=== vying is now known as Guest50651 === chriadam|away is now known as chriadam === Zic_ is now known as Guest26871 === rsalveti_ is now known as rsalveti === zenvoid_ is now known as zenvoid === mhall119_ is now known as mhall119 === achiang` is now known as achiang [02:19] Tried it again and it just worked this time... oh well. === chihchun_afk is now known as chihchun === nhaines_ is now known as nhaines [03:59] should i buy the google nexus7 or the asus nexus 7 2013 for ubuntu touch? === kaxing is now known as Guest53339 === chihchun is now known as chihchun_afk === chriadam is now known as chriadam|away [07:07] anyone using Touch on the Nexus 7 (2013) [07:07] ? [07:08] cwayne: hey, I've commented on the bug, seems we have some autopilot issues there. However, the MainViewStyle is also broken in a ense that the artwork is not found, and there are some components referenced that are not found. You have a binding which is not necessary at all. === chihchun_afk is now known as chihchun [07:18] didrocks: I thought there was some conclusion last cycle that the spreadsheet needed to be phased out. Now it's part of infrastructure and needs firewall holes and automation? [07:18] Erm, ECHAN. [07:19] infinity: the spreadsheet as it was needed to be phased out (it was all manual) [07:19] infinity: then, I was asked (with a few days credit to build everything) to do CI Train alone [07:19] didrocks: Driving our infrastructure from gdocs is not sane. [07:20] agreed, and that's not what will happen once the CI team will build the Airline [07:20] meanwhile, we took the cheapest decision to have sso integration, ui driving and automated refresh [07:32] good morning === chihchun is now known as chihchun_afk === oSoMoN_ is now known as oSoMoN === psivaa_ is now known as psivaa === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === oSoMoN__ is now known as oSoMoN === chihchun is now known as chihchun_afk [08:56] Good morning all; happy Monday, and happy Bunsen Burner Day! :-D === shadeslayer_ is now known as shadeslayer [09:19] pitti: hey! It seems we can't get any crash file on the phone anymore [09:19] pitti: that's maybe due to new lxc/apparmor, the apport logs are: http://paste.ubuntu.com/7183953/ [09:20] pitti: I guess, this is a consequence of: http://paste.ubuntu.com/7183958/ [09:26] JamesTait, yeehaw! [09:31] didrocks: the crash was fixed in apport recently, I can upload the fix todayh [09:31] pitti: do you know what triggered it into saturday's image? [09:31] pitti: as we didn't get it and apport was working before [09:32] dholbach: not exactly; supposedly some new apparmor restrictions? [09:32] (but yeah, getting the fix ASAP would be nice, we don't have any image view on the numerous crashers we have for 3 days) [09:32] pitti, eh? [09:32] dholbach: sorry, I meant didrocks [09:32] pitti: I guess so, I'll ask jdstrand to add that to his testsuite then [09:32] ah, yes :) [09:32] didrocks: bug 1296026 FTR [09:32] bug 1296026 in apport (Ubuntu) "apport crashes at reading cgroup" [Undecided,Fix committed] https://launchpad.net/bugs/1296026 [09:33] pitti: great! do you think you can upload it soon so that we rekick an image and see what current crashers we have? [09:33] didrocks: yes, I can; I didn't last week due to the freeze and forgot on Friday (too many meetings and couldn't type well, etc.) [09:33] pitti: no worry, thanks a lot! :) [09:38] plars: FYI ^ (no need to bisect I guess) [09:42] didrocks: bonjour :) [09:42] didrocks: are the click-app AP tests working? [09:42] zsombi: hey! yeah, nothing to signal in that front, why? [09:43] didrocks: we are trying to test UITK PPA for calendar-app, and we are getting crash for the phablet-click-test-setup [09:44] zsombi: I guess Mirv is currently running them manually, and I saw him discussing about parameters, maybe he can help you? ^ [09:44] zsombi: but to be clear: on the infra, it's working [09:45] didrocks: ok, thx! [09:45] yw ;) [09:45] zsombi: pastebin? [09:46] Mirv: pastebin what? error? [09:47] zsombi: there might be something going on, but you could edit /usr/bin/phablet-click-test-setup lines 65-66 and make it like at the top of: http://pastebin.ubuntu.com/7184049/ - as you don't need unity8/uitk tests downloaded for click testing [09:47] Mirv: http://pastebin.ubuntu.com/7183954/ [09:47] zsombi: well yes if you need help [09:47] zsombi: oh, that's indeed a known problem when you're testing UITK from landing PPA. in that case you really need to what I pasted to not have the click test setup try download uitk itself [09:48] I don't actually even know why it tries to do that, since both unity8 and UITK autopilot tests come from .deb packages [09:49] zsombi: after changing those lines it should work [09:50] Mirv: i think it does it because the phone is RO [09:50] the idea is you can run phablet-test-run without making your phone RW [09:51] popey: oh, right, makes sense. it just doesn't understand it'd need to get the tests from a PPA and fails since the version being tested is not in archives. === cjwatson_ is now known as cjwatson [10:09] Morning all [10:11] should i buy asus nexus7 or google nexus 7 2013 to run ubuntu touch on? [10:13] lotuspsychje: the nexus 7 2012 is no longer supported, but the nexus 7 2013 is [10:14] popey: so that would be the google nexus 7 2013 i need to buy right? [10:15] there's two, the nexus 7 2012 with a lower resolution screen, and the newer nexus 7 2013 with the higher resolution screen, the 2013 is the one we support [10:15] http://www.tabletcenter.be/product/359870/category-206965-nexus-7/nexus-7-32gb-2013.html [10:15] this one ok? [10:16] thats the nexus 7 2013 [10:16] popey: ok great tnx for info mate [10:50] hey, quick question, I'm in progress of flashing the devel channel onto a flo device, any ideas on how long it takes, ubuntu-device-flash has exited about 10 minutes ago, I see what looks like the ubuntu boot loader ot the screen (volume keys are responsive) and no progress, I can adb in and top shows no process consuming significant amount of CPU time [10:50] is that expected? I read https://wiki.ubuntu.com/Touch/Install but it dones't say how long the process takes [10:51] or should one eactually select something form the boot loader at the end === oSoMoN_ is now known as oSoMoN === MacSlow is now known as MacSlow|lunch === oSoMoN_ is now known as oSoMoN === oSoMoN_ is now known as oSoMoN [11:23] anyone? [11:25] zyga: what command did you use to flash it? [11:26] zyga: and what was it running before you ran the command? [11:26] popey: ubuntu-device-flash --channel=devel --bootstrap [11:26] popey: it was running stock android that it came with (just freshly out of the box) [11:26] popey: it was in the unlocked bootloader [11:27] popey: this is the log from the flash process: http://paste.ubuntu.com/7184405/ [11:27] whats on screen now? Ubuntu logo? [11:27] popey: should I select something from the menu on the screen of the device now? [11:27] there's a menu? [11:27] popey: the boot men, reboot system now, install zip, etc... [11:27] popey: yes [11:28] hmm, odd. [11:28] I can navigate with the volume keys [11:28] I'd be inclined to try a reboot first [11:28] popey: I did, once, it then goes to be stuck displaying the google logo [11:28] popey: I can try again [11:29] when its displaying the google logo, leave it [11:29] it takes a while [11:29] popey: the menu says, 'recovery v6.0.4.6' FYI [11:29] ok [11:29] rebooting [11:29] you may need to re-flash if you rebooted while it was doing its stuff [11:29] popey: I can adb shell immediately [11:29] popey: and this is 'ps' [11:30] http://paste.ubuntu.com/7184412/ [11:30] popey: it looks suspicuious [11:30] 223 0 1548 S {adbd} /bin/sh /scripts/panic/adbd [11:30] I'd re-run the flash and leave it alone [11:30] specifically this, it does nothing more than that [11:30] popey: I did, this is the third run now [11:30] and leave it [11:30] popey: I cannot find any images on the /media partition [11:50] is devel-proposed borked for anyone else [11:59] cwayne_, borked ? [12:00] ogra_: hey, could you have a look at the backlog and tell me if the flash instructions are borked or is the image/tool borked? [12:01] zyga, i flashed multiple times today, no issues here with the same command [12:01] ogra_, stuck at google log [12:01] logo* [12:01] bcwthats not helpful ... there are tons of conditions where it can be "stuck at google logo" :) [12:01] ogra_: ok, trying again [12:01] do you have adb ? [12:02] ogra_: if it fails, can I poke you for some help? I can adb in and do anything you may need me to [12:02] ogra_, nope [12:02] zyga, so what does the prompt show in adb ? [12:03] ogra_: I'm flashing it now but before I was stuck on the 'ubuntu recovery bootloader' screen and on the device I had something process with 'panic adb' in the name, probably because stuff failed and adb popped up as a last resort [12:03] ogra_: 13:30 < zyga> 223 0 1548 S {adbd} /bin/sh /scripts/panic/adbd [12:04] ogra_: I wiped my .cache/ubuntuimages and am getting the data again, just in case [12:04] ogra_: this is on flo, btw [12:04] if it panics in the initrd it could not find a root disk [12:04] oh., i havent tried flo todsay [12:04] ogra_: :) [12:04] * zyga would love if you could [12:04] check if your disk is full [12:04] ogra_: the device is fresh out of the box [12:05] ogra_: just got it a few hours ago [12:05] (there are a bunch of logging bugs that make the kern.log grow gigantically) [12:05] ok === _salem is now known as salem_ [12:05] ogra_: first thing after I'm out of adb [12:05] er [12:05] fastboot [12:05] zyga, and did it ever run ubuntu before? [12:05] ogra_: nope [12:05] you know you need to boot once into android completely ? [12:05] ogra_: I didd [12:05] * ogra_ is sure the install page says so [12:05] hmm [12:06] ogra_: got to the home screen and started the process from there [12:06] ok, that should be fine ... [12:06] ogra_: ok, flashing now [12:06] ogra_: this is what I have so far, looks normal: http://paste.ubuntu.com/7184505/ [12:07] ogra_: offtopic, is it possible to fix adb shell to return the process exit code? :) (there's an open bug about that on google and untouched for a few long years) [12:08] you mean to the host PC ? [12:08] ogra_: I see this now http://paste.ubuntu.com/7184511/ [12:08] ogra_: adb shell false; echo $? [12:08] ogra_: I mean the exit code is useless [12:08] ogra_: the scren has spinning ubuntu logo === alan_g is now known as alan_g|lunch [12:08] zyga, very tricky, adb does not sccess the device itself, there is a proxy [12:08] zyga: ogra_fwiw; after unlocking, boot once into android before flashing [12:09] sergiusens: hey, glad to see you [12:09] sergiusens: I have goodies for you later [12:09] sergiusens, yes, he apprently did [12:09] sergiusens: I did boot once, all the way to home screen [12:09] thats why i asked :) [12:10] sure, I'm eavesdropping and haven't read through the whole cycle [12:10] ogra_: back to 'recovery v6.0.4.6' [12:10] zyga: sounds good :-) [12:10] zyga, thats odd, it should boot into normal mode after uncompressing (teh dancing logo) [12:11] so what now? [12:11] sergiusens, any idea ? [12:11] thats a brandnew flo [12:13] ogra_: zyga check the logs in /cache/recovery/ [12:14] checking [12:14] http://paste.ubuntu.com/7184534/ [12:14] that /cache/recovery/log btw [12:15] (before killing android I did enable developer mode) [12:15] (I hope that is booting enough) [12:16] if you need any of the other log files then please tell me [12:16] you need to enable dev mode and unlock the bootloader ... then boot once into android and you should be fine [12:16] ogra_: I did all of that [12:17] right [12:17] ogra_: including the bootloader unlock [12:17] there seems to be a stale autodeploy.zip [12:18] I:using /data/media for /sdcard/autodeploy.zip. [12:18] W:failed to mount /dev/block/platform/msm_sdcc.1/by-name/userdata (Invalid argument) === MacSlow|lunch is now known as MacSlow [12:19] ogra_: anything actionable for me? [12:20] try removing it in recovery mode ... but i'd liek to hear what sergiusens thinks ... he is more familiar with the recovery logs [12:20] ogra_: remove what? autodeploy.zip? I cannot see it from adb shell [12:20] under sdcard ? [12:20] ogra_: /sdcard is empty [12:20] ok [12:20] then thats moot :P [12:21] it's a symlink to /data/media btw [12:24] stgraber, ogra_ are the any known issues with using container overrides? I could place files in the rootfs using them, but not overwrite init.rc, I had the phone (not a nexus) lock up on boot if I tried that [12:24] janimo, hmm, works for me usually [12:26] ogra_, ok thanks. Another file I tried overwriting in system/etc, while it did not get there on the rortfs at least that did not block the boot [12:26] janimo, ah ! [12:26] janimo, you didnt say /system ;) [12:27] ogra_, well both init.rc which is in root [12:27] thats indeed impossible without hacking up the mount code in initrd [12:27] and one under system/ [12:27] sergiusens: can you help me out with this, is there anything I should/could do or should I ask you some other day? [12:27] system is mounted readonly from the initrd [12:27] ogra_, init.rc was weird even if I tried replacing using the same copy of init.rc it would hang on boot [12:27] hence my asking whether there are known gotchas specific to touch [12:27] thats indeed weird, were the permissions correct ? [12:28] (for the file you copied in place) [12:28] ogra_: zyga no need for dev mode to flash [12:28] sergiusens, oh, indeed, we should remove that from the instructions i guess [12:28] ogra_, +x on AFAIK, yes. I will try again though [12:28] ogra_: zyga that autodeploy zip is just a warning; nothing else; rsalveti made me keep it for his debugging; it's obsolete for anyone else ;-) [12:29] janimo, yeah, +x they should be [12:29] sergiusens, well, any idea why zyga cant boot then ? [12:29] zyga: this is what I'd do: oem unlock; boot android; reboot to bootloader; ubuntu-devce-flash --channel devel --bootstrap; wait [12:30] sergiusens: that's exactly what I did 20 minutes ago, arriving here [12:30] zyga: from your logs it seemed it bootstrapped fine though [12:30] sergiusens: any other log I can pull out for you? [12:30] lines 321 to 334 from the first pastebin [12:31] zyga: hmmm, only thing I can think of is flashing stock android before [12:31] zyga: might be some partition label mess; ogra_back to you :-P [12:31] sergiusens: ok, let my do that now [12:33] stgraber, around ? [12:33] thanks, I'll stay in touch, ping you back when android is up again === josepht_ is now known as josepht [12:38] quick sanity check: https://developers.google.com/android/nexus/images#razor <= that's flo? [12:38] razor is nexus 2013? [12:38] nexus 7 2013 [12:39] sergiusens: ^^ [12:39] zyga: yes [12:39] thanks [12:39] flashing [12:39] zyga: os build vs hw [12:39] ok, android seems to work [12:39] sergiusens: ah, I see [12:52] sergiusens: ok, android restored [12:52] sergiusens: I guess trying to flash ubuntu again won't do anything new [12:52] sergiusens: do you want me to poke around the system that I'm running now for clues? [12:52] sergiusens: runnin android 4.4.2 now [12:53] zyga: I was expecting you to flash ubuntu now ;-) [12:53] sergiusens: wait, you asked me to restore android before, didn't you?\ [12:53] zyga: yeah, stock :) and then ubuntu === blaroche_ is now known as blaroche [12:53] sergiusens: ok, doing ubuntu again [12:54] sergiusens: do I need the PPA on trusty? [12:54] sergiusens: that's one of the steps I didn't do [12:54] nope [12:56] rsalveti, stgraber, i would like to drop the unpacking of the android initrd from lxc-android-config and instead merge it into the gzipping the initrd already does ... http://paste.ubuntu.com/7184677/ ... what do you guys think ? [12:57] sergiusens: hmm, it doens't turn back on [12:57] (saves us one gzip run during boot) === alan_g|lunch is now known as alan_g [12:58] sergiusens: it seems to be dead [12:58] sergiusens: I just powered it off from android [12:58] sergiusens: and now it's gone [12:58] zyga: no juice? [12:58] sergiusens: ah, it's on again :| [12:59] sergiusens: no, it was charged [12:59] sergiusens: I don't know what happened, I got scared for a second [12:59] sergiusens: I thought I wiped the bootloader somehow and that reboots weren't rebooting enough [12:59] ok, it is working [12:59] back to boot loader [12:59] and flashing again [13:00] sergiusens: is secure boot a problem? [13:00] ogra_, when you get a chance, https://code.launchpad.net/~mterry/ubuntu-touch-session/move-lightdm/+merge/211445 will help OEM customization of boot by letting them override lightdm config files with /custom ones [13:00] sergiusens: or modemmanager? (poking at all the serial lines) [13:02] sergiusens: ok, it's unpacking the tarball on the device now [13:03] sergiusens: ok, it's making SWAP, [13:03] sergiusens: and rebooted [13:03] sergiusens: and.... [13:04] sergiusens: it seems to work :D [13:04] sergiusens: I wonder what the problem was [13:04] sergiusens: adb shell looks much better than before [13:05] sergiusens: click installs stuff [13:05] ahhh [13:05] the beauty :) [13:05] sergiusens: it booted ok [13:06] :) [13:07] stgraber, ping [13:08] mterry, looks fine (as long as it works) ... i assume it needs changes to writable-paths too if the customization teams wants to put stuff into that dir ? [13:08] * cwayne_ perks up [13:11] ogra_, I feel like the customization team is already on top of that? They've been planning to use /custom for all sorts of stuff for a while, right cwayne? [13:12] i wouldnt know why they need to modify the greeter startup though [13:12] i dont think we have anything writable in there yet [13:12] mterry, yeah, we're allowed to write there, but once the phone's booted its r/o [13:13] wait i dont see anything in that MP about /custom.. [13:14] cwayne_, right. But what it does is move a file to a directory that is read before /custom is. So that you can actually override its contents with your own file if you want [13:15] ogra_, well, once we have a boot animation going, the lightdm config may be a convenient place to override that. Or whatever. I don't want to pre-judge what lightdm changes are going to be needed [13:15] ah, that makes more sense :) [13:19] fginther, ping [13:21] cwayne_, pong [13:23] fginther, hm sorry, never mind :) [13:26] cwayne_, just for future reference, please try to ping the ci vanguard in #ubuntu-ci-eng for ci requests or questions [13:30] hi, does anyone know how to put a N7 in landscape mode for testing purposes? I think there is a device-specific flag somewhere, but I'm not sure where it is [13:32] dpm: /etc/ubuntu-touch-session.d/ [13:32] but I dont think it works.. [13:33] popey, so assuming it'd work, what would I need to change in that dir? [13:35] dpm: the one for the device, in your case, flo.conf [13:35] but it already says landscape, which makes me think it doesn't work [13:36] ah, ok, thanks popey [13:36] right [13:36] it needs corresponding unity8 changes we dont have [13:38] sergiusens, hey, any idea why my custom tarball isn't being applied? [13:39] cwayne is this a new problem? [13:39] yeah [13:39] and it only seems to happen with trusty-proposed-customized-demo [13:40] my /cache/recovery is still completley populated too, looks like nothing was applied [13:43] sergiusens, ^ [13:44] cwayne is it too big? [13:44] he customization; the recovery partition is not that huge [13:44] the* [13:44] wow [13:44] skipped words like crazy [13:45] cwayne is the customization done bigger than before; the recovery partition size is not that great [13:45] ah, yeah, this particular custom image is much larger [13:45] as it's got all the demo bits in it [13:46] sergiusens, what's the size limit? [13:46] ogra_: have that info handy? ^^ [13:47] nope [13:48] * sergiusens looks for a device [13:48] seems to be 220M [13:48] err [13:48] /dev/disk/by-partlabel/cache 552M 11M 542M 2% /android/cache [13:48] sorry that was totally off [13:48] yeah [13:49] cwayne ^^ [13:49] you can't use all of it either [13:49] cwayne_: pong [13:50] pmcgowan: bug of the day https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1300236 [13:50] Ubuntu bug 1300236 in unity8 (Ubuntu) "Click Scope is showing ? instead of characters" [High,Confirmed] [13:50] stgraber, im having problems with trusty-proposed-customized-demo but i think it might just be because of the size [13:50] ogra_: so that'd work, however note that this means that rebooting the Android container may not work then (we obviously don't do that automatically but some developers might) [13:51] stgraber, rebooting it will work just not replacing the content [13:51] while installing ubuntu phone | http://askubuntu.com/q/441513 [13:51] ogra_: mounting a clean tmpfs at every container boot was done to simulate Android's ramfs as closely as possible (where if you reboot a standard Android system, you loose the content of the ramfs) [13:51] stgraber, right, but we waste .5 seconds to running gzip twice [13:52] and if there are no massive drawbacks i'l like to make that change, just looking for thiongs i could overseee here [13:53] ogra_: so long as developers always powercycle the whole device when they want to test an Android change, I don't see any obvious drawback with that change [13:53] wow, our lxc-android-config job needs quite some changes to fail gracefully (instead of hanging hard) [13:53] stgraber, well, developers could even stop the container, unzip a new initrd and start it again === oSoMoN_ is now known as oSoMoN [13:54] its a rw tmpfs after all [13:54] just one step more for them ... but one less for us on boot :) [13:54] sergiusens, did this size change recently? [13:54] zyga: hey, good to know it worked; not sure if the OEMs put in some stuff that we don't know about to break us; so flashing stock images always clears that up and runs/puts blobs we know about [13:55] sergiusens: yeah [13:55] cwayne you change check the ogra stats; not sure [13:55] sergiusens: though this is google :) [13:55] sergiusens: so it should be oem-free [13:56] stgraber, so the CPU issue ... i think when we dont use cpu cgroups the kernel allocates two cores by default to teh container, i'll try to enable that in a test kernel to see if i can force it to all four (for the boot) [13:56] stgraber, oh, and another thing, lxc-android-config has: "start on started cgroup-lite" did i see that being removed with a recent upload ? do we need to adjust anything ? [13:57] zyga: right; but google doesn't put in the real factory image on ;-) I bet there are deltas [13:57] ogra_: no, if you don't specify any cgroup limit, you inherit from the root cgroup which is unlimited [13:57] stgraber, hmm, then i dont get why it only uses 50% [13:58] ogra_: cgroup-lite still technically exists, however it's correct that it's not longer required as we now use cgmanager. So you could replace "start on started cgroup-lite" by "start on started cgmanager" and then then drop the cgroup-lite dependency since cgmanager is already a dependency of liblxc1 [13:58] i wonder if the fact that we use an android kernel might change that inheritance [13:58] stgraber, will do [13:59] ogra_: it shouldn't, but there may be a scheduler setting that may help, let me try to find it [14:00] ogra_: you may want to try "echo 0 > /proc/sys/kernel/sched_autogroup_enabled" [14:01] root@ubuntu-phablet:/# cat /proc/sys/kernel/sched_autogroup_enabled [14:01] cat: /proc/sys/kernel/sched_autogroup_enabled: No such file or directory [14:02] without that (and for mostly good reasons), the task scheduler considers a whole LXC containers as a single process, so during a busy boot, this may impact boot performance when there are boot dependencies between the host and the container [14:02] ah, then I guess it's off :) [14:02] i guess we miss a kernel option here [14:03] it may depend on the scheduler in use, not sure [14:03] * stgraber looks at the code [14:03] stgraber, oh, looking at cgmanager.conf, can http://paste.ubuntu.com/7184917/ be dropped from lxc-android-config.conf ? [14:04] probably, yes [14:04] #ifdef CONFIG_SCHED_AUTOGROUP [14:04] if that option isn't set, you get the equivalent of = 0 [14:05] ok [14:06] hmm, no /dev/cpuctl when i remove the code [14:06] so i guess we'll keep it [14:07] :) [14:07] ogra_, it feels like we need a dep between lxc-android-config and ubuntu-touch-session, the ones that disable surfaceflinger, don't we? [14:07] well ... [14:08] stgraber, hey, how often does system-image get updated with new builds [14:08] theoretically we do ... practically nobody uses these packages standalone :) [14:10] cwayne_: every 5 minutes [14:11] ogra_: well, that code won't work once you drop cgroup-lite... [14:11] ogra_: you'll need to mount cgroupfs yourself if things actually expect it to be mounted in /dev/cpu === beuno_ is now known as beuno [14:12] stgraber, hmm, whats the drawback to keep it as is ... the start of the container has significant influence on the overall boottime [14:12] stgraber, hm ok, still waiting on custom-demo-45 to show up, will wait for it [14:12] starting only after cgmanager will add significant delay i fear [14:13] ogra_, yeah, but I just mean that upgrading only one or the other will break [14:13] ogra_, or do you mean that we're not at the point when we care about such dep intricacies and just rely on the silo to land everything together? [14:13] OTOH you can't upgrade lxc-android-config anyway... [14:14] ogra_: why? both cgroup-lite and cgmanager share the same start condition so the actual boot time difference should be negligible (in either direction, I'm not actually sure which would be faster) [14:14] Saviq, lxc-abdroid-config only puths the file in place, it can well be chnaged and uploaded after the session stopped using the file [14:14] ogra_: your existing mount --bind could be replace by "mount -t cgroup cgroup /mnt -o cpu" and the "[ -d /sys/fs/cgroup/cpu ]" dropped [14:15] bzoltan: will you be able to help me build Trojita in click chroot? I'm still getting an error about it not finding GLESv2 [14:17] [10:17:13] CMake Error at /usr/lib/arm-linux-gnueabihf/cmake/Qt5Gui/Qt5GuiConfigExtras.cmake:16 (message): Failed to find "GLESv2" in "" with CMAKE_CXX_LIBRARY_ARCHITECTURE "arm-linux-gnueabihf". [14:17] I'm on Saucy, click choot is for Trust [14:18] Trusty === chihchun_afk is now known as chihchun [14:22] mhall119: is a choot like a chroot /me wonders off quickly ;) [14:22] stgraber, right, no difference when i change the start condition to cgmanager (just quickly did two bootcharts) [14:22] ogra_, it also disables surfaceflinger [14:22] ogra_, does it not? [14:23] Saviq, yes, but it does that based on the file it puts in place [14:23] which it does now anyway [14:23] ogra_, yeah, but if someone removed it [14:23] later the check for the file will be dropped [14:23] ogra_, it will now disregard it [14:23] yes [14:23] ogra_, and disable surfaceflinger anyway [14:23] like it will in the future [14:24] the package the file creates also uses teeh check [14:24] *the [14:24] ogra_: good to hear! === dandrader is now known as dandrader|afk [14:24] ogra_, while non-upgraded ubuntu-touch-session would still set up env to expect surfaceflinger running [14:24] ogra_, I think this is what should happen: http://pastebin.ubuntu.com/7184985/ [14:24] Saviq, right, lxc-android-config needs to land after your changes [14:24] ogra_, disregard the first three, just look at l-a-c and u-t-s [14:24] ogra_, no, it actually needs to land together [14:24] Who is the best person to talk to about how app descriptions are rendering in the apps scope? [14:25] Saviq, huh ? why ? [14:25] Saviq, you drop all code that cares for the file from the session [14:25] ogra_, yeah, but if l-a-c still does care [14:25] ogra_, surfaceflinger will start [14:25] so it doesnt matter when l-a-c loses it [14:26] ogra_, and mir will fail to start [14:26] why would sf start ? [14:26] ogra_, if someone has the file removed [14:26] * ogra_ doesnt get it [14:26] *we*put the file in place [14:26] on boot [14:26] from the very same package [14:26] ogra_, the last change you uploaded to l-a-c, does it not check for .display-mir? [14:26]