so, the timing is too off
[22:15] <Wellark_> robru: where do I push fixes now?
[22:15] <Wellark_> as you merged the previous MP
[22:15] <sergiusens> robru, and for what it's worth; all other branches are stacked
[22:15] <robru> Wellark_, you should be able to push to the same branch... worst case just submit a new mP
[22:16] <Wellark_> seb128: ok, thanks! need somebody (cyphermox, hello:) to do a more detailed review
[22:20] <sil2100> rsalveti, tvoss, Mirv: ok, so it seems that we really need to double-check these process-cpp landings
[22:20] <rsalveti> yeah
[22:21] <robru> sergiusens, sorry just want to confirm: that branch of yours that's in line 8 but not the other line, you don't need that anymore? can i delete line 8?
[22:21] <rsalveti> it seems process-cpp also broke the x86 emulator =\
[22:21] <sil2100> rsalveti, tvoss, Mirv: or make sure the test plans include 'recompile all reverse deps' ;/
[22:21] <rsalveti> but will investigate more
[22:21] <sergiusens> robru, yes, line 8 can be deleted
[22:21] <robru> sergiusens, ok
[22:21] <rsalveti> trying to rebuild a new image for it
[22:22] <Wellark_> sil2100: is everything compiled with gcc-4.7 that uses process-cpp ?
[22:22] <sil2100> rsalveti: we weren't in the mode of double-checking landings anymore, so I didn't check if things work anyways
[22:22] <sergiusens> sil2100, rsalveti or use the system tvoss engineered a couple years ago with fginther that does that automatically ;-)
[22:22] <Wellark_> if I understood correctly there was a dependency added to properties-cpp
[22:22] <Wellark_> which is compiled with g++-4.7
[22:23] <Wellark_> and anything using properties-cpp has to be compiled with g++-4.7
[22:23] <Wellark_> or there will be infinit runtime breakage
[22:23] <Wellark_> AFAIK
[22:23] <sil2100> Wellark_: it's supposed to, since process-cpp and dbus-cpp were already forced to build with 4.7 before
[22:24] <Wellark_> sil2100: and process-cpp does not expose anything coming from properties-cpp in it's public API?
[22:24] <fginther> sergiusens, heh, that had it's time in the sun. It's more than a bit obsolete now.
[22:24] <sil2100> Wellark_: not sure, we should probably ask tvoss once he's around
[22:25] <sergiusens> fginther, too bad; solved many issues we see today :-)
[22:25]  * sergiusens scratches item off list
[22:32] <sil2100> rsalveti, Wellark_: for now, we'll do a quick revert - I'm asking Didier to do it once he's up in the EU morning
[22:32] <sil2100> Since he has some tools to do it quickly
[22:32] <rsalveti> sil2100: sorry, revert what?
[22:32] <rsalveti> sil2100: the unity-mir issue with process-cpp is already fixed
[22:32] <rsalveti> now trying to understand what else might be broken with it
[22:33] <sil2100> Ah, ok
[22:33] <cjwatson> cyphermox: right now there's only a commented-out crontab entry at 0203
[22:33] <sil2100> rsalveti: how did you fix that issue in the end?
[22:34] <rsalveti> sil2100: missing dependency https://code.launchpad.net/~rsalveti/process-cpp/adding-missing-dev-dependency/+merge/210482
[22:34] <sil2100> rsalveti: you released it? Or is it still pending a release in CITrain?
[22:35] <sil2100> Oh, I see it's released
[22:35] <sil2100> Ok, nevermind then
[22:36] <rsalveti> yeah, let's just make sure we test build-deps when landing it in the future
[22:37] <sil2100> rsalveti: I already mentioned it to the guys - we need to explicitly make sure the TestPlan includes 'rebuild all rdeps' for every -cpp project that there is
[22:37] <Wellark_> all rdeps?
[22:37] <sil2100> Reverse-depends
[22:37] <Wellark_> that might mean dozen of projects
[22:37] <sil2100> Wellark_: I mean properties-cpp, process-cpp and dbus-cpp - those have only few rdeps
[22:38] <Wellark_> for now
[22:39] <sil2100> If it expands, then we think of automating it somehow
[22:39] <sil2100> For now, I don't want things to be broken
[22:40] <sil2100> See you tomorrow
[22:48] <robru> bah, I just noticed that it's 4PM and I haven't had lunch yet. bbl!
[23:02] <Wellark_> robru: could you update the MP linked to silo 1?
[23:02] <Wellark_> robru: this is the new one https://code.launchpad.net/~unity-api-team/connectivity-api/ci-testrun/+merge/210511