[00:00] ondra: awesome, please release that soon [00:00] :-) [00:00] rsalveti: actually just working on dualboot now, adding ubuntu and android update support [00:01] awesome [00:01] rsalveti: did you ask to use own ubuntu build with dualboot, or to have recovery functional unless you reboot to Ubuntu from the app? [00:03] rsalveti: BTW do we still need SWAP with 4.4.2 ports? [00:03] ondra: I was just wondering if there was any other way to update the image without downloading it from the app [00:03] as that's really really slow [00:04] ondra: we don't, will ping ogra_ tomorrow to get that disabled after our next promoted image [00:06] rsalveti: so I can share with you app we used at mwc, that has this feature already enabled [00:07] rsalveti: I did some extra tweaking and scripting, so that you can use already if you want [00:07] rsalveti: OK so I can drop swap from dualboot as well then [00:08] ondra: that would be nice, I believe Saviq might want that as well [00:08] rsalveti: I will drop this to my chinstrap then [00:09] * Saviq wants [00:11] rsalveti: okubik/public_html/humpolec/mwc [00:12] rsalveti: use script dualboot-one-device.sh [00:14] rsalveti: use param 'PUSH_CUSTOM' followed by list of the tar.xz you want to push to the phone [00:15] rsalveti: just remember to update app with one included there, at least once [00:18] ondra: that's great, will give it a try [00:18] rsalveti: let me know if it works, still a bit work in progress [00:19] sure [00:19] rsalveti: close app and let it work, it will open app and button should be at state "resume", it will resume with your own files [00:21] ondra: hm, got app not installed after trying to install the apk [00:21] ondra: do I need to remove it first? [00:22] hm, it's a built-in app [00:22] rsalveti: probably not signed properly [00:22] can't even remove it [00:23] ondra: should I use UPDATE-UbuntuInstaller.zip? [00:23] rsalveti: adb reboot bootloader && fastboot boot recovery && adb mount /system && adb push /system/app [00:24] alright [00:24] rsalveti: this will replace system app with one with our own key [00:24] ondra: PUSH_CUSTOM is not available in https://chinstrap.canonical.com/~okubik/humpolec/dualboot-one-device.sh [00:24] rsalveti: or you can remove it from system, and then you can update with adb install [00:24] rsalveti: humpolec/mwc/ [00:24] wonder how that'll work for me with adb going crazy :| [00:25] ondra: oh, sorry [00:25] rsalveti: sorry all the files are in mwc folder [00:29] ondra: then what should I give as arguments? device.tar.xz and ubunut.tar.xz? [00:31] rsalveti: yes [00:31] rsalveti: but ubuntu always as first! [00:31] rsalveti: since our brilliant overlaying, you need to have rootfs as first [00:31] ondra: oh, ok [00:31] damn, that's not gonna work, wth is happening ;| [00:32] rsalveti: it will take any number of packages [00:33] rsalveti: I put there also scripts which will convert you tar.gz preinstalled into system-image package which you need for dualbootr [00:33] cool, but I'll try with the official image first, let's see [00:33] rsalveti: same is there for converting device image, just point that script to your out/target/product/mako [00:33] ./dualboot-one-device.sh PUSH_CUSTOM ubuntu-c6b7105334c67a0c762f84904a3dc4e7818b7e7a1a95eef6f82c149ff0bc6f44.tar.xz device-456bc5a53bcbb0481be13acacc4de673433c4ea0255ffabe0cbc35925481715d.tar.xz [00:33] Saviq: what are you mumbling? :) [00:34] rsalveti: correct [00:34] ondra, my mako USB reconnects every 5s [00:34] rsalveti: then you are doomed, it needs almost minute to push all the files to the phone [00:34] Saviq: what do you get in your host syslog? [00:35] rsalveti, usb disconnect/connect, nothing more interesting [00:35] rsalveti: you can also passed to it device id as first param if you have more devices, as Saviq once requested :P [00:35] ;D [00:35] rsalveti, http://pastebin.ubuntu.com/7122863/ [00:35] ok, app was opened automatically [00:36] rinse&repeate [00:36] -e [00:36] rsalveti, Saviq if things go well, I will be able to use Ubuntu upgrader to update installed Ubuntu with deltas :) [00:36] ondra, /me awaits [00:37] ondra: hm, can't see any resume in here [00:37] rm: cannot remove ‘update_command’: No such file or directory [00:37] ondra, so basically instead of recovery doing that work, the app in android will? [00:38] * Saviq starts from scratch [00:38] Saviq: yeah [00:38] rsalveti: yeah, app will do it [00:39] ondra: pushed the app to /system/app, as you said [00:39] rsalveti: magic is to convince android not to delete downloaded update before anything happens [00:39] ondra, mhm [00:39] but still can't see any resume option [00:39] rsalveti: hmm, and app was not running and script started the app? [00:39] ondra: yeah [00:40] running it again with the app opened [00:40] rsalveti: OK I wonder if version of the app is wrong [00:40] says 0.3 here [00:40] rsalveti: no app needs to be closed [00:40] no, still nothing [00:40] rsalveti: so version should be right === chriadam|away is now known as chriadam [00:41] let me reboot and try again [00:44] no resume :-( [00:44] ondra: http://paste.ubuntu.com/7122890/ [00:44] we need to allow logging the phone in via QR code [00:44] typing a strong password through is crazies [00:45] hahah, yeah [00:45] every freakin' time [00:45] rsalveti: looks all good [00:45] rsalveti: OK uninstall the app [00:46] ondra: just remove it? [00:46] from recovery [00:46] rsalveti: no you can just uninstall it, we will install one from zip file [00:47] rsalveti: actually this could be the problem [00:47] rsalveti: you might still have old app [00:47] rsalveti: sometimes android is really stubborn to remove old app [00:47] rsalveti: so from recovery just delete /system/app/UbuntuInstaller.apk [00:48] right [00:49] removing, rebooting [00:49] ;( reflashing phone didn't help [00:49] reboot, biab [00:50] ondra: then use the zip in recovery? [00:50] rsalveti: no just pick app from that zip file [00:50] rsalveti: and install it with adb install [00:50] ondra: sorry, which zip? [00:51] https://chinstrap.canonical.com/~okubik/humpolec/mwc/recoveries/UPDATE-UbuntuInstaller-v0.3.zip ? [00:51] rsalveti: boot normally now and make sure app is gone [00:51] rsalveti: yeah that is the one [00:51] file 'UPDATE-UbuntuInstaller-v0.3.zip' does not contain AndroidManifest.xml [00:53] rsalveti: but you don't need that [00:53] rsalveti: hold a sec, i will get you the app [00:53] ondra: just grab the apk from that zip? [00:53] rsalveti: yeah [00:54] installed [00:54] rsalveti: I just replaced one in chinstrap with one from zip [00:54] rsalveti: with adb install? [00:54] rsalveti: and when you booted there wasn't dualboot app? [00:54] rsalveti: so now rerun script, fingers crossed [00:55] ondra: it wasn't there, so I guess it removed successfully [00:57] rsalveti: OK [00:58] ondra: ok, resume is available now \o/ [00:58] ondra, bootstrap possible with that new option, or do I need to bootstrap once before I can go with "resume install"? [01:01] rsalveti: sweet :) [01:01] Saviq: what do you mean by bootstrap now/ [01:01] ? [01:02] ondra, I freshly flashed my mako with android [01:02] Saviq: that's perfect [01:02] Saviq: app will backup your stock recovery [01:02] ondra, and installed the app through recovery [01:02] Saviq: and once you uninstall ubuntu it will restore it [01:02] Saviq: no need [01:03] Saviq: but you will need to install SU through recovery [01:03] right, SU, /me does [01:04] Saviq: if you don't mind app in the system partition, you can run FULL command, it will install some old Ubuntu though [01:04] ondra: we might just need to improve it to accept a build number [01:05] root@ubuntu-phablet:/# system-image-cli -i [01:05] current build number: 0 [01:05] other than that, working fine [01:05] Saviq: find function "complete_ubuntu_instalation()" [01:05] rsalveti: I'm working on it now [01:05] rsalveti: it will be getting build number from installed ubuntu [01:06] ondra: awesome, thanks so much, you can't image how easier my life will be now hahah [01:06] rsalveti: so if you pass to it "version.tar.xz" it will pick correct version [01:06] tired of waiting almost one hour to get this device flashed [01:06] rsalveti: welcome :) [01:06] ondra: great [01:06] ondra, SuperSU installed, can I just go and PUSH_CUSTOM now? [01:06] ondra: let me know when you get that working [01:07] rsalveti: hope soon, I will need some beta testers so I will ping you [01:07] Saviq: adb install UbuntuInstaller.apk? [01:07] thanks [01:07] ondra, it's already installed [01:07] Saviq: then you can just PUSH_CUSTOM [01:07] ondra, cool, let's see [01:09] ondra, hmm no "resume install", only lets me choose the channel [01:10] * Saviq might not be using the right app [01:12] Saviq: you need to close the app, run the script, once that is done, script with start app with extras in the intent pointing to sideloaded packages [01:13] ondra, yeah, I think I might not have installed the correct version of the app [01:13] Saviq: use one from the zip file, older but will work [01:13] used the UPDATE v0.3 [01:13] now [01:14] \o/ [01:15] ondra, it's alive! ;) [01:16] Saviq: nice :) [01:16] OK I'm off to get some sleep now [01:17] tomorrow guys === ondra is now known as ondra|away [01:17] cya [01:27] rsalveti, does the app lie about the installed ubuntu version for you as well? says image 200 from channel trusty for me, but it should be 247 (and also system-image has no idea about the number, that's what you said above?) === jhodapp|afk is now known as jhodapp [01:33] Saviq: yeah === jhodapp is now known as jhodapp|afk === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === tk is now known as Guest4571 [03:49] i have been trying to port ubuntu touch on i9082 for past several hours.. stuck on device boot [03:49] need suggestions === jonahbron is now known as jonahbron_ [04:33] Hello, I'm working on adding a MenuBar to my QML app. When I add it, the menu doesn't appear. Any idea why? === OutOfControl is now known as benonsoftware [05:21] hi all [05:22] is it going to be possible to get ubuntu touch experimentally on my samsung galaxy tab 3 smt210? [05:23] everything is possible [05:24] where can I find the img? [05:25] might have to port on your own [05:25] hmm sounds tricky === JoseeAntonioR is now known as jose [07:41] good morning [08:12] tsdgeos: I wonder if you could take a look at the calendar-app crasher bug #1294995 - trunk having it is a blocker for the actual calendar-app blocker, which is the last one preventing promoting Qt 5.2 based image [08:12] bug 1294995 in qtdeclarative-opensource-src (Ubuntu) "qmlscene crashed with SIGSEGV in value() when running calendar-app" [Undecided,New] https://launchpad.net/bugs/1294995 [08:13] it seems the trunk could see a revert that "fixes" the issue, but a crash would be good to be looked at anyway [08:25] Mirv: let me have a quick look [08:25] it's quite easy to trigger from trunk [08:25] Mirv: desktop also? [08:28] "Running the app on the desktop via qmlscene segfaults when manually selecting dayview as well." sayss the bug [08:28] tsdgeos: yes, especially desktop since I don't know yet if people are getting that on device and I fail to remember how to upgrade trunk for click packages on device [08:28] works for me [08:29] tsdgeos: bzr branch lp:ubuntu-calendar-app, cd u*, bzr bd, dpkg -i, /usr/bin/calendar-app? [08:30] no, not that [08:30] let me try [08:30] for me it crashes every time I click first "Week", then click "Week" again to reveal the lables again, then click Day [08:30] i'm just doing qmlscene calendar.qml [08:30] and that works [08:30] so they managed to trigger that bug in the latest commit which now then prevents landing fixes for the AP tests [08:30] let me see if doing what you say makes any difference [08:30] the /usr/bin doesn't do anything else either [08:33] also happens with LANG=C qmlscene /usr/share/calendar-app/calendar.qml [08:34] let's see if that helps me [08:34] maybe it just likes catalan [08:34] hey didrocks! mind a quick PM? [08:34] jose: sure [08:36] tsdgeos: there's CoreDump and friends in the bug too, though, if it helps [08:37] Mirv: i see that i did not have qtorganizer5-eds installed that seems to be needed [08:37] let me see if that does anything [08:37] may be the fact also that i have no appointments at all [08:37] Mirv: do you? [08:37] tsdgeos: yes to both [08:38] I've appointments via evolution, and qtorganizer5-eds installed [08:38] that may make it do more stuff than my empty thing [08:40] Mirv: i marked bug 1294671 as a dupe of your bug 1294995 - dunno if it has more tracing, but has core dump [08:40] Error: Launchpad bug 1294671 could not be found [08:40] bug 1294995 in qtdeclarative-opensource-src (Ubuntu) "qmlscene crashed with SIGSEGV in value() when running calendar-app" [Undecided,Confirmed] https://launchpad.net/bugs/1294995 [08:41] popey: thanks. the 995 has the best trace via lp:daisy usage [08:41] k [08:51] ogra_, would it be possible to start adbd if the container fails to start as well? [08:52] it helps debug while porting [08:59] janimo`, i'll try to come up with something [08:59] janimo`, until then just change the upstart job back to what it was i guess [09:00] ogra_, I did that, sure, just thought it would be good fixed properly :) [09:01] cwayne added a function to rootstock that can replace single files inside a build btw (for example the adbd job ;) ) [09:01] ogra_, maybe the container emit some error signal and have that too trigger adbd [09:01] yeah [09:01] * janimo` knows too little about upstart to suggest anything else [09:01] i need to find if thats possible [09:01] its not so much upstart but rather lxc [09:01] ogra_, for the phone I am working on I also have an overlay thingy copied over at buildd time so no biggie [09:01] ah, cool [09:02] ogra_, also changes can - as I learned yesterday - be added using the regular system-image deploymnt mechanism [09:02] very convenient [09:02] right [09:02] (rootstock install just copies that script :) ) [09:03] I no longer know where rootstock fits in now with so many scripts and different ways of flashing a device [09:04] I am glad we can finally make this port work with system-image [09:04] and not the legacy layout [09:04] well rootsotck is a tool for porters that dont want to bother with creating system images [09:05] Mirv: should adding meetings from the calendar-app work? [09:05] yes [09:05] doesn't work here [09:05] known bug editing them though [09:05] maybe something is missing a dependency to evolution-data-server ? [09:05] * tsdgeos tries installing that [09:06] it seems Olivier now provided a better backtrace with QV4_FORCE_INTERPRETER=1 [09:06] ogra_, right now I reuse pristine Ubuntu images and add an overlay at build time. I think this is simplest [09:06] janimo`, the installer is identical to the system-image script inside the recovery but you can hand over the android system.img of your build directly (and a rootfs traball frokm cdimage) the result is identic to a s-i install [09:06] that's better [09:07] and now i get it to crash [09:07] so yes it only crashes if you have data in your calendar [09:07] and you need eds to be installed [09:07] to be able to add data [09:13] my mouse pointer flickers with dual monitors (2 LED & 1 Built-in Display) , any easy fix ?? [09:13] ubuntu 13.10 [09:23] do you understand if https://code.launchpad.net/~timo-jyrinki/ubuntu-calendar-app/bug_1293489_mega-merge/+merge/211887 makes sense or not? ETOOMANYBRANCHES [09:35] Good morning all; happy World Storytelling Day! :-D [09:37] Mirv: so found http://pastebin.com/7kVvV6kr that seems to help, but i still need a testcase for Qt [09:46] tsdgeos: \o/ [09:46] good to have something, as one never knows when that crasher appears somewhere we can't workaround [09:53] Has someone tried putting a full length movie to Touch? I just copied one x264 trailer and it works, but a full length movie in same format does not show up at all [09:55] timppa: sounds like worth filing a bug against https://bugs.launchpad.net/mediascanner2 or something.. I think it's possible no-one has tested a full length movie [09:57] Mirv: is Video lens correct "term" for it? [09:58] Mirv: and where can I find logs for the mediascanner? [10:03] timppa: yes, video lens. my knowledge on the media side is not too deep, but it seems there are log files under .cache/upstart/mediascanner.log [10:03] Mirv: I'll have a look on those also [10:03] Mirv: thanks [10:03] np [10:04] yes, I can see stuff happening when I do tail -f mediascanner.log and then copy videos in [10:07] Mirv: I see error: Unable to discover /home/phablet/Videos/filename.mp4, error code: 5 === vrruiz_ is now known as rvr [10:08] timppa: oh, note also mediascanner-2.0.log [10:08] yes, that gives a different error [10:08] that seems quite clean, it gives an error for one of my media files and then "Added video" for others [10:08] Mirv, hmmm, jim has watched full movies with the media player [10:09] not sure he was monitoring the scanner [10:10] ok then, maybe it's related to the more limited scope of support then. like is h264 AVC high profile supported, or only mainline/base profiles? [10:15] Mirv: https://bugs.launchpad.net/mediascanner2/+bug/1295062 [10:15] Ubuntu bug 1295062 in mediascanner2 "Full length movie does not show up in Video lens" [Undecided,New] [10:16] probably wants a distro task for that? [10:20] timppa: mediascanner-service[2001]: ERROR error/fs-walker: Content discovery failed for "/home/phablet/Videos/Cosmos.A.Space.Time.Odyssey.S01E01.720p.HDTV.X264-DIMENSION.mkv": gst-core-error-quark(12): Your GStreamer installation is missing a plug-in. [10:20] related? [10:20] popey, shouldnt be if a trailer of the same format works [10:21] (if it really *is* the same indeed) [10:29] ogra_: I look codec info with vlc, it *looks* the same :) [10:29] popey: I did not have that kind of error messages [10:30] ok === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [10:50] ChickenCutlass: https://bugs.launchpad.net/music-app/+bug/1292306 [10:50] Ubuntu bug 1292306 in Ubuntu Music App "Upon upgrading to Qt5.2 the music app no longer plays the next song if the screen is off" [Critical,Triaged] [10:50] ChickenCutlass: there is a test attached, but doubt this will really cause the issue in automation beause adb etc. [10:50] will keep our device awake [10:50] so guess we would need a mechanism to put phone into deep sleep for N seconds [10:50] and have it auto wake up [10:50] can we do that? [11:11] asac, hard to tell, we would need to disable adb automatically for a certain time or so ... [11:16] ogra_: from host side? [11:16] no [11:16] on the device [11:16] Morning all [11:17] asac, a nohup script that calls: setprop persist.sys.usb.config mtp, sleeps a while and then adds adb back to the property might work [11:17] ChickenCutlass: do you know if the emulator has similar properties wrt to powermanagmenet as our devices? e.g. sleeping if no wake locks [11:18] not sure if nohup actually works since the shell will be completely dead when adbd is off [11:18] i'll do some testing later today [11:21] bah, sigh [11:21] http://people.canonical.com/~ogra/touch-bootcharts/ubuntu-phablet-trusty-246.png vs http://people.canonical.com/~ogra/touch-bootcharts/ubuntu-phablet-trusty-248.png ... the container starts 2sec later with 248 [11:21] * ogra_ wonders why [11:23] asac: I know we needed to turn off powerd in the emulator. it does not behave the same as the device === jhodapp|afk is now known as jhodapp [11:54] timppa, full length movies work just fine, it has more to do with supported codecs and profiles [11:54] timppa, I test playback with many different lengths of video files [12:07] dobey: hi! About your question on enabling account plugins only on desktop/touch, please file a bug [12:07] erm, is there no email client just the gmail one? also is there a way to access caldav/carddav/webdav resources? [12:07] dobey: I'm afraid that this will require a FFe, and in that case if you took care of the bug his would save me quite some time :-) [12:08] awafaa: an email client is in development [12:08] mardy: ah, thanks === MacSlow is now known as MacSlow|lunch [12:44] zsombi: any idea why this wasn't merged/landed yet? is it not ready or something? https://code.launchpad.net/~zsombi/ubuntu-ui-toolkit/statesaver-discard-shutdown/+merge/205042 [12:45] asac: it was blocked because of Qt5.2 landing, so we have to redo the tests now with 5.2. once those are done, we can land it [12:48] zsombi: ok please check with didrocks. thought it was one of the thigns pat scheduled for landing asap after qt5.2 [12:48] e.g. critical issues before promotion [12:48] so please follow through i would assume [12:48] zsombi: there is an issue with alarm sound not working [12:49] that popey and davmor2 discovered [12:49] is that supposed to fix it? [12:49] didrocks: that's not muy business :) charles may know more about that [12:49] zsombi: hum, so you just checked that there is no more denials? [12:49] in syslog [12:49] didrocks: I don't think so, that MR only fixes the statesaver state file [12:50] didrocks: Alarms API uses QOrganizer, and has nothing to do with setting sounds or not. Not even with playing. [12:50] zsombi: so, you think the denials are not due to a missing profile file and then is denied to play sound? [12:51] this wasn't what the denials were about? [12:52] didrocks: my comments were for 14:48 zsombi: there is an issue with alarm sound not working [12:53] zsombi: yeah, can you explain with dumb words what's the impact of the bug exactly, and what it fixes? [12:53] like "can't set an alarm clock" [12:53] or anything else? [12:56] didrocks: from teh comments in the bug it seemed that StateSaver is the guilty as the config path seemed to be related to that feature. the MR I proposed was about to solve the placement of the State database from ~!/.config to /tmp. That's what the bug sais. Nothing about the ability to get alarm sounds or not. [12:57] ~/config is the proper path [12:57] zsombi: right, and this sate file is used to set an alarm, right? [12:57] didrocks: nope === alan_g is now known as alan_g|lunch [12:58] didrocks: the StateSave is not doing anythiong for the alarms, unless the app uses it for that. But if it does, then it does wrong. StateSaver is to save th eapp state in case it is closed by the app lifecycle manager [12:59] zsombi: ah, ok, making sense and so, not that critical for now [12:59] thanks [12:59] didrocks: welcome :) [13:02] zsombi: can you check out this alarm bug? seems its a qt 5.2 regression and one of two things blocking promotion [13:02] zsombi: or find sonmeone in sdk team who can see? i think the alarm clock didnt change, so something else must have landed that broke it [13:03] didrocks: sergiusens: can you confirm that alarm app didnt land recently? [13:03] asac which bug? [13:03] zsombi: the bug you talked with didrocks about [13:03] alarm broken since qt5.2 landing [13:03] * asac wonders why that wasn't fixed. thought we had that solved - shrug [13:03] asac, I can't confirm; balloons has been doing most of the uploads of late in coordination with didrocks [13:03] or guess we didnt try the app [13:04] just the APs [13:04] asac: it didn't land [13:04] asac, by alarms you mean calendar? [13:04] clock [13:04] last version was before 5.2 landing [13:04] sergiusens: clock yep [13:04] sergiusens: didrocks has the details [13:04] not sure what the bugid is [13:04] I can check on the webstore [13:04] I just got the info it started to break with the 5.2 landing image [13:04] http://people.canonical.com/~ogra/touch-image-stats/20140314.1.changes [13:05] sergiusens: if you can double check, I quickly went to all diff from last image to that one ^ [13:05] New version: 1.0.373 on 2014-03-10 22:25 - 1 week, 2 days ago [13:05] popey: davmor2: do you have a bug for it? [13:05] Ran 20 tests in 545.735s [13:05] OK [13:05] sergiusens: yeah, so before 5.2 [13:05] asac: didrocks: so it ends up that the bug is on the critical path then? [13:05] didrocks: for which? [13:05] sergiusens: there is no test for alarm [13:05] zsombi: alarms broken, yes [13:05] zsombi: just before the promotion, yeah :/ [13:05] popey: the alarm regression [13:05] zsombi: it was even critical before qt5.2 [13:05] not yet [13:06] zsombi: not sure why its still not working. thought we fixed that. guess something fell somewhere through the crack. [13:06] didrocks, one thing that needs to be tested as well is upgrade from trusty --wipe to trusty-proposed; some apps don't seem to handle the transition [13:06] at least for me [13:06] asac: didrocks: are we talking about this bug #1288742? [13:06] bug 1288742 in ubuntu-ui-toolkit (Ubuntu) "clock app using incorrect path for config file" [High,In progress] https://launchpad.net/bugs/1288742 [13:07] zsombi: it was raised as a possible cause for the bug by the security team, but seems it's not as per our discussion [13:07] is there a bug report then for the bug that is being discussed now? [13:08] sergiusens: you lost data? I guess we'll have only time from stable to stable [13:08] didrocks: ok, I got confused again... :) [13:08] 14:05:18 didrocks | popey: davmor2: do you have a bug for it? [13:08] 14:05:30 popey | didrocks: for which? [13:08] 14:05:54 popey | not yet [13:08] t1mp: ^ [13:09] ok, thanks [13:09] popey: are you midst filing one if not I can [13:09] go for it [13:11] didrocks, no, music app doesn't launch [13:12] sergiusens, rsalveti [ 13.755104] init: cannot find '/system/bin/ubuntuappmanager.disabled', disabling 'ubuntuappmanager' [13:12] sergiusens: "nice" :/ [13:12] how about we finally rip that stuff out [13:12] sergiusens: do you have a bug report/reproducer for that? [13:12] sergiusens: while we are on the list of criticals [13:12] popey: on it [13:14] jhodapp: which codecs are supported? [13:14] didrocks, ogra_ that's been like that since we switched to mir... [13:14] ? [13:15] timppa, depends on your hardware [13:15] sergiusens: ah, so it's more a stable to stable… [13:15] test [13:15] then [13:15] (sorry my comment was unrelated to the other conversation) [13:15] not for next image [13:15] ogra_, that is only used for surfaceflinger [13:15] jhodapp: Nexus4 and Nexus7 (2013) [13:15] timppa, check out /etc/media_codecs.xml [13:16] sergiusens, well, didnt we want to drop support for it once screenshooting works on Mir ? [13:16] jhodapp: ok [13:16] didrocks, yeah, my test is from trusty to trusty proposed assuming the next trusty proposed is going to be the next stable [13:16] sergiusens: so it's not there since Mir is in [13:16] ogra_, that's a tough question; I would of dropped it ages ago [13:16] ogra_, that things was supposed to last a week [13:16] thing* [13:16] heh [13:17] * didrocks is confused [13:17] lets talk about it in the standup [13:17] jhodapp: no such file!? [13:17] timppa, do a find for it, must have been moved [13:17] davmor2: any bug that we can give to zsombi and t1mp? [13:17] sergiusens: tell me once you get a bug for it, seems it needs to be on the blocker list [13:18] t1mp, popey, didrocks, zsombi: bug #1295122 [13:18] jhodapp: /android/system/etc/media_codecs.xml [13:18] bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Undecided,New] https://launchpad.net/bugs/1295122 [13:18] timppa, yeah that's it [13:19] didrocks, yeah, I'll need to reproduce cleanly which means wiping MY phone ;) [13:19] didrocks: ok, this is something charles can do [13:19] jhodapp: basically if the file works on android it should work with ubuntu? [13:19] davmor2: hum, why indicator-datetime? [13:19] zsombi: is charles avaialble? [13:19] didrocks: what else would it be? [13:19] didrocks: popey said that it was the indicator that triggered the alarm [13:20] timppa, for the most part, but the most tested codec pair is h.264 and aac [13:20] sergiusens: doing that now? I think it's not something that we should wait for having another firedrill at the last minute [13:20] asac he should be available later yes [13:20] didrocks: setting the alarm is working so the clock app is doing it's bit [13:20] zsombi: how long? [13:20] popey: davmor2: I don't think the service is datetime [13:20] charles: wake up :) [13:20] didrocks, well I'm dealing with the music app now [13:20] charles: important mission for you [13:20] didrocks, I can't parallelize this [13:21] sergiusens: I thought the music app issue wasn't critical where you change seems to be? [13:21] didrocks: feel free to change it [13:21] popey, davmor2 do you have devices to test this? [13:21] sergiusens: "this"? [13:21] no, that [13:21] s/change/bug/ [13:21] popey, davmor2 -> ubuntu-device-flash --wipe --channel devel ; ubuntu-device-flash --channel devel-proposed (a simulated upgrade path) [13:21] jhodapp: ok, thanks! I'll try to see if the file is indeed in incorrect format [13:21] ok === _salem is now known as salem_ [13:22] sergiusens: what is that to achieve? [13:22] hi folks :) [13:23] thostr_: when charles will be available? [13:23] popey, keep launching the music app :-) [13:23] zsombi: in about an hour or 1.5 usually [13:23] cjwatson: seems we have an issue that makes music-app not start when you upgrade a device from current promoted to current dev image [13:23] asac: ^ [13:23] cjwatson: but that doesnt happen if you do a fresh install. think this could be us? [13:23] or rather sdk? [13:24] sergiusens: ok [13:24] I was wondering whether I could contribute some things to get ubuntu / ubuntu touch working on RK3188-based tablets like the Vido N90 I have [13:24] cjwatson: see with sergiusens, didrocks and friends for details ... think they filed a bug alread [13:24] tedg: ^^ see the two lines for colin above [13:24] guess you might also know [13:24] asac: I'm waiting for sergiusens to file one :) but seems popey will reproduce first [13:25] I could provide the partition scheme and boot options, firmware files if needed :) [13:25] sergiusens: flashing 237 on flo [13:26] asac, Do we know what error it's giving? [13:27] asac: do we have logs or anything? [13:27] snap [13:27] :-) [13:27] could anybody give me some advice ? [13:28] sergiusens: didrocks: have to run to lunch, can you give tedg and cjwatson info about the startup issue? [13:28] dholbach: maybe you would know (porting guide? ^) [13:28] thanks [13:28] will be back in 20 i hope [13:28] I think I found a bug. My current build number is 237, but if I want to upgrade, the system is upgrading to 237. phone is nexus 4. [13:28] asac: sergiusens has the info, I don't… [13:28] We need a way to convince apport to file bugs on click packages. [13:28] We've got an "app failed" upstart job, we could just upload automatically. [13:29] errors would be better for that [13:29] tedg, what I see with upstart-app-launch and my instance of music is that it doesn't error but I don't get music to start up in any way [13:29] asac: lunch? lucky you :p [13:29] afaik ev has that in his agenda [13:29] Certainly, but isn't apport effectively the frontend for errors? [13:29] sure, you just specifically said "bugs" that's all [13:30] sergiusens, Anything in the log file in ~/.cache/upstart/application*music*.log [13:30] cjwatson, Heh, we need a name for errors DB entries. pre-bugs? ;-) [13:30] tedg, there is no log file created [13:31] sergiusens: want me to put music on this and play with the app between flashes? [13:31] No log file... [13:31] sergiusens, Do you have a device doing this right now? [13:31] sergiusens: can you check whether the music entry in ~/.local/share/applications/ exists and whether the various paths in it all point to sane places? [13:32] cjwatson, sure one sec [13:32] sergiusens, If so, can you run this and see if a log file gets created on next run? initctl set-env --global G_MESSAGES_DEBUG=all [13:32] tedg, yes [13:39] tedg, cjwatson paths seem to be correct http://paste.ubuntu.com/7125271/ ; thanks for the set-env thing [13:39] I do notice there's still a -I in there; not sure if added back or never removed as it's using the import path upstart-app-launch sets now [13:40] the error would be different as well [13:40] sergiusens: paths> agreed. doesn't look like a click problem [13:41] well...that's sad...I'd love to participate, but if noone cares... [13:42] crawler2014, see the porting doc from the channel topic [13:42] sergiusens, Yeah, I don't see anything odd in that log other than the icon seems to be in the root instead of in "images" [13:42] Clearly shouldn't break starting up though. [13:43] crawler2014, we officially only suppport a few nexus devices, other doing ports is up to the community [13:45] sergiusens, I'm a little concerned that you're not getting lttng errors. [13:46] I hadn't figured out how to suppress those yet. [13:46] tedg, music is unconfined [13:47] sergiusens, Ah, okay. [13:48] tedg, wait; am I supposed to see the qt import dir setup when launching with debug? [13:48] or is that not printed/logged? [13:48] sergiusens, Let me check [13:49] sergiusens, No, not printed. [13:50] sergiusens, Where is the vala comment comming from? [13:54] tedg, let me launch again and see if I can get the process [13:54] Ah, got it, that's the zg posting [13:54] fginther, so autopilot-testrunner-touch-saucy involves unlocking the screen [13:54] fginther, and I'm trying to land a branch that separates the greeter [13:55] fginther, what's the best way to land a branch against autopilot-testrunner-touch-saucy in sync with a code change like that? Can it be part of the same silo? [13:55] tedg, hmm, is it ok for upstart-app-launch to return 255? [13:56] guess not [13:57] sergiusens: that kind of sounds like an exec failure ... === alan_g|lunch is now known as alan_g [13:58] sergiusens: does that aa-exec-click command from the .desktop file work if you run it with the current directory set to the value of Path? [13:58] er and I suppose you have to substitute %f too [13:59] I think it expands to empty? [14:00] sergiusens: flashed latest unpromoted on flo.. unity is in a restart loop ☹ === chihchun is now known as chihchun_afk [14:00] popey, not here with OTA [14:00] cjwatson, no worries; not the first time I launch with that; but I would need to set the import paths manually (which u-a-l does for me) [14:00] ogra_: i clean flashed promoted image with --wipe, put some music on it, then flashed latest unpromoted [14:00] sergiusens: ^^ [14:01] popey, that's worse than my issues [14:01] popey, right, i just did an OTA upgrade here, no issues with that (though i dont have any music on it) [14:01] ogra_, well you've been getting gradual improvements; this simulates ota from trusty current to trusty next [14:02] which has a bundle of changes === dandrader is now known as dandrader|afk [14:03] yeah, rebooted device and now it's just not starting unity without dying [14:04] sergiusens, Sorry, got pulled into a HO. [14:05] sergiusens, Yes, I'd agree with cjwatson that it's probably an exec failure. [14:05] sergiusens: on my flo the music app opened [14:05] davmor2, with simulated upgrade or a fresh flash [14:05] * sergiusens knows that it works fresh [14:06] sergiusens: with the command you gave me [14:06] hmm [14:06] davmor2, ok; then I'll drop this one for now [14:06] sergiusens: meh hang on [14:07] uh oh [14:07] sergiusens: it looks like it only ran the first flash and not the second I'm on 237 [14:08] sergiusens: reflashing the second part [14:08] sergiusens, I guess the next thing I'd do is see if the apparmor profile is correct. [14:08] sergiusens, Is there a deny in the system logs? [14:09] tedg, nope; it's unconfined :-) [14:09] sergiusens, It can't be unconfined and be a click, it has to a null confinement. [14:10] sergiusens: D'oh that's why. The second flash tried to run as soon as the first exited but it was still rebooting the device at the time [14:10] sergiusens, sudo aa-status | grep music [14:11] well, I don't know what 'null confinement' means [14:11] I like to use the term 'effectively unconfined profile' [14:11] Heh [14:12] or 'it uses the unconfined template' [14:12] phablet@ubuntu-phablet:~$ sudo aa-status | grep music com.ubuntu.music_music_1.3.385 [14:12] tedg, it uses the unconfined template as jdstrand mentions [14:13] popey, Are you implying you don't have an apparmor profile for it? [14:13] popey: curious-- does 'apparmor_parser -p /var/lib/apparmor/profiles/click_com.ubuntu.music_music_1.3.385 || echo fail' say fail? [14:13] i am not implying anything [14:13] i pasted what you typed ☻ [14:13] popey, Was there output? [14:13] tedg: his paste shows it has a profile [14:13] i get stuff [14:13] com.ubuntu.music_music_1.3.385 was the output tedg [14:13] Oh, I didn't get the return. [14:13] sorry [14:14] #blameadb [14:14] sergiusens: music app still opened it just took it's time [14:14] jdstrand: no, it doesn't say fail [14:14] ok, good [14:15] tedg, http://paste.ubuntu.com/7125434/ [14:15] more worrying to me is that unity flat out wont start [14:15] popey: so, you have the "after upgrade" issue? [14:15] didrocks, he has a different issue though [14:15] didrocks: I have _an_ issue, dunno if it's one others have seen [14:15] hum [14:15] unity keeps respawning [14:16] nice [14:16] didrocks, the music stopping problem isn't music app; seems platform [14:16] all signals are blocked on suspend [14:16] blimey, finally stopped respawning, no idea why. [14:17] sergiusens: yeah, discussing that with ogra_ === chihchun_afk is now known as chihchun [14:17] sergiusens: so alarm can be the same… [14:17] yeah [14:17] yup [14:17] i would say thats very likely [14:18] * ogra_ pets if you set an alarm that goes off when suspended you will hear the alarm as soon as you wake up the device [14:19] ogra_: no you see the notification but not the alarm [14:19] ah [14:19] ogra_: if you wake it just before it is meant to go off it works [14:19] right [14:20] sergiusens, Hmm, this is weird. Can you copy the /usr/share/upstart/sessions/application-click.conf to ~/.config/upstart/ and then change the "exec" to a script dumping the environment? That'd tell us if it's getting to exec-line-exec at least. [14:21] ogra_: but you know my cure all for getting things fixed is I blame popey (tm) he blames this ogra_ dude and magically it works in the next image ;) === MacSlow|lunch is now known as MacSlow [14:23] tedg, sure; give me a sec on that though [14:24] sergiusens, Thanks! [14:24] asac, zsombi, thostr_: looking at the ticket now [14:24] thanks [14:24] very urgent :P [14:25] davmor2, heh [14:25] asac, so there are at least three things going on in this ticket [14:25] didrocks: gonna try and reproduce. I suspect it may be because I had a /userdata/.writable_image hanging around. trying to prove/disprove that. [14:26] charles: fyi: (03:15:10 PM) ogra_: asac, thostr_, so we nailed the issue down to "*all* events get queued when suspended" volume changed as well as playing the next song from the playlist (and i suspect also alarms) get queued up as long as the screen is off ... and get applied immediately when we wake up the device [14:26] cool [14:26] hope thats it === dandrader|afk is now known as dandrader [14:26] the issue spreads across too many channels :P [14:26] popey: and, let's hope so [14:26] popey: there is a bug for that, that I think didrocks assigned himself to that the .writable_image isn't being removed by ubuntu-device-flash [14:26] (3 at least) [14:27] thats a separate issue, but yes, i am aware of it davmor2 thanks ㋛ [14:27] davmor2, didrocks right, and I don't intend to fix that; should work just like the on image upgrader but with full images [14:33] TOPIC: qt events not getting processed when screen off - PoC: Saviq and kgunn :) [14:33] :) [14:33] ;) [14:34] thostr_, popey, asac, zsombi, ogra_: it's already a known issue that the alarms don't work correctly on suspend, we need the hardware alarm API to land in platform-api before that can be resolved [14:34] from popey's video it looks like it's working when not suspended [14:35] charles: what introduced this bug? [14:35] but wrt to music; it was working before qt 5.2 [14:35] charles, ok, so it is not related to the swallowing of events with all other Qt apps then [14:35] charles: i was told alarms were fixed even before we landed qt5.2 :) [14:35] they used to work, yeah [14:35] sergiusens, exactly, I suspect that's what can be causing it - Qt might have decided that "since I'm not rendering, I can't do anything, so why would I deliver any events" [14:35] asac, nothing's introduced this bug; it's always been there. The plan all along has been for tvoss to add hardware alarm support to platform-api s.t. indicator-datetime can wake up the phone to make snap decisions [14:35] the cases that didnt work are ok to continue not working [14:35] (as long as you were on UTC) [14:36] charles: right, but we see a new bug [14:36] afaik [14:36] sergiusens, which kind of makes sense for a UI app [14:36] that one is to be investigated [14:36] the old one we dont care for today [14:36] ogra_, they never worked, n4 just didn't go to deep sleep [14:36] * ogra_ knows that davmor2 and popey have been testing alarms before [14:36] asac, walk me through what you're seeing that's different here? [14:37] so this is not "alarms dont work when they used to" but "suspend now works where it never used to"? [14:37] charles, we see an issue where all events (vol up/down, music playback) get queued when the device is suspended and get applied as soon as you wake it up [14:37] popey: I believe that's correct, yes [14:37] ok. [14:37] charles, there was suspicion that alarms are affected by the same issue [14:38] popey, right, deep sleep was broken on the n4 [14:38] ChickenCutlass, ^ correct? [14:38] tvoss, still is afaik [14:38] ogra_: I've only tested it on active systems as I was informed that the alarms didn't activate the phone if it was in sleep, popey it seems went the extra step [14:38] tvoss, it wasnt fixed yet [14:38] unless rsalveti did a secret upload with a fix :P [14:38] tvoss, yes, deep sleep is currently not correct on the n4 [14:38] ogra_, charles, tvoss, I don't think the "all events are queued" is related to deep sleep, it probably can't go into deep sleep within the 5s that the issue still shows if you suspend after having moved to the end of a song [14:38] Indeed, on #237 (pre-5.2, stable image) I can set an alarm, and turn off the sceen. The alarm *does* sound. [14:38] ChickenCutlass, but that didnt change [14:38] right [14:39] did not change [14:39] Saviq, right, no changes wrt deep sleep [14:39] or vol up/down straight after turning the display off [14:39] it was and still is broken [14:39] so what I *think* is happening is Qt internally queues all events when it can't render, only thing that makes sense [14:39] Saviq, yeah [14:39] Saviq, yup, that makes sense [14:39] Saviq, your comment makes sense [14:39] i'm with you on this one [14:40] tvoss: ogra_: deep sleep is broken by default :-) [14:40] and won't be fixed [14:40] Saviq: yes, now you mention it I have seen the clock on the welcome screen do a "fast forward" from 12:18 to 12:40 and all minutes in between when i woke my phone earlier [14:40] this is an issue with the original kernel and image [14:40] rsalveti, thats mean ! [14:40] modem issue [14:40] yeah [14:40] just reproduced it again [14:41] popey, oh yeah, that always happens [14:41] yep [14:41] ogra_: bug 1267570 [14:41] i have only ever seen it jump [14:41] bug 1267570 in linux-mako (Ubuntu) "Mako not always entering suspend (msm_hsic_host wakelock)" [High,Confirmed] https://launchpad.net/bugs/1267570 [14:41] not seen it actually display every minute inbetween. hey ho [14:41] popey, you don't pay attention [14:41] hah [14:41] true [14:41] ooh, shiny [14:41] popey, it's worse now with deep sleep [14:41] no deep sleep [14:41] see the bug above [14:41] or whatever [14:42] popey, 'cause it would take 5s to update after a night on the nightstand [14:42] for me [14:42] hm, one thing here is that indicator-datetime isn't (yet) using the Qt stack to play sounds [14:42] Saviq, might be the nightstand [14:42] :P [14:42] Saviq, fwiw; the music problem and the EOS signal happens with the cable plugged in [14:42] same for vol up/dn [14:42] when I turn the screen on, it gets the EOS signal and goes to the next track [14:43] (see https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1283065 -- it's currently going directly to PA) [14:43] sergiusens, not for me [14:43] Ubuntu bug 1283065 in indicator-datetime (Ubuntu) "Use the existing audio stack to play event sounds" [High,Triaged] [14:43] so that might explain why popey is hearing the alarm go off even when the screen is darkened... [14:43] sergiusens, If I have USB connected -- I still do not get events [14:43] ChickenCutlass, i think thats what he said [14:43] ChickenCutlass, yeah, that's what I said; but constrained it to EOS [14:43] ogra_, sergiusens lol right [14:43] sorry [14:44] sergiusens, Saviq: confirmed I have the lead in my flo and the end of the track playing has just happened and it has stopped playing [14:44] ChickenCutlass, just trying to say I had an adb shell tail -f running [14:44] right [14:44] so deep sleep seems rather impossible [14:44] it is not a deep sleep issue [14:45] exactly [14:45] * sergiusens tries to clarify and rule out, seems to confuse [14:45] :) [14:45] sergiusens, AFAICT unrelated to any cables :) [14:46] ogra_: we can only remove 'ubuntuappmanager' once we officially drop support for SF [14:46] rsalveti, today ? [14:46] :P [14:46] i think the only bit that held us back was the missing screenshot capability ... which now works [14:47] oh [14:47] and Kaleo :P [14:47] rsalveti, ignore me, i totally forgot he wants to do performance tests [14:47] hehe [14:48] ok, so... /me grabs a patch from Daniel that disables the "stop rendering" in mir [14:48] oh ! we have such a thing ? [14:48] awesome [14:51] sergiusens: didn't Saviq just reproduce the issue when forcing the screen to be on with powerd? [14:52] rsalveti, no [14:52] rsalveti, active != display on [14:52] * Saviq tries again [14:52] Saviq: so if you force the screen to be on, or disable powerd, it works? [14:52] just want to make sure it's really related with screen being off [14:52] rsalveti, verifying again, but last I tried, yes - even if greeter is on screen [14:53] rsalveti, yes only when screen is off [14:53] charles: alarm was reported to work with screen off before, now its broken [14:53] i cant walk you though. i am just here to facilityate and connect people [14:53] bug 1295170 ☹ [14:53] asac: it might be related with a suspended device [14:53] bug 1295170 in unity8 (Ubuntu) "unity8 in restart loop on flo" [Undecided,New] https://launchpad.net/bugs/1295170 [14:53] charles: just look around whoelse is active here and engage. we all here work on the promotion issues [14:53] which might all be related [14:54] I mean, there is a wakelock that is used by the modem, that keeps the device not suspending from time to time [14:54] rsalveti, so yeah, as long as you keep the screen on (even on greeter) [14:54] rsalveti, it works fine [14:54] right, screen on processes all events ... screen off queues them [14:54] rsalveti, there's a wakelock requested over dbus for music that is never released as well [14:54] screen on again processes the queue [14:54] awesome, so it's indeed screen related [14:55] not suspend and such [14:55] nope [14:55] nope [14:55] rsalveti, right [14:55] sergiusens: right, that could help as well [14:55] half of the work is done now that we convinced rsalveti :-) [14:55] ok, so dandrader's code won't help, he just didn't SIGSTOP the apps [14:55] \o/ [14:55] lol [14:55] heh [14:55] * ogra_ makes some meeting coffee [14:57] sergiusens: just making sure because people were running like crazy spending time on this as well [14:57] kgunn, ok, can we get a mir patch that would allow apps to render even though screen is off? [14:57] sergiusens, I think rsalveti just does not beleive us [14:57] lol [14:57] Saviq: maybe... [14:57] true [14:58] kgunn, would help greatly with testing the theory [14:58] ChickenCutlass, pfft ... the beachboy ... [14:58] Saviq: because if the hw doesn't flip then we'd queue the buffers and be stuck... [14:58] depends a little on the hw ability to consume [14:58] kgunn, or discard them? [14:58] yeah...we can hack it somehow [14:58] kgunn, the usual frame drops? [14:58] you don't care about artefacts right ? [14:58] kgunn, nah [14:59] wonder when this started to happen [14:59] AlbertA: ^ could we just ignore screen state ? [14:59] kgunn, I just want to test the theory that indeed stopping the rendering messes with the Qt event system [14:59] hack for testing [14:59] probably part of mir that landed right before qt [14:59] rsalveti, 5,2 [14:59] well, around that time at least [14:59] rsalveti, yeah, another thing that makes me think it's Qt → this must've happened with Qt 5.2 [14:59] ogra_: not sure if only qt related, could be qt + mir or just mir [15:00] rsalveti, right, we landed to much around the same time (again) [15:00] which one's the first Qt 5.2 image again? [15:00] but we got an image before doing the qt switch [15:00] Saviq: so do you actually want 2 test patches ??....e.g. one that allows renders while screen off...and then a toggle to "stop rendering" in mir/block basically while screen is on ? [15:00] AlbertA: ^ [15:00] kgunn, that could work, too [15:01] Saviq, http://people.canonical.com/~ogra/touch-image-stats/238.changes [15:01] kgunn, as long as I can make mir block the rendering without SIGSTOPing the app [15:01] ogra_, ! [15:01] ok so no mir upgrade there [15:01] Saviq, so 237 was the last non 5.2 one [15:01] * Saviq flashes238 [15:01] AlbertA: i was about to say as quick and dirty as you like...but not with SIGSTOP on apps :) [15:01] he already had that patch :) [15:01] yeah [15:01] let me flash 237 [15:02] 243 had a Mir update then [15:02] Saviq, I confirm that 238 is the image where bug 1292306 appeared, worked fine on 237 [15:02] bug 1292306 in Ubuntu Music App "Upon upgrading to Qt5.2 the music app no longer plays the next song if the screen is off" [Critical,Triaged] https://launchpad.net/bugs/1292306 [15:02] jibel, right, so Qt [15:03] and 246 had another Mir update [15:04] * Saviq uses ondra's WiP way to flash dual-boot without downloading on the device :> [15:04] elopio, maybe you can help me with a testing problem with my split greeter branches... How would I propose a branch to autopilot-testrunner-touch-saucy that would land in conjunction with a code change? Can I just add it to the silo? [15:05] awe, sorry I had to leave early, I had another call. One more question: Is it ever the case that when you use the same APN for Internet and for MMS, you use different user name and/or password for Internet and for MMS? [15:05] mterry: I have never touched autopilot-testrunner-touch-saucy. But I have seen people adding to the silos all the changes that need to be landed at the same time, even if they are from different projects. [15:06] mterry: the ci vanguard should help you with that. [15:06] mterry, there's no way to do that atm :| [15:06] elopio, good point [15:06] Saviq, bummer :( [15:07] Saviq, ah... but I bet I can make my change work with pre and post split code.... [15:07] mterry, only way is to actually flash a device and run the tests that you wanted to run after dist-upgrading from a silo [15:07] Saviq: because is on autopilot-testrunner-touch-saucy, or because we can't land two projects in conjunction? [15:07] elopio, because the job doesn't know how to use PPAs [15:07] popey: sergiusens: on the upgrade issue, do we do one or two bugs? [15:08] Saviq: ah, but the autopilot guys added a script to the jenkins job that added the ppa and did the dist-upgrade [15:08] didrocks: i have one. [15:08] I'm not sure if that's doable for all the jobs. [15:08] popey: please send my way [15:08] elopio, I'm not saying it's difficult, I'm just saying it's not there tet [15:08] yet [15:08] mterry, btw, whenever you're ready for a silo, let me know [15:08] didrocks: bug 1295170 [15:08] bug 1295170 in unity8 (Ubuntu) "unity8 in restart loop on flo" [Undecided,New] https://launchpad.net/bugs/1295170 [15:08] popey: you have that on flo right? [15:08] yes [15:09] I understand [15:09] popey: so why didn't I [15:09] only triggers once you enable wifi (I suspect it's the service going off to get media for all the music) [15:09] look at the very explicit instructions [15:09] mterry, we can now have multiple silos for a single project, only we need to make sure that we reconcile after one of them lands [15:09] about connecting to wifi and music [15:09] ah I enabled wifi in the second flash not the first let me try that again [15:09] sergiusens: do you have a bug for your issue maybe? [15:09] sergiusens: upgrade one [15:09] as it seems quite separate [15:09] Saviq, I suppose I'm ready for a silo [15:10] Saviq, all the branches are listed in https://code.launchpad.net/~mterry/unity8/split/+merge/210664 [15:10] Saviq, though... [15:10] popey, please unity8.log to bug #1295170 [15:10] bug 1295170 in unity8 (Ubuntu) "unity8 in restart loop on flo" [Undecided,New] https://launchpad.net/bugs/1295170 [15:10] Saviq, did I hear something about how a branch can't be in two silos? [15:10] Saviq: k [15:10] mterry, it can [15:10] mterry, only when it lands in one [15:10] Saviq, ok, great [15:10] Saviq, I expect some of these branches I need will land separately [15:11] mterry, sure, that's fine [15:11] mterry, we'll remove them from the silo then [15:11] mterry, and resync stuff with trunks [15:11] Saviq, so all the branches listed in that merge, plus the merge itself is needed (lp:~mterry/unity8/split) [15:11] mterry, ok, will let you know [15:12] popey, and there's no unity8 .crash is there? [15:12] Saviq, for lightdm 1.9.13, I guess just do lp:lightdm [15:12] Saviq: no [15:12] mterry, does it need a rebuild? [15:12] Saviq: uploaded unity8.log [15:12] popey, thanks! [15:12] Saviq, does what? [15:13] mterry, does lightdm need a rebuild? or is it just not yet released? [15:13] Saviq, just not yet released [15:13] Saviq, branch got merged to trunk, but not in trusty [15:13] mterry, please do a temporary no-change MP [15:13] Saviq, for silo purposes? [15:13] mterry, yeah, they can only deal with MP [15:13] ok [15:14] mterry, or well, we could dput, but same difference [15:14] mterry, can be set to WiP straight away or something [15:14] mterry, silos don't care [15:14] didrocks, sorry - I was not around... what about the porting guide? [15:15] :'( ondra's way doesn't boot [15:15] dholbach: I don't have the trace and can't find it easily, when I poke you, someone wanted to have porting info [15:16] Saviq, https://code.launchpad.net/~mterry/lightdm/fake-no-change-mp/+merge/211969 [15:16] didrocks, yeah, I'm not an expert - I never ported anything myself - I just helped put the docs up and advertise them [15:16] yeah, not sure if the doc is up to date though [15:17] didrocks, we always said "yeah, we need to rework some bits again", but probably never got around (if the comment was about some information missing?) [15:17] ok [15:17] mterry, thanks [15:18] mterry, that not released yet either https://code.launchpad.net/~robertcarr/mir/remove-ensure-display-powered/+merge/209734 ? [15:18] Saviq, right [15:18] * Saviq wonders how silos will deal with that [15:19] didrocks, I wanted to prepare a silo with ↑ for testing the split greeter, but that branch is against lp:mir/devel, will silos cope with that? [15:20] Saviq, you can just do mir/devel then [15:20] Saviq, oh but that's not an MP [15:20] mterry, yeah, that MP will be fine I think [15:20] mterry, or well, we might need another no-change against devel [15:20] boiko, what's the plan for a release of telephony-service with those AccountsService changes? [15:21] Saviq, you don't think silos like merged branches? [15:21] mterry, not sure, we'll find out soon enough :) [15:22] tedg, I assume you saw the comments in your indicator-sound FFe? [15:24] Saviq: the branch should be against trunk? I don't think we want to ship devel? [15:24] didrocks, it will get into trunk on its own [15:24] didrocks, we'll strip that silo a lot before landing [15:24] didrocks, we just wanted to prepare for it === alan_g is now known as alan_g|tea [15:25] didrocks, but yeah, I'm asking because it's against devel, we don't plan on landing it [15:25] Saviq: not sure you'll want a silo then [15:25] didrocks, dholbach, the porting guide is really behind ... but luckily we have some poeple doing ports and reporting to the ML, once these ports are successfull we should go over the thread and update the wiki accordingly [15:25] didrocks, but silos are so nice! [15:25] Saviq: yeah, but it's not the airline :) [15:25] popey: I can't replicate your issue and I followed the steps in the bug and everything :D [15:25] didrocks, what would be the issue? [15:26] there's probably one interesting question... ABI breakage in mir devel [15:26] mterry, Yeah, working on it. [15:26] ogra_, yeah, that'd be great - or the folks who get good advice can go and update the guide? [15:26] Saviq: blocking, need everyone to set "force" on that components and you will stuck a silo where we are already low on resources [15:26] we'd probably need to rebuild qtubuntu, platform-api and unity-mir at least [15:26] didrocks: I can't replicate what popey is seeing at all [15:26] davmor2: music app is starting? [15:26] as well? [15:26] * mterry hugs tedg [15:26] didrocks, :| [15:26] davmor2: i did it twice ☻ [15:26] didrocks, yeah I get what you mean [15:27] dholbach, yeah, prob is that porting isnt usually done in one shot so telling the good from the bad advise is needed [15:27] mterry, it might be too early for a silo then [15:27] Saviq: you really want the airline which is to address that :) [15:27] didrocks, of course I do [15:27] didrocks: music starts which is sergiusens issue blown out the water [15:27] if they just dump in everything the guide will be mess again [15:27] didrocks, thought locking is off, though... I could have two unity8 silos with nothing special? [15:27] davmor2: I just reflashed #137 [15:27] davmor2: let me upgrade [15:27] didrocks: and popey 's restarting loop of unity8 I can't reproduce [15:27] tedg, seems my changes aren't taking effect; even tried replacing the original conf [15:28] Saviq: it's an exceptional case as told [15:28] didrocks, ok [15:28] Saviq: and the consequence is that we are low on silos now :p [15:28] * mterry reads back [15:28] tedg, init-checkconfig says conf is ok and I did reload the configurarion [15:28] ogra_, right [15:28] sergiusens, Try putting something in the pre-start? Perhaps that is failing? [15:28] didrocks, it's all juju-ified, isn't it? just crank the dial up! ;D [15:28] Saviq: ahah, not the LP-side though :p [15:29] mterry, so yeah, a little bit early for a silo [15:29] Saviq: you are building the Mir thingy to hope it will help the music one? [15:29] didrocks, although I'd say the split greeter would be a big transition enough to warrant the exception [15:29] Saviq, because of the Mir issue? I'll have to keep poking kgunn then :) [15:29] sergiusens: of course we are all fools for following your advice for the upgrade simulation ubuntu-device-flash --channel=devel-proposed --revision=237 --wipe would be the way to do it then you can actually upgrade :) [15:29] Saviq: yeah, but clearly not until we can promote an image and get to a normal silo level [15:30] mterry, not only that [15:30] tedg, was pre start [15:30] didrocks, yes, understand [15:30] sergiusens, Did you try the aa-exec-click -p thing? I'm curious if the apparmor profile is failing to load? (that happens between pre-start and and the job itself) [15:30] Saviq: but agreed then [15:30] didrocks, waiting for a patch from AlbertA, what Daniel had was only dealing with SIGSTOP/SIGCONT [15:30] Saviq: ok, but we all agree that the alarm one is different? === chihchun is now known as chihchun_afk [15:30] tedg, not yet; happens next [15:30] and it's on thostr_'s team [15:30] didrocks, ENOCOMMENT, dunno the alarm one [15:31] didrocks, but people said so [15:31] Saviq: lucky you :p [15:31] tedg, http://paste.ubuntu.com/7125780/ that's with env from pre start [15:31] Saviq: kgunn is off today? [15:31] he was around [15:31] ok [15:31] * didrocks invites him to the meeting [15:32] didrocks: which one in particular [15:32] davmor2: I'm upgrading to latest proposed image as we speak === chihchun_afk is now known as chihchun === alan_g|tea is now known as alan_g [15:32] from latest promoted [15:32] sergiusens, Can you throw an "echo foo" after the click-exec just to make sure it happens? [15:32] Mirv: you can add the patch to our Qt, i have a patch+test for upstream (that is slighlity different since they are only accepting 5.3 now) https://codereview.qt-project.org/#change,81454 [15:33] and see if I can reproduce popey's/sergiusens issues [15:33] didrocks, he lost connection like a quarter ago or something [15:33] Saviq: ok :) [15:33] sergiusens, Just trying to clear up possibilities. [15:33] thanks [15:33] hello i am new to development world and i would like to port ubuntu touch on my SONY XPERIA P can someone guide me. [15:34] tedg, wrapped in prefoo and postfoo http://paste.ubuntu.com/7125796/ [15:35] sergiusens, So, no postfoo? === gatox is now known as gatox_lunch [15:35] nope [15:36] Hmm, there's no error handling between that last print out and the "return 0" [15:36] sergiusens, No crash file for UAL, right? [15:37] nope [15:37] sergiusens, Did you make it so it wasn't "exec click-exec" just a call? [15:38] tedg, the pre-start script has exec originally [15:39] sergiusens, Yes, just checking that you removed it. Because the exec there would make postfoo not print. [15:39] didrocks, so are you experiencing the alarm bug that's being discussed this morning? [15:39] right [15:39] charles: popey and davmor2 are, right [15:41] tedg http://paste.ubuntu.com/7125821/ ... [15:41] * sergiusens goes back to the upstart handbook [15:41] davmor2, didn't make sense of what you just said [15:42] sergiusens, Heh. And did you put a prebar and postbar around exec-line-exec? [15:42] charles: you can't reproduce? [15:43] sergiusens: I can start music app [15:43] charles, didrocks: I didn't follow up, too much IRC channels and a meeting, but didn't Saviq said that the issue for alarm/music was down to Mir/events being queued/blocked? [15:43] quite slow to start [15:43] sergiusens: so if we reflash with --revision=237 (last promoted) we can do a real upgrade to 248 without having to reflash [15:43] seb128: issue for music yeah [15:43] davmor2, oh, well devel-proposed only keeps the last 10 images; my comment was a generic one; you are just lucky it was rolled out [15:43] didrocks, tbh I'm still not sure what the new issue is, the bug reported is about indicator-datetime not showing snap decisions while the screen is off [15:43] seb128: alarm doesn't use Qt [15:43] charles: what is playing the alarm music, do you know? [15:43] didrocks, ^ that's qt :p [15:43] seb128: charles told the indicator isn't using Qt [15:44] didrocks, right now it's pulseaudio [15:44] charles: what is asking pulseaudio to play a music? [15:44] didrocks, (I was talking about notification, not the sound) === vying is now known as Guest59247 [15:45] (didrocks, seb128, there's a bug ticket to switch from PA to use ricmm's upcoming media API for that, bug #1283065) [15:45] bug 1283065 in indicator-datetime (Ubuntu) "Use the existing audio stack to play event sounds" [High,Triaged] https://launchpad.net/bugs/1283065 [15:45] tedg, I have another env there so it should of reprinted [15:45] charles: sure, but we are talking about the present :) [15:45] :) [15:45] charles: well, the existing audio stack is PA [15:45] so, indicator-datetime is triggering the alarm sound? [15:45] didrocks, indicator-datetime is what tells PA to play the alarm and show the snap [15:45] didrocks: what is the issue? (alarm) [15:45] if do we have a bug? [15:45] rsalveti, that's what we're trying to figure out [15:46] rsalveti: sure https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295122 [15:46] Ubuntu bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Critical,New] [15:46] sergiusens, Can you pastebin the job, I'm getting myself confused :-) [15:46] charles: the bug was pointed to you by thostr_ an hour ago IIRC :) [15:46] didrocks, yes I know. We discussed it then but the discussion petered out and I had a pair of hangouts [15:47] so I'm back and continuing the conversation, trying to figure out what actually is new in 1295122, if anything at all :-) [15:47] wonder why image # were not set though [15:47] so there's an existing known issue about not working when the phone is suspended, we need for hardware alarms to land in platform-api before indicator-datetime can wake the phone up to show snap decisions then [15:47] previously it seemed to work because the phone wasn't actually suspending... [15:47] tedg, here, let's make it atomic: http://paste.ubuntu.com/7125841/ and http://paste.ubuntu.com/7125847/ [15:47] charles: how come was it working pre-Qt 5.2 then? [15:48] charles: on #237? [15:48] didrocks: previously it seemed to work because the phone wasn't actually suspending... [15:48] rsalveti: but it seems you punned that one off, right? ^ [15:48] mterry, we're getting a silo after all ;D [15:48] * didrocks seems to be back one hour earlier [15:48] :) [15:49] mterry, only we need not to complain if they flush us out of it when needed [15:49] didrocks: no, it just suspends correctly sometimes :-) [15:49] rsalveti: hum, but what changed? seems to be reliably working on #237 [15:49] and not on #238 [15:49] davmor2: popey: can you guys try to reproduce bug 1231455 after running 'powerd-cli active' on another terminal? [15:49] bug 1231455 in powerd "powerd-cli help could do with some cleaning up and fixing" [Undecided,In progress] https://launchpad.net/bugs/1231455 [15:50] jibel: davmor2: popey: right? ^ [15:50] sergiusens, Cool, great idea! That clears things up. So it seems to not be getting to the exec section of the job. [15:50] didrocks: nothing changed, but the wakelock is usually not around when the antenna signal is strong enough for the modem to be quiet [15:51] rsalveti: you sure about that bug number? [15:51] Popey: See's the alarm notification but not hears no sound. [15:51] this is a different issue [15:51] popey: no, that's wrong [15:51] lol [15:51] popey: davmor2: but 1295122 [15:51] bug 1295122 [15:52] bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Critical,New] https://launchpad.net/bugs/1295122 === chihchun is now known as chihchun_afk [15:52] * rsalveti needs coffee [15:52] I was beginning to wonder then popey beat me to the ask [15:52] being able to see the alarm and not hearing is an interesting issue [15:53] sergiusens, Can you dump the output of this while starting the app please? dbus-monitor --session interface=com.ubuntu.Upstart0_6.Job === chihchun_afk is now known as chihchun [15:53] Saviq, OK :) [15:53] rsalveti: am I running that as root via adb or via sudo -u phablet -i in adb? [15:53] mterry: saw you talkin' bout silo in scrollback... [15:53] sergiusens: assigning you with popey for the upgrade issue to get more investigation, neither davmor2 or I can reproduce it [15:54] you need any help ? (leave Saviq free for real work :) [15:54] davmor2: as root [15:54] rsalveti: does that stop it blank? [15:54] kgunn, sure you could help set it up instead of Saviq [15:54] rsalveti: roger wilko [15:54] kgunn, it's there in row 42 [15:54] popey: no, just stop suspending [15:54] k [15:54] Saviq: mterry....glad i could help ;D [15:54] kgunn, sil2100 and didrocks decided to give us the silo with the caveat that we might lose it if it gets tight [15:56] kgunn, so the row is good, only potential issue is if mir broke ABI between trunk and devel (or that the branch does), in which case we'll need to rebuild papi, qtubuntu and unity-mir [15:56] and whatever else you rebuild on Mir ABI break, but that you'll know better [15:56] rsalveti: it's the same, i mean, with the screen blanked (I tapped power) I see nothing until I wake the screen [15:56] popey: davmor2: rsalveti: screen blanked works here [15:56] Saviq: is that landing row good MR-wise? [15:56] rsalveti: anyway to force the suspend? [15:56] sil2100, yes [15:57] didrocks: nops [15:57] anyway to turn off the alarm as well? [15:57] rsalveti: and it rings [15:57] Anyone know the status of the QtQuick.Controls.MenuBar widget? It's not working for me. [15:58] hm, confused haha [15:58] sergiusens, Actually this might be more informative: dbus-monitor --session interface=com.ubuntu.Upstart0_6,member=EventEmitted,arg0=stopped [15:58] rsalveti: so, to have is suspended for sure, need to unplug the cable and wait for x minutes ? [15:58] popey: so broken the same way for you [15:58] and working fine for didrocks and davmor2 [15:58] left comment on bug [15:58] didrocks: in theory yes, but with mako that is not always true as the modem might be holding a wake_lock [15:58] didrocks: testing with flo is better in this case [15:58] rsalveti: I only have mako [15:59] i have flo [15:59] mterry: Saviq...yeah, that ought to be ok...that mp doesn't break abi & will use trusty branch as base [15:59] rsalveti: I have mako/flo/manta [15:59] should be fine [15:59] kgunn, will not [15:59] now, if only unity didnt die on me on flo [15:59] kgunn, it's gonna use devel as base [15:59] davmor2: but you said it worked for you on mako, right? [15:59] kgunn, that branch actually does delete a public method [15:59] not sure why it would fail for popey [15:59] are you guys using the same image? [15:59] Saviq: oh my bad... [15:59] kgunn, since it's directed to devel, it will build devel [16:00] -rw-rw-rw- 1 root root 0 Mar 20 12:45 .writable_image [16:00] ffs [16:00] * popey reflashes his device clean [16:00] didrocks: but we know it'll be broken if the device is properly suspended [16:00] kgunn, when is 0.1.18 planned? [16:00] what's the preferred way to reboot the device with ubuntu touch? [16:01] that is a known issue and not a regression [16:01] s/the/a/ [16:01] rsalveti: ah hang on retried and now it seems to of trigger I hit dismiss and it's carried on [16:01] just hold the power buttong? [16:01] feels a bit rough ;D [16:01] didrocks: the possible regressing is alarm not working when the device is not suspended [16:01] mterry: let me add 2 mp's [16:01] rsalveti: well, it's an user regression (from his point of view) [16:01] rsalveti: and seeing how hard the line management is taking, I think they are going to block any new promotion [16:01] * davmor2 throws the phone out of the window to stop it ringing [16:02] didrocks: right, but nothing can be done to get that fixed, only when we land hw alarms [16:02] didrocks: it's not a regression [16:02] davmor2, popey, https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295122/comments/6 [16:02] Ubuntu bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Critical,New] [16:02] didrocks: all you need to do is get low signal so the modem can take a wakelock [16:02] + rsalveti, didrocks, ^ [16:02] rsalveti: the second test should of gone off at 16:00:00 it went off at 16:00:32 and then refused to stop ringing [16:02] rsalveti: well, it's a "4.4 regression" actually as the device sleep is fixed [16:03] rsalveti: just talking about user's view [16:03] it's not :-) [16:03] it was always broken [16:03] it works sometimes [16:03] ah, so deep sleep isn't something fixed in 4.4 [16:03] ok [16:03] no, broken the same way [16:03] not sure why it was so reliably working/not working for popey and davmor2 [16:03] tedg, http://paste.ubuntu.com/7125933/ [16:03] maybe with usb cable connected [16:03] charging [16:03] that's plausible [16:04] caused by the music-app bug (not removing the suspend blocker) [16:04] ok [16:04] popey: davmor2: tell us about the second test, otherwise, happy to remove it from the list [16:04] but if the alarm is broken when the device is not suspended, then we have a bug [16:04] based on what rsalveti told [16:04] rsalveti: tried 3 times already [16:04] here [16:05] and all ok [16:05] is that the second test in popey's video? [16:05] i am re-flashing my device cleanly now [16:05] its possible there was crap on it, I don't know. [16:05] davmor2: going off before the desired time is a bug for sure [16:05] yeah, trying reproducing after a clean flash and clean boot [16:06] reflashing here === chriadam is now known as chriadam|away [16:06] (...we really need a nicer alarm tone :) [16:06] it's all clean here (well, after an upgrade from a clean last promoted image) [16:06] rsalveti, fwiw the suspend blocker is a powerd bug ;-) [16:06] sergiusens: right :-) [16:06] rsalveti: that is after the desired time by 32 seconds. but the issues is the fact that I hit dismiss which stopped it the first time and didn't the second I had to power down the device to make it stop ringing [16:06] jdstrand, Is there a way that we could load the music app profile to see if it's failing somehow? [16:07] jdstrand, If I just called "aa-click-exec -p music ls" would that work? [16:07] davmor2: yeah, that might even be a different bug [16:07] let's see if charles is able to reproduce that [16:08] jdstrand, We seem to be getting out of pre-start but not into the main job, and the only thing I can think of that happens there is switching to the apparmor profile. [16:08] always keep 'powerd-cli active' running when testing the alarms [16:08] so we know it's not suspend related [16:08] also flashing latest [16:08] rsalveti: I'm just now starting the flash, so this may take a bit ;) [16:09] jdstrand, So looking for an independent way that I can ask sergiusens to verify that the apparmor part isn't failing. [16:09] rsalveti, didrocks: right I'm going to start an album playing let the device sleep with the powerd-cli active on and see if we get to track 2 [16:09] davmor2, could you walk me through the steps to try & reproduce the wont-stop-ringing issue you saw [16:09] * sergiusens checks /var/lib/apparmor [16:09] davmor2: Saviq tried that already, and failed the same way [16:09] davmor2: the music-app issue happens when the screen is off, not necessarily suspended [16:10] rsalveti: :( [16:10] tedg, jdstrand think I found the issue :-/ [16:10] tedg, phablet@ubuntu-phablet:/var/lib/apparmor$ ls clicks/|grep music [16:10] com.ubuntu.music_music_1.1.359.json [16:10] wrong version [16:11] Oh, heh. Computers are such sticklers about numbers. [16:12] mterry: unfortunately landings are blocked until an image is promoted [16:12] charles: the time was 15:55 so I set the first alarm for 15:57 and hit the power button to stop the screen, alarm went off fine and I was able to wake the phone and hit dismiss and it did, I then tried again for 16:00 this time the alarm went off 32 seconds late and wouldn't stop ringing [16:12] davmor2, the alarm does wake the device up, though, right? [16:12] rsalveti: did miss one with mako this time, so yeah, probably suspended [16:13] boiko, makes sense [16:13] didrocks: right [16:13] davmor2, so it will light up the screen, so the "screen off" issue shouldn't matter [16:13] hopefully we'll be able to land hw alarms next week anyway [16:13] Saviq: this was with powerd-cli active in place the alarm went off both times correctly just wouldn't stop ringing [16:13] sergiusens, So I'm gonna assume I'm out of this one now then. I think it's cjwatson vs. jdstrand for this bug now :-) [16:13] davmor2, ah ok, so different [16:14] Saviq: alarm didn't wake the screen I had to [16:14] tedg, would be hard to now has this happened or why it hasn't regenerated (relinked) [16:14] davmor2, hmm not sure what the UX design is, but I imagine it should [16:14] tedg, thanks [16:14] I don't think we have anyone waking up the screen reliably though [16:14] Saviq: but that is a known bug [16:14] hard to know* [16:14] Saviq: https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295122/comments/6 [16:15] Ubuntu bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Critical,New] [16:15] sergiusens, Yeah, I'm guessing it's either a click hook didn't run or when it ran it didn't generate it correctly. [16:15] sergiusens, I guess if there's no symlink, it's probably a click thing. [16:15] davmor2, oh ok [16:15] tedg, there is a symlink, with an incorrect version :-P [16:16] awe, another question: About how long does it take for ofono to tell whether manually entered APN settings work? One second? Five? Ten? [16:16] right, so nobody should be waking up the screen atm [16:16] and that is expected [16:16] davmor2, right, the screen will wake up but powerd hasn't done that yet because iiuc it's waiting on hw alarms in platform-api [16:16] but the sound should still play [16:16] charles: sound not playing when device is not suspended is a bug [16:17] rsalveti, agreed. I'm trying to reproduce that [16:17] and not being able to stop it is another bug [16:17] tedg, seems it happens since I installed a click for that package that doesn't belong to the normal in image clicks (as an update or out of band install); seems similar to what kenvandine saw yesterday [16:17] who wants to stop alarms anyway :P [16:17] rsalveti: davmor2: powerd-cli active [16:18] davmor2: do you get the 'unable to stop the alarm sound' issue when forcing 'powerd-cli display on bright'? [16:18] rsalveti: davmor2: no ringing here :/ [16:18] (we could just call it a feature ... it reminds the whole day you that you didnt forget to set the alarm) [16:18] sergiusens, i rebuilt my click and it worked [16:18] didrocks: right, that's a bug [16:18] no changes [16:18] not sure what was up with that [16:18] charles: ^ [16:18] charles: so no deep sleep [16:18] and no alarm either [16:19] wonder if pulse is smart enough to not play when the screen is off? [16:19] * ogra_ thought it was clearified 2h ago by rsalveti that we have no deep sleep on mako [16:19] rsalveti: worked once though [16:19] right [16:19] with screen off [16:19] (why is this still discussed, we need to improve our communication seriously [16:19] ) [16:19] ogra_, It's because everyone was in a deep sleep and didn't get the message. [16:20] ogra_: we have, that's the issue, but not always (deep sleep on mako) [16:20] ogra_: well, rsalveti told the contrary here (I thought we had no deep sleep, but seems we can have it) [16:20] ogra_: you need to backlog even more! :) [16:20] ah, k [16:20] seems ted is right then ... [16:20] (wrt /me) [16:21] davmor2, with the second alarm that went off 32 second late -- did you get a snap decision when it went off? [16:21] ara: Hi hi. I'm working on chromium-browser touch support. You mentioned that you have a touch-pad and you tried two-finger pinch, and it didn't do what you expect. Does *any* app you try understand two-finger pinch? [16:21] charles: see https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295122/comments/7 [16:21] charles: Yes I hit dismiss it just didn't [16:21] Ubuntu bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Critical,New] === gatox_lunch is now known as gatox [16:22] qengho: hello, I would need to try, do you have any recommendation of an app that should understand 2-finger pinch? [16:23] ara: web browser [16:23] ara: the OSM (webapp) understands 2 finger pinch [16:24] popey: is it available for the desktop? [16:24] ara: ya, apt-get install webbrowser-app [16:25] popey: Thanks! When you say it should, have you tried it on something not a touch-screen, like a touch-pad? [16:25] nope [16:26] i switch touch-pads off on my laptops ☻ [16:26] hm, can't get alarm to work [16:26] interesting that in the upcoming indicator the alarm is set with UTC [16:26] charles, rsalveti: so with powerd-cli display on bright it is popping up the alarm snap decision and dismissing correctly each time it has worked 3 times now [16:26] davmor2: how are you doing that? [16:27] rsalveti: hum [16:27] qengho: I need to reboot, will get back to you in a second [16:27] rsalveti: It didn't ring even screen on [16:27] rsalveti: I wonder if the issue isn't rather "first alarm(s?) work" [16:27] then none [16:27] popey: adb shell and then type in powerd-cli display on bright the nset the alarm etc [16:28] argh, why I can't get alarm to work [16:28] ok [16:28] following didrocks' suggestion of adb shell + powerd-cli active + manually turning off the screen, I get a ringtone at the right time, and then when I turn on the screen I see the snap decision & am able to stop the ring by dismissing the snap [16:28] rsalveti: that powerd-cli command though is opening directly to the clock-app/apps scope bypassing the welcome screen is that what is meant to happen [16:28] charles: did you try to setup another alarm then? [16:28] having trouble editing alarms in the clock app, but that's a different issue... [16:28] charles: even screen on now I don't get any new alarm [16:28] didrocks: each time I've done it, it's been creating a new alarm from scratch [16:28] how can I remove an alarm? [16:29] charles: do you delete old alarms? [16:29] didrocks: if the recipe involves editing an existing alarm, I'm not able to edi tthem [16:29] rsalveti: I tried rebooting just in case [16:29] then, I'll go on removing alarms [16:29] if I can :) [16:29] rsalveti: in the clock app, slide them to the side and you'll get a trashcan button [16:29] charles: no, I keep appending [16:29] oh, awesome [16:29] charles: yeah you just have to delete the old alarm [16:29] didrocks: I deleted and repeated. I'll try appending instead [16:29] I wonder if it can be a ETOOMANYALARMS [16:29] then [16:29] lol, app is frozen now after removing second alarm [16:30] apport is on fire [16:30] gah! [16:30] lol [16:30] _usr_lib_arm-linux-gnueabihf_qt5_bin_qmlscene.32011.crash [16:30] yay [16:30] * charles blames it all on ubuntu-clock-app, then runs away [16:30] freshly flashed my device with 248. keeping display on i can see the popup but hear no sound. [16:31] ok, it's ringing [16:31] bzoltan: is it possible to tell qtc to look in a subdirectory of a project for the click and apparmor manifest files? [16:31] after a reboot [16:31] instead of hte project rot [16:31] rsalveti: My apport is on fire [16:31] * popey reboots [16:31] let me try again to add more [16:31] lol [16:31] it's like blind men feeling the elephant [16:32] qengho: confirmed that 2-finger pinch does not work either in webbrowser-app [16:32] so it's not the number of alarms [16:32] ara, All, I'm glad I'm in this channel, so people can tell me I'm full of nonsense. I don't think any touch-pads will support two-finger pinch in ubuntu. The underlying layers take two-finger gestures and make them into motion or scroll events, not touch events. It's an ancient hack to support apps that didn't understand 2-finger scrolls in early 2000s. So, terminal gets a PageUp/Down. At three fingers, it flips to interpretation to touch events. [16:32] ok, 2nd alarm since reboot rings as well [16:32] I have played with it, and I'm no expert, but I don't see a way to get pinch info from a touchpad. [16:33] 3rd works… [16:33] Not until xinput2 get rewritten. [16:33] if the 4th works, I have no idea… [16:33] so, appending a new alarm (instead of deleting the old one) seems to not affect anything, getting the same behavior with "powerd-cli active" + screen off that I saw before [16:33] ara_: touch*screens* should work. They don't have the historical baggage. [16:34] charles: even screen on, I had nothing after a while [16:34] after reboot alarm works. [16:34] 4th worked :( [16:34] I don't know… [16:34] I tried dismiss/ok as well [16:34] qengho: do you happen to have a bug around we can reference? [16:35] ok, trying to get a low power now [16:35] and let's see if it breaks alarm forever [16:35] (until you reboot) [16:35] ara_: Not yet. I'm just discovering it. I'm hoping to be told I'm wrong. [16:35] popey, you're able to reproduce the no-sound issue? [16:35] oh, not after reboot. damn :/ [16:36] charles: on first boot, yes [16:36] after reboot, sound works [16:36] charles: I'm at the continued ringing after hitting dismiss on powerd-cli active what info is useful? [16:37] qengho, phones with touch pads ? [16:37] brb, need to grab some food [16:37] ogra_: I'm not here to talk about phones uniquely. [16:37] qengho, well, but this is the channel for them :) [16:37] Sorry. [16:37] qengho: maybe we should ask in #ubuntu-x [16:38] (teh channel name is misguiding) [16:38] Good idea. [16:38] yeah [16:38] * ogra_ was about to suggest that [16:38] * qengho leaves. [16:38] charles: it went off at 16:35:26 rather than 16:35:00 [16:38] stay :) [16:38] we like you around [16:38] * qengho hugs ogra_. [16:38] * ogra_ hugs qengho [16:38] charles: only seems to keep ringing if it is late [16:38] (and ara_ ) [16:39] davmor2, and you're getting the snap decision late too? [16:39] ie, the popup shows up at 16:35:26 rather than 16:35:00, just as the ring does? [16:39] charles: pass I don't unblank the screen till the alarm sound === chihchun is now known as chihchun_afk [16:41] charles: rsalveti bah, passing everytime on that one… [16:42] charles: I do confirm though that waiting for too long doesn't enable us to dismiss the ringing [16:42] davmor2: opening a bug on that? [16:42] didrocks: will do [16:43] davmor2: before the meeting please :) [16:43] charles: can I shutdown the phone the ringing is driving me crazy now if there are no logs you need [16:43] davmor2, try this for me please [16:44] ssh in as phablet [16:44] $ stop indicator-datetime [16:44] $ G_MESSAGES_DEBUG=all /usr/lib/arm-linux-gnueabihf/indicator-datetime/indicator-datetime-service [16:44] Saviq: sorry I had to go deal with my kid [16:44] davmor2: ^ this will start dumping debug messages to the terminal [16:45] AlbertA, nw, plenty of things to do :) [16:45] charles: It stopped ringing === dandrader is now known as dandrader|lunch [16:45] Saviq: here's a branch that should do what you want [16:45] AlbertA, did you survive? (or did he/she?) [16:45] davmor2: this may help us to see why it's running late sometimes [16:45] Saviq: lp:~albaguirre/mir/test-hack-consume-buffers-display-off [16:45] davmor2: then, while the debug messages are dumping to the terminal, try to make the alarm trigger late again [16:46] charles: right yeap so I have the stuff dumping in the cli now and then trigger the alarms again right? [16:46] Saviq: yeah we survived :) [16:46] davmor2: hm, it would be nicer if these messages had timestamps. oh well [16:46] MacSlow, poke [16:46] davmor2: yes specifically a late alarm [16:46] mterry, what's up? [16:46] MacSlow, got some feedback about spinner, but not final design note [16:46] charles: okay [16:46] davmor2: since there's no timestamps on here, a little more info on what I'm looking for: [16:46] MacSlow, dropping text, likely use ubuntu logo instead of spinner, and likely back background. Will know more tomorrow [16:47] davmor2: I'm wanting to know which (if any) messages show up at the correct time, and which show up only when the alarm goes off late [16:47] charles: right okay [16:48] mterry, that sound like easy changes... which I could do on the side... once final... still very busy with the "avoid fullscreen notifications messing up visual queue"-issue. [16:48] MacSlow, cool [16:48] AlbertA, thanks [16:49] charles: http://paste.ubuntu.com/7126132/ no alarm at all [16:49] and now I have (process:6560): Indicator-Datetime-WARNING **: Unable to show snap decision for 'Alarm': Timeout was reached [16:49] oho! [16:50] interesting. [16:50] (process:6560): libnotify-WARNING **: Failed to connect to proxy [16:51] Does anyone know why the QtQuick.Controls.MenuBar isn't working on Trusty? [16:51] davmor2: that unblocks me on the never-stops-ringing question anyway [16:52] davmor2: so you got no snap decision popup, is that right? [16:52] charles: I got one when it didn't stop ringing [16:53] charles: with that one I got no alarm so I didn't hit the power button so saw no snap decision either [16:53] davmor2: meeting approaching, got a bug for me? [16:53] davmor2: easier to setup for the meeting with the list :) [16:53] didrocks: writing it now [16:54] didrocks: one issue, indicator-datetime will keep playing the sound until the user dismisses the snap decision. Dismissing the snap decision is impossible if libnotify can't be reached. In that case, the sound plays forever. :/ [16:55] didrocks: (process:6560): libnotify-WARNING **: Failed to connect to proxy [16:55] charles: yeah, in the case I had, I saw the snap decision, clicked dismissed, but no effect [16:55] (in the 2 cases) [16:55] so probably multiple sources [16:56] ya, agreed [16:58] charles: isn't indicator-datetime just notifying for minute change only when about to sleep? [16:58] so the alarm will either pop up before, or after sleep, and never inbetween sleep cycles? [16:58] brb [16:59] didrocks: https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295237 [16:59] cyphermox, no that's not correct. minute_change gets fired whenever the clock notices that the time has changed. This can happen through sleep/wakeup, or through normal time passing [16:59] Ubuntu bug 1295237 in indicator-datetime (Ubuntu) "Sometimes alarms trigger at xx seconds past their time and dismiss on snap doesn't stop the ringing" [Undecided,New] [17:00] davmor2: could you please paste into that ticket a log of the debug messages that get logged during one of those late-by-XX-seconds runs? [17:00] davmor2: is http://paste.ubuntu.com/7126132/ an example of that? [17:02] charles: where is the code that does that? I see LiveClock has a timer, but it's not used in Clock or ClockWatcher. Am I looking at the right place? [17:02] cyphermox: inbetween sleep cycles, it's tested this way: http://bazaar.launchpad.net/~indicator-applet-developers/indicator-datetime/trunk.14.04/view/head:/src/clock-live.cpp#L105 [17:03] cyphermox, in that file, line 112 is the event that fires out to listeners [17:03] cyphermox, and clockwatcher registers for it http://bazaar.launchpad.net/~indicator-applet-developers/indicator-datetime/trunk.14.04/view/head:/src/clock-watcher.cpp#L41 === chihchun_afk is now known as chihchun === om26er_ is now known as om26er [17:23] Saviq: news? [17:23] kgunn: ? [17:23] did we find out whats the probl with those messages? [17:23] asac, building mir [17:23] asac, "those messages"? [17:23] Saviq: err. events not getting dispatched [17:23] asac, I did confirm it's Qt 5.2 - i.e. between images 237 and 238 [17:24] Saviq: cool. and building mir is part of finding out more? [17:24] Saviq: just noticed that there would be race in the branch I sent you, I updated it and adding some logs [17:24] Saviq: do you need more help? anyone i could activitate? [17:25] asac, yes, confirming that it's indeed the stopped rendering that breaks the events [17:25] or not [17:25] AlbertA, ok [17:25] Saviq: you should see it compositing but not posting after screen off with adb shell tail -f /var/log/lightdm/unity-system-compositor.log [17:26] AlbertA, atm I got http://pastebin.ubuntu.com/7126300/ [17:27] AlbertA, although I did merge your branch into trunk, if that makes any difference [17:29] Saviq: I based it off trunk [17:29] AlbertA, right, yeah, the diff was minimal [17:29] popey, davmor2, hey guys, chatting with charles about the alarms bug. Have either of you attempted to get a bustle/dbus-monitor log of the failure? [17:30] Saviq: I can send you the two binary libraries if you want [17:30] AlbertA, armhf/ [17:30] ? [17:31] Saviq: yeah well libmirserver.so.17 and libmirplatformgraphics.so specifically [17:31] AlbertA, sure, that would work [17:32] tedg: no idea what or where I'm trying to get the alarm to not stop to get feedback from the terminal nothing else using G_MESSAGES_DEBUG=all /usr/lib/arm-linux-gnueabihf/indicator-datetime/indicator-datetime-service [17:36] davmor2, Ah, good, that'll be helpful as well. [17:36] didrocks: davmor2: when setting up an alarm, can you see it in the right timezone when checking it in the upcoming indicator? [17:36] in the indicator mine is always in UTC [17:36] and can't get the alarm to be triggered here [17:36] rsalveti: I always uses it on UTC for that test [17:37] right :-) [17:37] rsalveti: have you set the time zone in settings? [17:37] that is probably why [17:37] davmor2: yes [17:37] rsalveti: hmm might be an issue with either eds or clock app then maybe [17:37] popey, https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295271 [17:37] Ubuntu bug 1295271 in unity-notifications (Ubuntu) "On phone, alarms being shown without dismiss/show buttons" [Undecided,New] [17:38] charles: wow, what a lovely bug-bot you are ☻ [17:38] popey: I like the bugs that are easy to trace :-) [17:39] Hah, so it seems that notifications is broken in general. [17:39] tedg, maybe [17:39] Trying to tab complete gdbus on org.freedesktop.Notifications I get "Timeout" [17:39] didrocks: davmor2: yeah, alarm only works when you set it with UTC [17:39] all we know for sure is that we can't connect to it [17:39] add an alarm for 3 hours ago, and worked fine this time [17:39] * rsalveti opening another bug [17:40] rsalveti: bugfest ;) [17:40] rsalveti: you're meant to be fixing them not creating more ;) [17:41] AlbertA, a script? you're making me obsolete! :D [17:41] popey: what was the link to that screenshot again please [17:41] the timezone issue sounds a little like https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1283236, which says it's In Progress [17:41] Ubuntu bug 1283236 in Ubuntu UI Toolkit "Alarms are triggered in UTC tz despite saving in local tz" [High,In progress] [17:41] Saviq: :) [17:41] davmor2: see the bug above [17:41] from charles [17:41] davmor2: it's in https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1283236 [17:41] popey: I want to add it to a bug as I've seen it to now :) [17:42] davmor2: or fish it out of this lot http://popey.com/~alan/phablet/gallery.php [17:42] no, wrong pahttps://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295271 [17:42] Ubuntu bug 1295271 in unity-notifications (Ubuntu) "On phone, alarms being shown without dismiss/show buttons" [Undecided,New] [17:42] d-feet the same. [17:42] I'm totally willing to blame Unity8 now. [17:42] :-) [17:42] this also explains why no sound [17:42] this also explains why popey had those test runs with no sound, I think [17:43] indicator-datetime doesn't play a sound if there's no way to dismiss the popup [17:43] and if there are not buttons, there's no way to dismiss [17:45] sergiusens,tedg: so, sorry, I had a couple of meetings [17:45] davmor2: didrocks: the timezone issue is already opened: bug 1283236 [17:45] bug 1283236 in Ubuntu UI Toolkit "Alarms are triggered in UTC tz despite saving in local tz" [High,In progress] https://launchpad.net/bugs/1283236 [17:45] sergiusens: could you please pastebin "find /usr/share/click/preinstalled /opt/click.ubuntu.com /var/cache/apparmor /var/lib/apparmor -ls" for me? [17:46] rsalveti: ok, so nothing really "new" (not that a good news, but not a bad either) [17:46] rsalveti: \o/ [17:46] yeah [17:46] charles: sorry, so what do you think is currently causing these alarm issues? [17:47] (no sound, no way to dismiss, etc) [17:49] it seems to be bug 1295237 [17:49] bug 1295237 in unity-notifications "Sometimes alarms trigger at xx seconds past their time and dismiss on snap doesn't stop the ringing" [High,Confirmed] https://launchpad.net/bugs/1295237 [17:49] rsalveti: I am tempted to close the first ticket as being too vague. https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295271 and https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295237 are more useful [17:49] Ubuntu bug 1295271 in indicator-datetime (Ubuntu) "On phone, alarms being shown without dismiss/show buttons" [Undecided,Triaged] [17:49] so it might be the same issue [17:49] rsalveti: I've done writeups in both of those two tickets about what I think the cause is [17:49] charles: right, mark one as duplicate of the other [17:50] rsalveti: but the tl;dr seems to be (1) unity-notifications is not responding (so there may be a bug on their end?) and (2) indicator-datetime needs to fail more gracefully when it can't reach the libnotify server (which is a bug on my end :) [17:51] ogra_: , I have u-boot able load the ramdisk and find the init within it, but I'm getting the message that: "initrd: couldn't find data partition" [17:51] hedz09, so try adding datapart= to your kernel cdmline [17:51] cjwatson, sure; one sec [17:51] hedz09, (with the device that holds your rootfs) [17:52] rsalveti: no, 1295271 and 1295237 are slightly different from each other. But I'd be ok with closing bug #1295122 in favor of those two more descriptive ones [17:52] bug 1295122 in indicator-datetime (Ubuntu) "Alarms not going off reliably on recent touch images" [Critical,New] https://launchpad.net/bugs/1295122 [17:52] cjwatson, and no worries; just affecting me [17:53] charles: yeah, sounds good [17:53] davmor2, since you reported 1295122 -- does ^ that sound right to you? === dandrader|lunch is now known as dandrader [17:54] cjwatson, http://paste.ubuntu.com/7126410/ === chihchun is now known as chihchun_afk === bfiller is now known as bfiller_afk [17:58] asac, rsalveti, ogra_, davmor2, sergiusens, charles, @all: https://bugs.launchpad.net/music-app/+bug/1292306/comments/8 [17:59] charles: so I think they are slightly different but related. I'd be happy for you to merge some together or mark both as fixed once there is a fix in place [17:59] Ubuntu bug 1292306 in Ubuntu Music App "Upon upgrading to Qt5.2 the music app no longer plays the next song if the screen is off" [Critical,Triaged] [17:59] confirmed, Qt 5.2 queues events when it can't render === alan_g is now known as alan_g|EOD [17:59] Saviq, nice [17:59] Saviq: awesome [17:59] Saviq: I'll have a look in a bit teatime has been called [18:00] Ah guessing that's the same thing. [18:00] Saviq, fun [18:00] Saviq, the find, not the situation it put us in :-) [18:00] ogra_: initrd: That seems to be doing something, but it fails again: "mounting mmcblk0p3", "btrfs: open mmcblk0p3 failed", "initrd: couldn't find a system partition" [18:00] It's kinda fun with notify-send, you can get it to pause right until you hit the screen on function. [18:00] hedz09, do you have the ubuntu system.img in /userdata/system.img on that partition ? [18:01] (note that you first need to assemble that system.img like rootstock-touch-install does) [18:01] sergiusens, indeed [18:02] Saviq, Do you have a bug tracking that we should make the others dup to? [18:02] tedg, "the others"? [18:02] tedg, that above was the master bug [18:03] we have other bugs for that ? [18:03] Saviq, I'm pretty sure bug 1295237 is the same, if the screen is off the dbus messages don't get processed. [18:03] ogra_: I don't. That partition has system.img at its root abd doesn't reside in a /userdata/ folder within it. [18:03] bug 1295237 in unity-notifications "Sometimes alarms trigger at xx seconds past their time and dismiss on snap doesn't stop the ringing" [High,Confirmed] https://launchpad.net/bugs/1295237 [18:04] ogra_: I have assembled the system.img and am trying to test that I put it together correctly. [18:04] Saviq, I can literally turn off processing them or not by hitting the lock button. [18:04] hedz09, wipe the partition, look at the rootstock-touch-install script and try to do the same by hand that it does and it should just boot [18:04] tedg, possible [18:04] tedg, try the binaries from there [18:05] from my comment, that is [18:05] hedz09, the "initrd: couldn't find a system partition" is only printed if the initrd could not find a system.img in the right place [18:05] k [18:05] * Saviq called it 18hrs ago! WIN [18:06] charles: tedg: thostr_: https://bugs.launchpad.net/ubuntu/+source/indicator-datetime/+bug/1295237 [18:06] Ubuntu bug 1295237 in unity-notifications "Sometimes alarms trigger at xx seconds past their time and dismiss on snap doesn't stop the ringing" [High,Confirmed] [18:06] who is the ultimate owner? [18:06] asac, read up [18:06] ah [18:06] does that answer my question :)? [18:06] asac, it's most probably the same Qt 5.2 events issue === chihchun_afk is now known as chihchun [18:06] right [18:06] asac: we're on it [18:06] still we should keep folks confirming etc. [18:07] and not rest because it might be this :) [18:07] asac, tedg is confirming with the binaries from there [18:07] yeah thanks [18:07] kk [18:07] Saviq, cant be ... everyone denied that when i said 4h ago that the event processing affects alarms :P [18:07] tedg: its on you! :_ [18:07] ogra_, you weren't pushing enough ;) [18:07] Saviq, lol [18:07] Heh [18:07] ogra_, so alarms are processed fine [18:08] ogra_, but when the notification is sent [18:08] ogra_, it reaches unity8, which can't render [18:08] yeah, thats what i said before :) [18:08] ogra_, and Qt doesn't deliver it up until later [18:08] and everything goes apeshit [18:08] poo [18:09] ogra_: ok. Would btrfs: open mmcblk0p3 failed have anything to do with it? [18:09] hedz09, why do you even have btrfs in your kernel ? [18:10] (it doesnt really do well on MMC/SD) [18:10] ogra_: to be honest I don't even know what it is [18:10] its a filesystem [18:11] but you should use ext4 on SD cards anyway, so you dont need it [18:11] sergiusens: is there anything in /var/log/upstart/click-system-hooks.log? [18:12] Saviq, Works for the notify-send case, checking with an alarm. [18:12] asac, your ANN to ubuntu-devel got bitch-slapped into Junk by my dspam ;P [18:12] fginther, you around to help me practice running mediumtests? [18:13] consider yourself ignored ;) [18:13] cjwatson, not there, I do see it in the apparmor one http://paste.ubuntu.com/7126501/ [18:13] Saviq: haha. well, i needed to setup special rules for mark, rick etc. to avoid that [18:13] so its on you :) [18:13] Saviq: its just telling you what will happen :) [18:13] if you ignore it will still happen [18:13] but i know you are joking [18:13] just stop using all caps subjects in mails ;) [18:13] :D [18:14] its good that it even came through [18:14] mterry, can I help? you mean mediumtests as in the jenkins jobs directly? or locally? [18:14] Saviq, locally on my phone [18:14] thought i wouldn't go to list since i was kicked out of the core-dev and ubuntu member club a while back [18:14] mterry, what project in particular? [18:14] Saviq, is there a wiki page for that? [18:14] because of my laziness :) [18:14] sergiusens: mm, that's different, that's processing the symlinks that are in place. I'm trying to work out how the symlinks got there [18:14] mterry, if unity8, `phablet-test-run -p unity8-autopilot -n unity8` [18:14] Saviq, unity8 with my split stuff -- I wanted to test my proposed fixes for the mediumtests scripts to work in that scenario [18:15] asac, how is that lock playing out with the desktop LTS for shared component where we need fixes? [18:15] mterry, then just ./run_on_device [18:15] Saviq, how do I point it at my branch of the mediumtests scripts? [18:15] cjwatson, I dont have a click-system-hooks log though [18:15] mterry, and then PYTHONPATH=~/shell/tests/autopilot autopilot run unity8 [18:15] seb128: everything that goes on touch image is part of this [18:15] special exceptions can be discussed with me or rickspencer3 as pointed out in mail [18:16] sergiusens: so you said this is basically just you at the moment? (and I agree, it doesn't seem like a regression, more a corner case we hadn't previously found) if so I can look at it tomorrow, as I need to leave pretty soon [18:16] asac, what happen to the desktop LTS if the touch image doesn't recover before the release? [18:16] seb128: but we won't really make exceptions from what i understood [18:16] mterry, you need to make install (or ninja, depending on whether you got it on the phone) [18:16] rickspencer3, ^ [18:16] cjwatson, yeah, don't worry about it [18:16] seb128: if that happens we are all fired :) [18:16] Saviq, that's all basic running unity8 tests on phone stuff [18:16] sergiusens: not that I expect I could land a fix for it at the moment anyway :P [18:16] seb128: we have a veery strong team, with smart people. they are actively working on this as priority [18:16] mterry, and the only other point is you need to `stop unity8` before (it will tell you) and keep the screen on [18:16] cjwatson, I don't use this phone for click testing that much (as I use it daily) just flashed to devel-proposed this week [18:16] Saviq, but I'm specifically talking about using the same scripts that jenkins does, but custom versions [18:17] seb128: this is a matter of hours and at most days to sort out [18:17] :-) [18:17] sergiusens: yeah, it has an excitingly baroque set of packages installed apparently [18:17] might have some work to do on garbage collection :) [18:17] Saviq, hrm. They seem very particular to QA. Like connecting to 'ubuntu-qa-g-wpa-d' wifi [18:17] asac, let's see, last time it took you guys a week to unfreeze things when that happened [18:17] mterry, those scripts don't do anything other than setting the device up (or should not) [18:17] Saviq, yeah but notably, they also unlock the screen [18:17] seb128: right. hope we are getting better. also we just have those 3 bugs [18:17] mterry, not for unity8 tests [18:17] Saviq, which is going to change with split [18:17] asac, I'm not sure why your smart people need a freeze also to fix the bugs [18:18] mterry, where it's the test itself that will do it [18:18] seb128: otherwise new bugs will come along [18:18] mterry, unity8 tests are excepted from unlockign [18:18] asac, well, as long as they are not regressions or blockers [18:18] mterry, and even for other tests, they use (or should use, in the case of smoke testing) the process_helpers from unity8 [18:18] Saviq, well. they actually do unlock screen there, but then also stop unity8 [18:18] mterry, so all in all [18:18] seb128: every regression is a blocker [18:18] mterry, just test locally, everything should be the same [18:18] seb128: thats the current standing rule [18:18] asac, what about changes on shared component that don't affect the touch runtime paths? [18:19] seb128: we discuss that, so might change, but for now we continue to stay strong [18:19] Saviq, yeah, but process_helpers.unlock_unity in unity8 will hopefully go away with split (unity8 doesn't need it) [18:19] mterry, if it's not, we generally need to fix the scripts, not the tests [18:19] stgraber, barry are there any plans to enable 3rd party gpg keys for system-image updates? From what I see the option for porter devices is disabling signature verification in recovery, or flash a fake archive-master key [18:19] hi seb128 [18:19] asac, screwing us on the desktop/LTS on the way... [18:19] seb128: if there is any likelyhood that this affects runtime behaviour its covered [18:19] rickspencer3, hey [18:19] seb128: otherwise you can move on [18:19] Saviq, and mediumtests I was planning to fix to modify lightdm config and swipe greeter itself [18:19] mterry, sure, there should be a unity-greeter.process_helpers then [18:19] seb128, you guys have important fixes that you are blocked on getting into the archive for desktop LTS now? [18:19] mterry, or an equivalent [18:19] or, are you just worried that that will happen? [18:19] rickspencer3, not now, but starting tomorrow it seems, from asac's email [18:20] rickspencer3, the later one [18:20] janimo`: not on the official s-i.u.c site afaik. but you can run your own site with your own keys, as some folks are already doing [18:20] seb128, right [18:20] ogra_: ahh, yeah the mmcblk0p3 partition is ext4. And I am not having any luck finding out where it's looking for system.img. It seems like /data/system.img but it isn't finding it there. [18:20] rickspencer3, I've fixes on indicators waiting for review that we were planning to land tomorrow [18:20] mterry, so yeah, if you're after fixing the scripts, you'll need fginther/vanguard for mediumtests and doanac/plars for smoketesting [18:20] Saviq, yeah but it's tricky. We need to modify lightdm config to run greeter with testability. I could leave the unlock_unity method in unity8 source if we wanted, but unity8 wouldn't use it itself. Made sense to move it into mediumtests [18:20] so, I read this: [18:20] We also kindly ask our core-devs with direct upload power for their [18:20] support. Please wait with your uploads that affect touch until we have [18:20] fully recovered. [18:20] rickspencer3, so it's not an hypothetic situation [18:20] janimo`: we have okay support for that now (not perfect, but we'll be improving that after 14.04 i suspect) [18:20] rickspencer3, right, that could be after the LTS release for what we know :p [18:20] barry, right, and if using new keys the recovery image needs to have another archive-master right? [18:21] mterry, it needs to be in some source [18:21] rickspencer3: i think his indicators go through CI TRain. if its desktop only they can float of course, but seems there are shared components in his landing [18:21] mterry, to verify it works [18:21] seb128, well, I think, let's fix the last issues right away [18:21] right, I get it [18:21] mterry, whether it's lightdm, unity-greeter or unity8, you decide [18:21] kk [18:21] barry, yes it is ok support indeed ( I use that now) but was wondering if there are other plans. I udnerstand it is risky to have other keys signed by the official Ubuntu ones [18:21] janimo`: did you read https://www.stgraber.org/2014/02/11/your-own-ubuntu-touch-image-server/ ? [18:21] I think that it's not an issue today [18:21] hedz09, right, it needs to be in /data/system.img on that partition [18:21] rickspencer3, I don't get why we need a lockdown to fix bugs, but let's see [18:21] and tomorrow will be a pain, but not an issue [18:21] Saviq, ? [18:21] mterry, but it does need to be *tested* with every -ci job [18:21] janimo`: I'm using my own gpg keys in my system-image server [18:21] mterry, it can't live with the mediumtest runners etc. [18:22] stgraber, yes, and I have set up a local server (more or less). I think there's a typo there, it needs newgrp instead of newgroups [18:22] janimo`: we can also build a trust chain from the main system-image server using the device keyring, but we haven't done so at this point because we didn't have any reason to [18:22] rickspencer3, well every day we delay desktop landing is less days to verify the fixes that should land in the LTS, I'm not sure we are doing the right tradeoff there [18:22] janimo`: I mean, the ones generated by generate-keys, so I suppose those are testing only (but the signature gets verified in recovery) [18:22] rickspencer3, I get that touch is important, but the desktop LTS is as well [18:22] mterry, we (lightdm, unity8 devs) need to be responsible for keeping those helpers working [18:22] seb128, well, like I say, let's just get this fixed and promoted so it doesn't become an issue [18:22] Tassadar, so you did not need to change archive-master then, ok. I need to see why I had verification failed there [18:23] rickspencer3, agreed, and we could do that with a lockdown of the uploads... [18:23] you have to put your archive-master into the recovery image [18:23] lockdown doesn't resolve any byg [18:23] bug [18:23] most people aren't qualified to help of course [18:23] ogra_: Okay, it's there but it just isn't find it. System.img is supposed to be ext2, right? Couldit be a permission issue on the folder? I can't imagine... [18:23] mterry, so they need to be with our code, not with mediumtest scripts (which should do as little as possible so that we can reproduce the same env locally) [18:23] Tassadar, indeed, what I figured [18:23] rickspencer3, sorry, *without* [18:23] seb128, I understood [18:23] so they just have to stop landing, it's not like our developers are interchangeable [18:23] hedz09, you need to create it the same way that the prepare_ubuntu_system() function from rootstock-touch-install uses, then unpack the tarball into it and copy the converted android system.img into var/lib/lxc/android/ in there [18:23] cjwatson, right, I get it [18:23] mterry, bug #1262879, too [18:23] bug 1262879 in Ubuntu CI Services "There should only be one, documented, way to run tests on devices" [Undecided,Confirmed] https://launchpad.net/bugs/1262879 [18:24] stgraber, right, the trust chain plans is what I was wondering about === chihchun is now known as chihchun_afk [18:24] Saviq, fine. That tiny bit of python code can continue to live in unity8 code, even though unity8 tests don't need it. But that doesn't change the fact that mediumtests need changing for split mode, to run greeter with testabiliity [18:24] hedz09, and it is system.img, not System.img (no capital S) [18:24] Tassadar, janimo` if you find things missing from the client, please do file wishlist bugs. for the sake of stability i don't want to do much more with s-i in 14.04 except fix critical bugs, but i'm interested in doing more feature work after trusty is released [18:24] mterry, they need changing to use whatever helper you provide to run greeter with testability [18:25] mterry, the mediumtests scripts can't know what happens [18:25] barry, I will if I find any bug :) [18:25] Saviq, it's not that simple, due to the way lightdm works [18:25] mterry, I understand [18:25] barry: except for the duplicate keyring files bug it's fine [18:25] Tassadar: yep [18:25] mterry, but we still need to build as little knowledge as possible into mediumtests, and as much as possible into the components that the knowledge is of [18:25] janimo`: cool [18:26] mterry, or they will get stale, [18:26] mterry, things will fail here, but not there [18:26] mterry, we won't be able to reproduce locally [18:26] Saviq, but this isn't component knowledge, it's system knowledge [18:26] janimo`: oh, yeah, look at http://bazaar.launchpad.net/~vbocek/+junk/system-image-server-tasemnice/revision/221 and the bug linked in the commit message [18:26] stgraber, any plans to package the server? [18:26] mterry, good, then we need system helpers [18:26] Saviq, this particular instance of "run greeter with testability" [18:26] mterry, maybe they need to live with the integration tests project that elopio is starting to build [18:27] Tassadar, thanks I had just figured this out and used a full image to flash the phone. I had not done anything with deltas or actual upgrades [18:27] mterry, with as much as possible abstracted closer to the affected services (and tested there) [18:27] janimo`: I don't have any such plan personally, ubuntu-cdimage isn't package so I don't see a good reason to package ubuntu-system-image besides the server on which this actually runs is on precise and we don't have package installation rights, so having it packaged would just make things harder for me [18:27] janimo`: it'd bite you later and you'd have to edit the index.json files by hand [18:27] mterry, I just want to save us the pain of having to run around after all the scripts that somewhere somewhen implemented a workaround [18:28] mterry, we just need to say: to unlock the device, call foo.foo_unlock [18:28] mterry, and that foo.foo_unlock needs to live somewhere outside of mediumtests [18:28] Saviq, I'm not able to revive the screen with this mir binaries. Either by locking with the button or timeout. [18:28] Saviq, Is that expected? [18:28] Saviq, I get the vision. It's just that some calls have side effects. Like in this case, probably a reboot and a system config change that will later need to be undone [18:28] tedg, hmm just noticed [18:29] tedg, actually no, works here [18:29] tedg, try touching the screen [18:29] tedg, maybe the compositor needs to catch up [18:29] mterry, I understand [18:29] Saviq, Yeah, I think you have to generate an event. [18:29] tedg, that can be expected, yeah [18:30] mterry, but that "change, reboot, clean" logic might still be outside of mediumtests, and that's completely fine [18:30] mterry, as long as that's the only documented and supported way of doing that [18:31] Saviq, sure. I just think it makes it unsuitable for unity8 source. But I can toss it in some elopio repository [18:31] mterry, of course, that's why I said lightdm / unity8 / unity-greeter, but maybe even something global [18:32] mterry, truth be told, there's probably only a few autopilot tests in unity8 that should remain there [18:32] Saviq, sure. I thought mediumtests was a reasonable global place to put it (it has the rest of the kitchen sink in there). But since we are apparently trying to get away from that, I'm happy to talk to elopio [18:32] mterry, after all, they are integration tests, and how can you integrate just one project? [18:32] * Saviq AFK === chihchun_afk is now known as chihchun [18:34] Saviq, asac, confirmed to be the same bug, Saviq's binaries make it so alarms show. [18:34] (and make noise) [18:34] finally [18:34] tedg: cool [18:34] :P [18:35] now Saviq needs to RTK :) [18:35] hehe [18:35] (if people would just have listened to me 4.5h ago we would have had far less noise :P ) [18:36] ogra_, I'm strongly resisting the urge of suggesting that perhaps messaging is causing problems in the alarms, but I guess that would be more noise? [18:36] * rickspencer3 ducks [18:36] lol [18:36] * rickspencer3 notes to listen to ogra early and often [18:37] * ogra_ hugs rickspencer3 [18:37] * rickspencer3 hugs ogra_ :) [18:37] we're getting there [18:37] so close to promoting with Qt 5.2 !!! [18:38] ogra_, can you tell rickspencer3 to not lockdown the archive tomorrow? ;-) [18:38] hahahaha [18:38] ogra_: /nick oracle_ogra [18:39] seb128, we'll all be fine, there will be a fix tomorrow and nothing will be blocked [18:39] ogra_, I like that prediction ;-) [18:39] ;) [18:40] I bet we have a promotable image tomorrow, and promote it Monday [18:40] that's my prediction [18:40] bets? takers? [18:40] Promote it tomorrow if the fixes land early enough [18:41] davmor2, your betting on promotion tomorrow? [18:41] I like the optimism [18:41] I would happily pay a round of beers if that happens ;-) [18:41] though, I am considering that we hold it back just to annoy seb128 [18:41] :) [18:41] lol [18:41] rickspencer3: yeah but hey I'm insane ask anyone [18:42] rickspencer3: I think the devs have a good handle on the causes now it's just the fixin' :) [18:42] oh and testin' [18:42] yeah [18:43] I like how the last few issues all boiled down to the same thing [18:43] rickspencer3, you think that having to sit outside in the sun on a friday, because the archive is locked down and we can't work, is annoying me? ;-) [18:43] Im really happy that we have a root cause [18:43] seb128, that's what I keep saying, stopping the line means *less* work! [18:43] lol [18:43] :) [18:44] note that we still do other landings until tomorrow morning [18:44] if something slips in, it defers a bit further [18:44] but i am confident, so yeah. monday we go to green [18:44] asac, you mean if a regression slips in? [18:45] well, there were quite a bunch of changes in the most recent image that just finished [18:45] http://people.canonical.com/~ogra/touch-image-stats/249.changes [18:45] yikes :) [18:45] rickspencer3: yes, we still have careful landings tonight. those might bring regressions (dont know whats coming right nwo) [18:45] so lets hope we still get 100% in the tests :) [18:45] I assume that those people were super careful to ensure that their changes would not cause regressions [18:45] ack [18:45] under the circs [18:45] thats the idea === chihchun is now known as chihchun_afk [18:46] I don't think anyone would want to be the one that causes us to go another day without promoting [18:46] they'll have to answer to seb128 otherwise [18:46] :) [18:46] lol [18:47] rickspencer3, our issue is that everyone tests their changes in their silo against the most recent image ... but you can only test for bugs that come out of interaction of all the landings in the final build ... so there is always a risk of one landing influenceing the other [18:47] rickspencer3, I still think archive lockdown is not a solution and the wrong approach (that's the archive moving that created those issues or is delaying the fixes) but I think we need to agree on disagreeing there [18:47] ogra_: rigth, but at this time we go very careful about combining those [18:47] that first testing has to be 100% to land ... but that second testing is hardly predictable [18:47] rickspencer3, sorry, "that's *not* the archive..." [18:47] ogra_: take less per image cycle etc. [18:48] asac, yes, thats why i want more images :) [18:48] (i know, we'll get ther) [18:48] yes, i want that too - and a pony :) [18:48] * ogra_ looks up where to order ponies online ... [18:49] and a narwhal :) [18:49] lol [18:49] if you get it, can i get the other thing ? [18:49] :) [18:49] ogra_: send the pony to colin and adam :) [18:49] lol, k [18:49] maybe make it two [18:49] half a pony won't last that long [18:50] :) [18:54] rsalveti, Saviq, popey: confirming that the pushed files allows a second track to play will test alarms now === chihchun_afk is now known as chihchun [18:54] great [18:54] now we just need to fix the issue ;-) [18:54] details [18:55] ogra_: https://bugs.launchpad.net/music-app/+bug/1292306/comments/8 [18:55] Ubuntu bug 1292306 in Ubuntu Music App "Upon upgrading to Qt5.2 the music app no longer plays the next song if the screen is off" [Critical,Triaged] [18:55] davmor2, yes ? [18:55] (see what was talked about for the last 30min) [18:55] ;) [18:56] I was a tea I was just confirming as I'd been asked to test :P [18:56] davmor2, dont foget to test the vol up/dn keys too ;) [18:57] ogra_: I don't care about them there are only 2 volumes Damned loud and mute [18:57] heh [18:58] sound on the alarm isn't playing 100% correctly but it worked [18:58] great [18:59] tries for a second time to be sure [19:01] i think the semi broken sound behavior was mentioned by charles before [19:07] I'm trying to use QtQuick.Controls.MenuBar in my QML app, but when I run, the menu does not appear. Any idea why? [19:08] 2 successful alarms \o/ again sound is abit wonky but I can live with wonky sound if it works :) [19:11] Saviq, rsalveti: it's like a phone that works again and everything woohoo! \o/ [19:11] crazy [19:11] shhhh or popey might break it again [19:11] well, it isnt fixed yet [19:12] we dont have a fix [19:12] (what you are testing is a hack to prove the point) [19:18] yeah [19:18] ogra_: right and now the point it made people know how to fix it right? [19:18] not really [19:19] the prob is to identify where the fix needs to live [19:19] there are two options === chihchun is now known as chihchun_afk [19:19] * ogra_ goes for dinner [19:22] jonahbron, try #ubuntu-app-devel === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === dandrader is now known as dandrader|afk [19:55] niemeyer, hi, when calling a QML objects Set method from go, should that trigger the onChanged signal in qml? [19:55] cwayne: Yeah, it does [19:56] niemeyer, hm, it doesn't seem to be for me, I must be doing something wrong [19:56] cwayne: It certainly does, as Set is using tradition Qt logic to change the propertie [19:56] s [19:57] cwayne: It's Qt that calls the signals === bfiller_afk is now known as bfiller [19:57] s/tradition/traditional/ [19:57] cwayne: If you have a small snippet, I'd be happy to have a look [19:59] niemeyer, thanks, let me do some trial and error first, I think I may know what I was doing wrong, if I can't figure it out, I'll shoot you a ping tomorrow :) [19:59] cwayne: Sounds good, will be around [20:01] evening everyone! [20:02] is there a guide for syncing google calendar to touch? [20:03] timppa: http://askubuntu.com/questions/360554/how-do-i-sync-google-contacts [20:03] rsalveti: cool, thanks! [20:03] rsalveti: get it added to the info bot :) [20:04] davmor2: I believe the official support should be landing soon :-) [20:04] rsalveti: um, but thats for contacts? [20:05] right, it indeed, didn't see you asked for contacts specifically [20:05] maybe renato can help [20:05] or calendar... :D [20:05] right, my brain is damaged [20:05] asac: you can get me a colleague to help take some of the click development/maintenance load so that I can focus on the build stuff for more than five minutes at a time. oh wait, we're doing that :-) [20:06] renato: ping [20:09] timppa, try this one: http://paste.ubuntu.com/7127033/ [20:10] renato: great! Thanks, I'll try [20:11] niemeyer, so if you took this example https://github.com/go-qml/qml/blob/master/examples/customtype/customtype.go, and if you called v.SetText from some other function, that should still send that signal right? [20:14] FYI all-devs, http://bit.ly/Nex7-2013_has_only_2years_to-be_fully_operational_cos_price_will_fall_to_70pounds_or_99dollars_in_US (a Proof-Link, this type of advert will also re-occur, again). [20:14] leaves, in confidence of this project [20:20] niemeyer, oops sorry, ignore me, i'd forgotten qml.Changed :) === dandrader|afk is now known as dandrader [20:22] cwayne: No.. [20:22] cwayne: In that case you're not calling obj.Set.. [20:23] cwayne: You're calling pure Go logic, which QML has no way to know about [20:23] cwayne: To notify about changes in that method, you need to call qml.Changed(&v, &v.Text) [20:23] niemeyer, yeah, so if I do that and then call qml.Changed it works [20:23] cwayne: Right [20:24] cwayne: I need to step out, but will be back soon in case you want to talk further about this [20:27] cjwatson: yes, oh yes. 10 days from now :) [20:27] i am pinging him like every other day :) [20:27] because i am not sure if tis a dream [20:27] hehe [20:39] robru, kgunn: so renaming session-manager-touch is kind of a pain at this point in the cycle [20:39] robru, kgunn: means a new source package in archive [20:40] mterry, hmmm [20:40] mterry, the other thing is that then the source package name won't match the binary [20:40] mterry, so maybe it's easier to just register a new lp project with the right name? [20:40] robru: mterry ...might need to be a manual aspect for this one wrt ci train [20:41] e.g. xorg is like this right ? [20:41] "extra packages" [20:41] kgunn, xorg isn't in ci train, it's done with source package uploads only [20:42] kgunn, robru: for renaming a project, we need to file an LP question right? [20:42] mterry, no.... just register a new project. that's how i've always done it [20:42] robru, but we can rename and I think retain internal LP links [20:43] mterry, oh, ok, never heard of that [20:43] robru, but it's crazy -- you have to file a Question against the LP project [20:44] robru, kgunn: https://answers.launchpad.net/launchpad/+question/245816 [20:44] mterry, seems like a lot of hassle and waiting. if we just register a new project we can copy the branches over immediately. [20:45] robru, they are actually pretty quick usually. But we can always light a fire [20:45] ok [20:45] kgunn, i'll update my MP not to rename the source package [20:45] robru, plus, if we make new project, we'd just have to file a request to delete old one at some opint [20:46] mterry, true, but deleting the old one doesn't block getting this silo built right now ;-) [20:46] robru, fair [20:46] kgunn, actually come to think of it, i don't think the package name was a blocker, just a warning. the real problem was the native/split version scheme. [20:48] kgunn, ok, try adding my MP and rebuilding now [20:50] robru: ack [20:50] mterry: robru...webops can help, they told me to skip the lp question thing last time i did something similar [20:51] kgunn, oh, if you have an in, please use it ;-) [20:51] i don't know those guys at all [20:51] kgunn, I asked on the lp ops channel [20:51] is that different than web ops? [20:52] mterry: its basically the same people....altho less are in lp ops...use #webops [20:55] bzoltan: Mirv: will qt 5.2 be available for saucy in the sdk team release PPA sometime soon? [20:55] or will 5.2 on saucy not be supported [21:00] ogra_, poke -- any objection to renaming session-manager-touch project? === chihchun is now known as chihchun_afk [21:08] cwayne: hey - new upstart app launch support is in silo 3 PPA, and ready to land. If you want to get a preview before it lands, now's the time to grab those packages [21:08] cwayne: but I'm hoping it'll land today sometime, so you won't have much longer to wait [21:08] thomi, \o/ [21:09] youre my hero :) [21:09] cwayne: sorry it took so long :( [21:09] thomi, no worries, we're all super busy, stuff takes awhile :) === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === kenvandine_ is now known as kenvandine === salem_ is now known as _salem [22:44] um hi [22:44] hello [22:45] from texas, thought about trying ubuntu on the nexus, seems pretty neat [22:46] relley, a nexus in texas? [22:46] i know, it's a crazy thought [22:47] relley, i have a Nexu 4 and 7, they're pretty great little gadgets [22:48] i'm thinking mine is a 7... android is/was fun to play and do work on, especially while wife is watching TV === chihchun_afk is now known as chihchun [22:53] popey, balloons, Could we push the current music-app to the click store? There have a few fixes for search via the HUD and opening tracks from the dash/Filemanager--some of which relate to Qt5.2. [22:53] vthompson, sure [22:59] rev389 going in [22:59] vthompson, ^^ [23:00] balloons, excellent! Thank you! === chihchun is now known as chihchun_afk [23:11] I have multiple apps open. In the apps scope, when i go to opened apps and log press to mark for closing, when i hit the first x to close an app, all the other x-es all disappear. I did not see any bug like that mentioned on the mailing list [23:11] But I may well have missed that [23:12] anyone else seeing this bug? [23:13] matv1, what image revision are you running? [23:13] 248 === chihchun_afk is now known as chihchun [23:13] on maguro though [23:13] i should add [23:14] matv1, I haven't seen that bug previously. I'm on 249 now on mako. [23:15] okay. I will update to that as well and see what gives [23:15] matv1, Maybe it exists on 248, but I never noticed [23:16] vthompson if I get it on 249 as well, what kind of logfiles would be usefull for a bug report? [23:16] if any :) [23:17] well, there's a unity8 log here: /home/phablet/.cache/upstart/unity8.log... that might have something [23:17] otherwise, I'm not sure [23:19] note that maguro is no longer supported.. [23:20] popey I aware thanks :) [23:20] but I've not seen anyone report that issue [23:20] although the term supported was never actualy an apt one [23:20] well nobody tests it on maguro now [23:21] and if a bug manifests itself on maguro only we wont fix it [23:21] ofcourse. dont worry I know [23:21] does it happen in the emulator would be worth checking [23:21] which apps did you have open? [23:22] it doesnt matter which [23:22] any combi will do it [23:22] and is it only the first app you close, the first in the list [23:22] or does it happen with any app you close? [23:22] funily enough i am now noticing that it will only happen when you start killing them from the right :) [23:22] go figure [23:23] i never do that [23:23] so worth me testing [23:23] one mo [23:23] well go for it! [23:23] sadly not on nexus 4 on #249 [23:23] hapily, i would say [23:24] then i wont report it [23:24] sorry about that [23:25] uhm. I am confused. If you think it could be helpfull i will popey [23:26] confused? [23:26] i just said sorry that I can't reproduce it on nexus 4 [23:26] i thought you meant sorry to hear i am not reporting it?? [23:27] ah! [23:27] heh no ☻ [23:28] i will try on the emulator later. If it shows there, i will definitely file. [23:28] thanks [23:28] popey against what would i file it? unity? === chihchun is now known as chihchun_afk [23:28] yes [23:28] cool [23:29] so yeah, https://bugreports.qt-project.org/browse/QTBUG-37677 [23:29] let's see what happens with that [23:30] oof [23:32] would be nice to squash the three blockers in one go [23:34] Saviq: do we have contact with upstream to get a prompt response? [23:35] Saviq, thanks for filing that. I hope it gets resolved relatively soon. If not jhodapp might be able to push the media hub's MediaPlayer component out quicker and the music-app can try to addapt [23:35] *adapt [23:36] popey, yeah, in EU timezone [23:37] vthompson, yeah, let's see what happens [23:37] balloons: vthompson I am running autopilot on #249 with your latest music click [23:37] and the screen is blanking part way through! [23:38] i dont remember this happening before [23:38] popey, is the app actually loading? [23:38] yes [23:38] but after a while, the screen goes black as usual when you're not using it [23:39] by blanking do you mean white? Or the screen shuts off? [23:39] shuts off [23:39] * popey re-runs [23:39] I have a test in the works that should allow the screen to shut off to test the lifecycle--but that hasn't merged yet [23:39] right, it just dimmed [23:39] and now blank [23:40] hm, maybe the shuttle test took longer than a minute due to randomness being not so random [23:40] 23:39:41.844 INFO globals:58 - Starting test music_app.tests.test_music.TestMainWindow.test_artists_tab_album (with touch) [23:41] thats what it was running at the time [23:41] also, i still see metadata about my music in the app [23:41] odd, I'll run as well [23:41] which may be throwing it off [23:41] need to delete the mediascanner database and music.. where is that? [23:41] ah, yes. bah. So what I do is remove /home/phablet/.cache/mediascanner and reboot the device [23:42] ok [23:42] the test should delete your music, um correctly [23:42] the test should do no such thing! :D [23:42] but the mediascanner db stays in place [23:42] ha, no disagreement here [23:42] I know balloons has a long standing bug on that [23:43] We have an issue currently that we cant mock home for the click package test. There's a branch in the works to allow us to do so [23:43] balloons is blocked on this ATM: https://code.launchpad.net/~elopio/ubuntu-ui-toolkit/initctl_env_var/+merge/208612 === chihchun_afk is now known as chihchun [23:45] vthompson: much better now I ditched that db [23:47] popey, much better until you want to copy your music back over ;) [23:51] vthompson: Feel free to replace those three songs any time you like! [23:52] will probably be doing so when I restructure for the new mediascanner2.0 db [23:54] vthompson: oh, while you're about, we had a hangout today for music app.. [23:54] Also, I've grone to love Kerensa's autotuned voice [23:54] wanted to let you know we're going to have a few hack days next week [23:54] Tues-Weds [23:54] lol [23:54] freak [23:54] ha [23:54] Tuesday will be Music & Reminders [23:54] o/ [23:54] will blog about it tomorrow [23:54] o/ [23:55] excellent [23:55] Ran 13 tests in 676.460s [23:55] OK [23:55] boom! [23:55] boom is much better than flop [23:55] approved music into the store balloons [23:55] so it'll be on image #250 at ~3AM UTC [23:56] cool [23:56] The hack day will focus on bitesize bugs, polish, getting people involved, more testing etc [23:57] i didnt realise until today that the touch-coreapps metapackage in the daily ppa didn't pull in music-app, so fixed that [23:57] and made sure it installs on trusty and saucy [23:57] so no excuse for people not to test on the desktop [23:58] Right, time for me to turn back into a pumpkin. nn all!