=== rww_ is now known as rww [00:08] stgraber: 56 and still going, off to bed now.. will leave it going [00:08] popey: that sounds much better! [00:08] yeah [00:08] popey: here I just got past 100 and it's still going [00:08] sounds great, thanks! [00:33] stgraber: bah, just went to walk away and noticed it's wedged 13 mins ago [00:33] stgraber: 73 loops [00:34] stgraber: adb devices shows nothing === fginther|away is now known as fginther [00:45] popey: hmm, I really don't get why we're not seeing the same thing... Here I'm at 148 and still going... if there's still a bug I'd really like to also hit it so I can actually fix it === chriadam|away is now known as chriadam [00:59] stgraber: what can i get off my device to debug? [00:59] dammit it just booted, not into bootloader even though i had my hand on the vol button [01:00] stgraber: is wifi enabled on yours, 3g? does it have a sim? [01:00] stgraber: mine has wifi on, and a 3g enabled sim on board. fwiw [01:16] popey: I've got wifi on and a SIM inside, data isn't enabled though (foreign sim card) [01:16] not that it really should matter that early in the boot sequence... [01:25] stgraber: yeah, wild speculation ☻ [01:28] popey: getting /var/log/upstart/cgmanager.log and /var/log/upstart/cgproxy.log after the hang may help, though if it's like last week, the chances that the file is actually synced by the time you force the reboot aren't particularly high... [01:37] stgraber: ok, gonna leave it now and if it's barfed by the morning i'll try that thanks. === sarnold_ is now known as sarnold === axisys is now known as info === info is now known as axisys === vying is now known as Guest63859 === chihchun is now known as chihchun_afk === Stskeepz is now known as Stskeeps [04:31] hello [04:32] maybe someone here can help, i am trying to flash a Nexus 7 flo with ubuntu touch [04:32] i installed everything, bootloader is already unlocked [04:32] i plugged in and did adb reboot bootloader [04:32] but when i try to do the bootstrap [04:33] it’s stuck at “Expecting the device to be in the bootloader…. waiting" [04:33] tried googling and stuff but nothing :| === chihchun_afk is now known as chihchun [05:37] o/ all [05:42] Hey, rickspencer3! [05:42] hi RAOF [05:43] I'm in Beijing if youa re wondering what I am doing up at such an hour :) [05:44] Ah. [05:44] Not London with everyone else? :) [05:44] RAOF, I'm not sure whoe "eveyrone else" is, but yeah, guess not :) [05:45] I guess it's only 6:43 there :) [05:49] Oh, are they on daylight savings already? [05:54] RAOF, dunno, I always just ask Google what time it is somewhere [05:54] :_ [06:04] Europe switched a week ago too === rickspencer3_ is now known as rickspencer3 [06:52] good morning [06:57] hi [06:57] [07:00] pwd [07:00] su [07:01] c,i [07:04] how can i want to install ubuntu on galaxy s3 [07:58] 2014/04/08 08:56:00 Start pushing /home/mpt/.cache/ubuntuimages/ubuntu-touch/devel/mako/version-250.tar.xz to device [07:58] 2014/04/08 08:56:00 Cannot push /home/mpt/.cache/ubuntuimages/gpg/image-master.tar.xz to device [07:59] Dear #ubuntu-touch, what do? [08:00] stgraber: ogra_ did this and left over night http://paste.ubuntu.com/7219206/ - it's now on 540 loops.. [08:00] It doesn’t give me any other error message. The phone is on the “Recovery” screen. [08:00] mpt: what did you run? [08:04] popey, “ubuntu-device-flash --channel=devel --bootstrap” (and yes, it was at the bootloader to start with) [08:05] Hmmm, the wiki page contradicts itself [08:05] oh? how so? [08:06] popey, says (a) “the recommended command is to be run from the bootloader: ubuntu-device-flash…” and (b) “ubuntu-device-flash will not work unless you have booted your device (it must not be displaying the boot loader screen…)” [08:06] So it both must and must not be at the bootloader [08:07] it depends if you're bootstrapping i believe [08:07] if you are then it must be in the bootloader, if not, it doesn't [08:08] I’m not sure what bootstrapping is, but if it’s flashing the device for the first time, I’m doing that and I was at the bootloader [08:11] mpt: yes, that. [08:13] Ah, now it works [08:14] popey, the only thing I did different this time was check “Always trust this computer” (I don’t remember the exact wording) in the Android USB debugging prompt [08:14] how odd.. [08:15] popey, great, stop the looping then ... (250 would have been enough to convince me :) ) [08:15] i'll land the change today and we should be done with that bug [08:15] yay [08:15] * popey reflashes [08:50] :-? === vying is now known as Guest57420 [09:17] Good morning all; happy Be Kind To Lawyers Day! :-D === vrruiz_ is now known as rvr === VDVsx_ is now known as VDVsx === chriadam is now known as chriadam|away === chrisgward is now known as chrisward [10:23] Morning all === MacSlow is now known as MacSlow|lunch [11:25] dbarth, so it seems the aebapps actually hit OOM ... i just tested with 281 here (commented on the bug) [11:26] ogra_: even though they're still running? [11:27] hello [11:27] popey, the ones that the log claimed to kill are gone [11:28] can i make ubuntu touch go horizontaly on my nexus7? [11:28] the AOSP does not produce .zip to install the device builds anymore is this normal? [11:28] (this is not for the default devices btw) [11:28] popey, i think we have two bugs here ... one is the apps hitting OOM ... the other is that when killing an app not everything gets torn down [11:28] annerajb, yes, AOSP never supported zips [11:28] ogra_, how do you install the builds now then? [11:29] just flash the system.img and boot.img? [11:29] use rootstock-ng ... (see my mail to the ML from 6-8 weeks ago) [11:29] ogra_, ill search for it [11:30] bzr branch lp:project-rootstock-ng; cd project-rootstock-ng; ./rootstock-touch-install /path/to/rootfs.tgz /path/to/system.img [11:30] (and before flash boot and recovery with fastboot) [11:30] (and the device needs to be in (the new) recovery when you run the rootstock-touch-install script ... [11:30] ) [11:34] so before running rootstock-ng i flash the boot and recovery [11:34] right [11:34] then boot into recovery and hand the cdimage tarball and your self built system.img to the script [11:35] there is also a README in the branch (and the script has a --help option) [11:36] dbarth, oh, btw, are --webappUrlPatterns supposed to work yet ? seems they don't ... all links open inside the wepapps === pete-woods is now known as pete-woods-lunch [11:38] the instructions on the ml talk about a rsalvetti ppa and mir is that "needed" [11:39] no, that was for the experimental 4.4 images [11:39] ignore that [11:39] follow what i wrote above (or follow the README) [11:40] mandel: hey, did you see my comment about u-d-m not building in the PPA for non x86 arches? [11:40] sergiusens, nope, I missed it, can you point it to me? [11:40] mandel: https://launchpad.net/~ci-train-ppa-service/+archive/landing-019/+packages [11:41] mandel: summary, all non x86 builds had test timeouts [11:41] ogra_: not yet, not in #281, it's coming in silo 8 [11:41] dbarth: \o/ [11:41] dbarth, ah, fine then [11:41] ogra_: this is bug https://bugs.launchpad.net/webbrowser-app/+bug/1294279 on the webapps-blocker tag [11:41] sergiusens, oh, great, I have a fix for that I'm finishing in a few mins :) [11:41] * sergiusens mentions that the feeling is different when it's your main device [11:41] dbarth, so i think we have multiple bugs, not only one ... [11:41] sergiusens, I did notice that last night [11:43] dbarth, i think something is wrong with the way the webapps register with unity-mir/upstart-app-launch ... else OOM would properly tear down everything (and probably even quietly restart it) ... and the other is that we hit OOM at all [11:44] dbarth, and given that webapps that use to old api are not affected i dont think the issue lies within unity-mir here [11:44] Saviq, ^^^ [11:44] s/to/the/ [11:45] ogra_: hmm, ok [11:45] oSoMoN: ^^ [11:45] maybe there is an issue with the process group [11:45] i remember that we had this issue initially with qtwebkit [11:46] yeah [11:46] and also, i was suspecting a bad response to a memory pressure callback, which could incite the app managr to get rid of non cooperating apps [11:59] dbarth, hmm, so looking at my processlist i see actually two times the same webapp-container process for each webapp [11:59] once the app dies from the UI one of them is gone [12:00] (nice work with the scrollbar btw) [12:01] 3750 ? Tsl 0:12 0 41 589950 71168 3.7 webapp-container --enable-back-forward --webappUrlPatterns=https?://sueddeutsche.de/* http://m.sueddeutsche.de/ [12:01] 3776 ? T 0:00 0 41 235138 15236 0.8 webapp-container --enable-back-forward --webappUrlPatterns=https?://sueddeutsche.de/* http://m.sueddeutsche.de/ [12:01] i.e. that is what i see for a running app [12:02] after the app dies only the first one is gone from the processlist [12:07] cant flip my nexus7 horizontaly with ubuntu touch, is this normal? [12:09] lotuspsychje, the shell has no rotation support yet ... some apps do though [12:10] ogra_: ok tnx [12:10] tedg, mdeslaur: hey-- I observed the music-app needing this access: http://paste.ubuntu.com/7221312/ [12:11] tedg, mdeslaur: I was wondering if you guys knew what it was and if that would be safe for all apps to do? [12:11] tedg, mdeslaur: (ie, considering adding it to the templates) [12:13] music-app is a dbus service? [12:13] sounds fishy === cul- is now known as cul_ [12:20] ogra_, well, we rely on OOM to kill apps that have been stopped... [12:21] Saviq, right, but only one of the grou is killed [12:21] *group [12:22] ogra_, mhm [12:22] (i dont get at all why i see two webapps-container processes) [12:22] (for each running app) [12:23] hi, is there anyone here that can help me? I'm creating a "qml extension library + tabbed ui" in qt creator... but i'm not sure how to execute that app so it opens the proper window with the ui... i don't know if there is any missing default setting or what [12:23] Saviq, running an old webapp that uses the 13.10 api i only get oone process [12:23] dbarth, ^^^ [12:24] ogra_, you should get another process that's rendering anyway [12:24] cjwatson: hi, do you know if there's any documentation on how to use "click chroot" to create a chroot that's ok for building apps for our latest image ? [12:24] Saviq, right, but not two "webapp-container" processes with identical commandlines [12:25] Saviq, ogra_: i can get that too [12:25] ogra_, yeah [12:25] 2 webapps-container's running ebay, with their oxide render children [12:25] right [12:26] ogra_, I only have one container... === _salem is now known as salem_ [12:26] and no separate rendering process... [12:26] Saviq, for an app thats using the new API ? [12:27] ogra_, webapp-container --webapp --enable-back-forward --webappUrlPatterns=https?://m.jakdojade.pl/* http://m.jakdojade.pl [12:27] and no oxide processes? [12:27] ogra_, do you need qemu-arm-static as it says on the ml? [12:27] annerajb, yup ... [12:27] which ppa is that on?? [12:27] annerajb, oh, wait [12:27] annerajb, for installing you dont [12:27] ogra_, dbarth right, started g+, two containers [12:27] ah ok thanks [12:27] only for the rootfs build script [12:28] Saviq, yeah, i guess jakdojade uses the 13.10 API [12:28] ogra_, probably, yeah [12:29] smells like webapp-container re-execs itself or some such [12:29] ogra_, dbarth yeah, I can confirm webapp spawns twice [12:29] http://pastebin.ubuntu.com/7221375/ [12:29] and unity-mir cant catch that apparrently [12:29] ogra_, OTOH when I stopped it, both went away [12:30] ogra_, but it might be that if one dies, that's when it goes fishy [12:30] Saviq, right, but not if it gets killed automatically [12:30] ogra_, yup, we don't really support multi-process apps yet, the web process is whitelisted [12:30] i guess the killing instance simply doesnt know about the second process ... and the partent doesnt tear it down alongside [12:31] nerochiaro: "click chroot -a armhf -f ubuntu-sdk-14.04 create" is supposed to do it [12:31] teach the parent to tear it down and i guess we have at least one issue fixed [12:31] cjwatson: is this process documented anywhere ? [12:32] cjwatson: i'm trying it now, thanks [12:32] nerochiaro: "man click" to some extent. but you aren't really supposed to need to do it directly, normally, QtCreator is supposed to take care of it for you [12:32] mdeslaur (and tedg): well, it seemed related to the whole GApplication discussion-- but I don't know, which is why I asked [12:33] cjwatson: forcing people to rely on a GUI tool isn't that great, and even in that case it should be documented online somewhere [12:34] cjwatson: just my 2 cents [12:34] nerochiaro: sure, whatever [12:34] nerochiaro: I did give you a pointer to a man page as well [12:34] nerochiaro: I'm not forcing anyone to do anything [12:35] jdstrand: no idea what the security impact of that is, need to wait for tedg to explain why it's needed [12:35] cjwatson: sorry if it came up as a complaint, i was just giving out a suggestion so that things would be easier for all developers interested to use the platform [12:36] nerochiaro: I don't think me documenting a single line would be helpful; it should be documented in the context of a broader workflow, which is beyond what I can help with [12:37] (hm, I'd probably better change the default for -f in time for release ...) [12:37] oSoMoN, dbarth: hi guys! [12:38] cjwatson: yes, i wasn't implying you did that yourself or just documenting that single line. i was just wondering if that workflow documentation already existed somewhere and if not who would be responsible for writing it. perhaps dpm knows ? [12:38] mdeslaur (and tedg): seems like https://bugzilla.gnome.org/show_bug.cgi?id=699259 and http://standards.freedesktop.org/desktop-entry-spec/desktop-entry-spec-latest.html#dbus [12:38] it would be something in dpm or mhall119's area, I guess [12:38] on developer.ubuntu.com [12:38] oSoMoN, dbarth: so, seb128 noticed that there are 2 landings that have webbrowser-app locked in, it seems robru ignored conflicts for one of those - is there any chance we can merge those two landings into one? [12:38] hi nerochiaro, what's up? [12:38] oSoMoN, dbarth, sil2100: or land one and then rebuild the other silo [12:38] dpm: see scrollback [12:38] jdstrand: yeah, sounds really scary, wonder why it's needed for that app [12:38] Yeah, that's how usually it's done [12:39] dpm: hi, is there any documentation anywhere on how to build a click app from the command line, install it to the device, run tests on it, etc ? [12:39] seb128: but this approach prooved a bit flawed, since people 'ignore' warnings and in the morning we had to fix a package that was broken because of that [12:39] dpm: i found something about tests in the wiki https://wiki.ubuntu.com/Touch/Testing but that's about it [12:39] seb128: so I would prefer to merge those two requests into one for clarity... [12:39] sil2100, yeah, that works for me [12:40] dpm: i feel very lost, coming from a world of debs and having to start working on click packages. missing a proper document describing the whole workflow [12:40] mdeslaur: I'm not sure why it is scary. I think this is so that if an app is launched by dbus, it can be passed url and things. the thing is, if all apps bind to org.freedesktop.Application, how are they differentiated when something is sent to them? [12:40] mdeslaur: definitely need more info [12:40] jdstrand: the scary part is that the app gets launched by dbus..why would we want that? [12:41] we don't on touch with application lifecycle (and we block it currently). but I don't know about on converged [12:41] nerochiaro, what i do for my webapps is: create manifest, .json and .desktop file in a dir, add an icon to that dir too, copy it to /home/phablet/ on the device, adb shell in ... cd to that dir and run "click build" [12:41] sil2100: i know, trying to make room [12:41] ogra_: urgh, ideally don't recommend on-device building to people [12:42] that's not the intended long-term workflow [12:42] cjwatson, ok [12:42] saves me from installing click on my desktop ... i'm just lazy [12:42] eh, apt-get install click, is it hard? :) [12:43] nerochiaro, yes, we don't have any documentation on click packages on d.u.c right now, other than cjwatson's click docs on readthedocs. I agree we should have at least something on the site, but it's been low on our priority list, as we've been working on the HTML5 and scope docs recently. Would you mind filing a bug about it on https://bugs.launchpad.net/ubuntudeveloperportal/+filebug ? I can't promise we can address that straight away, but we've got o [12:43] ur next docs day this Thursday, where we'll go through the bugs in the site and prioritize things [12:43] for a webapp there's absolutely no point copying things around, click build on a desktop will work just as well [12:43] yeah, when i started that wasnt possible iirc ... and the procedure stuck [12:43] ogra_, the script has been on transferring rootfs for a while now. I know the rootfs is big 500mbs but is been more than 10 minutes [12:44] annerajb, give it another few [12:44] kk [12:44] jdstrand: what do you mean by "how are they differenteiated"? [12:44] ogra_: um, ok, I guess there's no point trying to debug historical problems but I've had click installed on my desktop since roughly day zero [12:45] only packagekit-plugin-click is problematic [12:45] jdstrand: it's the interface name underneath their own object path, no? [12:45] well, i remember it wasnt possible to get it on precise initially ... and my laptop doesnt run trusty for that long yet [12:45] oh, precise, sure [12:45] for you fossils [12:45] haha [12:45] mdeslaur: music-app is binding to org.freedesktop.Application [12:45] my desktop has so many steam games :) [12:45] jdstrand: under what path? [12:45] dont want to risk breaking them [12:45] I didn't think bind had a path [12:45] let me see [12:46] jdstrand: look at gedit in d-feet for an example of org.freedesktop.Application [12:47] mdeslaur: this is the denial: apparmor="DENIED" operation="dbus_bind" bus="session" name="org.freedesktop.Application" mask="bind" pid=3544 profile="com.ubuntu.music_music_1.3.389" [12:47] mdeslaur: I guess what would happen is after it binds to that, then its path would need to be used for its methods [12:48] dpm: ok, will file a bug. i would suggest we really make it a priority though, otherwise how do we expect developers to be able to develop for our platform ? [12:48] jdstrand: you don't see paths when you get bind requests? [12:48] and if we were to allow this (still not saying we should), then we would have another rule with the path [12:48] mdeslaur: I pasted what I see [12:48] nerochiaro: it's worth noting you only need a chroot for packages with native code === pete-woods-lunch is now known as pete-woods [12:49] mdeslaur: maybe music-app is doing it wrong, I don't know [12:49] cjwatson: good point, this one has it [12:49] Porting Ubuntu Touch On Samsung Galaxy S2 Plus | http://askubuntu.com/q/444953 [12:49] which gets back to asking about it :) [12:49] nerochiaro, they would in principle use Qt Creator, which would do it for them. I agree that it's worth documenting, but it's not that anyone is blocked on that [12:49] how can i change default root password on ubuntu touch? [12:49] cjwatson: so once i create the chroot, click-buddy --arch armhf from the branch directory will build it properly, right ? [12:50] I'm afraid I don't know about click-buddy [12:50] You'd need to ask its author [12:50] ok, sorry [12:51] i.e. sergiusens [12:51] sergiusens: just looking at click-buddy source, were you aware that session support in click chroot landed a little while back? [12:52] nerochiaro: ah, looks like today click-buddy won't *quite* work because it relies on the default value of click chroot -f and that's currently ubuntu-sdk-13.10 ... [12:53] let's see if I can at least make that possible [12:53] cjwatson: dpm: i see, so from the qtcreator, after opening the CMakeLists.txt, how do i build a click package and deploy it to the device ? [12:54] build seems to build locally [12:54] jdstrand: now I'm confused...music-app is qml, right? [12:56] mdeslaur: it is qml with C++ [12:57] nerochiaro: https://code.launchpad.net/~cjwatson/phablet-tools/click-buddy-pass-framework/+merge/214744 ought to at least make "click-buddy --arch armhf --framework ubuntu-sdk-14.04" work, I think [12:57] jdstrand: oh, any idea where the C++ parts live? /me can't find in tree [12:57] nerochiaro, the UX is a bit clunky on the QtC version on the archive, but this is going to change very soon (pending a FFe). Right now, you'll need to 1) create a click chroot first with Tools > Options > Ubuntu > Create click target [12:58] nerochiaro, and then 2) Build > Build in chroot [12:59] mdeslaur: I'm told lp:music-app is the bzr branch. vthompson would know more [12:59] ogra_, it finished. it is now stuck on the LG logo screen. [13:00] mdeslaur, jdstrand there are no C++ parts in lp:music-app, they live in separate projects. Which bit are you looking for? [13:00] dpm: we're looking for whatever is trying to bind org.freedesktop.Application? [13:00] mdeslaur, what do you mind by bind? [13:01] err, *mean, sorry [13:01] ogra_, when you flash the boot.img and system.img does it modify anything that the recovery uses? [13:01] dpm: music-app is trying to bind to org.freedesktop.Application on dbus [13:01] since recovery was working before running rootstock-ng [13:02] mdeslaur, I've no idea what would do that, sorry. In any case, the only C++ plugin we're using is at https://code.launchpad.net/qtgrilo (mediascanner 1) - although we're migrating to the new mediascanner 2 and we're trying to drop that soon [13:03] annerajb, you flashed recovery and boot with the files crated by your build ? [13:03] i flash recovery and it was working fine and showed the ubuntu logo [13:03] then i flash boot and the rootstock-ng script [13:04] and can't seem to get into recovery [13:04] well, rootstock doesnt touch the recovery at all [13:04] ogra_, does flashing boot.img affect the recovery?? [13:05] not if you flash to the right partition [13:05] yeah should be flashed to the right onw [13:05] one* [13:05] well, rootstock doent touch anything but the content of the data partition ... [13:06] *doesn't [13:06] ok then ill have to figure out if i am hitting it right [13:06] dpm: i don't have "Ubuntu" under Tool > Options [13:06] does recovery uses it's own recovery kernel? ogra_ [13:06] its a boot.img usually [13:06] but that may vary between devices [13:07] ok got it it was the key combo [13:07] (boot.img = kernel plus initrd) [13:07] now to figure out why it dosnt work [13:07] boot again [13:07] til it hangs [13:07] jdstrand: my best guess for now is that some component it's using is doing it wrong...it should be trying to bind that [13:07] then reboot directly into recovery again [13:07] (make sure it is really the direct newxt boot) [13:08] then adb shell cat /proc/last_kmgs >messages.txt [13:08] nerochiaro, are you on 14.04 and using Qt Creator from the archive? [13:08] ogra_, where was the temp file that holded the logs /proc/last_kmsg? [13:08] k [13:08] that should have the log of your last boot [13:08] *kmsg [13:09] jdstrand: sorry, meant "should not be trying to bind that2 [13:10] if i built the kernel without swap and you guys are using swap could it hang or does it fallback into using memory? === VargaD_ is now known as VargaD [13:11] it should just fall back [13:11] dpm: yes, i'm using 3.0.1-0ubuntu3 [13:11] ogra_, just realized my device does not appear to be on a boot loop it just stays on the logo [13:11] it dosnt turn off and go back to it [13:12] thats good [13:12] did you check if you have adb ? [13:12] yeah it dosnt show up [13:13] also i just remembered ogra_ do we still have to modify the kernel cmdline for the console? [13:13] like setting console=tty1,S115200? [13:13] that would have gotten you a reboot loop [13:13] so check last_kmsg [13:13] see how far the boot got [13:14] there is no last_kmsg [13:14] hmm [13:14] zbenjamin, could you perhaps help nerochiaro with http://pastebin.ubuntu.com/7221533/ ? I'm using a non-released QtC, so I don't know how the UI looked before that [13:14] did you disable it in the kernel config ? [13:14] ogra_, which flag is it? [13:14] dunno, just know there is one :P [13:15] ogra_, http://stackoverflow.com/questions/9682306/android-how-to-get-kernel-logs-after-kernel-panic [13:15] annerajb, RAM_CONSOLE [13:15] make sure thats on [13:16] it is but none of the other ones on the stackoverflow post i linked are on the config [13:16] dpm: if there is no Ubuntu options page he maybe does not have the qtcreator-plugin-ubuntu installed [13:17] nerochiaro, ^ [13:21] nerochiaro, you need to run the "ubuntu sdk" not qtcreator [13:22] actually that seems ok now [13:26] mhall119, can you update the content-hub docs on d.u.c? [13:29] didrocks, I actually had a dream that my phone had a system update available last night [13:32] mterry, you should seriously see a psychologist :) [13:33] rsalveti: ping [13:33] dreaming of work isnt a good sign :) [13:33] ogra_, :) [13:43] popey, ogra_: similar result here, left it running overnight and got up to around 380, then it eventually got stuck though looking at it it's because the kernel of the box it's plug into decided to shutdown the port for some reason :) [13:43] stgraber, yeah, i'll land the fix today [13:44] just need to coordinate with the telephony landing which uses lxc-android-config too === dpm is now known as dpm-afk [14:06] kenvandine: doc import is failing, there's no ResponsiveGridView HTML being generated [14:06] running "make doc" [14:07] mhall119, make doc is failing? [14:11] no, the import is failing because that qml class is referenced in the index, but the doc file for it doesn't exist [14:12] it's in ubuntu-content-qml-api.index [14:13] mterry: I stopped dreaming, I live in reality! :) [14:13] mandel: https://launchpad.net/~ci-train-ppa-service/+archive/landing-019/+build/5889264 [14:14] tedg: ping [14:15] mhall119, Good morning. [14:15] good morning tedg, does upstart-app-launch change the $XDG_ variables? For example $XDG_CONFIG_HOME, to point to a location that the confined app can write to? [14:17] mhall119, Well, we make sure it is set correctly, but that directly isn't a place that they can write to. [14:17] mhall119, For instance they can write under their own package name in that directory. [14:18] didrocks, :) one of these days we'll promote an image! [14:18] mterry: let's see and put a candle [14:21] tedg: oh, so ~/.config// is writable? [14:21] mhall119, I believe so, though jdstrand would be the best guy to ask on that. [14:22] is it the full APP_ID? or just the last part? [14:23] mhall119, You can also look in /usr/share/apparmor/easyprof/templates/ubuntu/1.1/ubuntu-sdk [14:23] mhall119, It's not the app id, it's the package name. [14:23] oh, package name? [14:24] Looks like line 420-ish is what you want in that file. [14:25] ok, I think that might be my problem [14:27] mhall119: the "name" field in the click manifest. this intentionally corresponds to APP_ID.split('_')[0] [14:27] How should I design an Ubuntu app icon for Unity 8? | http://askubuntu.com/q/444997 [14:27] https://wiki.ubuntu.com/AppStore/Interfaces/ApplicationId [14:28] jdstrand: ok, that's my problem then [14:28] mhall119: aa-exec-click sets all these things up too. you can examine /usr/bin/aa-exec-click to see what gets set and how [14:29] mhall119: also see 'Runtime environment' in http://developer.ubuntu.com/publish/apps/security-policy-for-click-packages/ [14:29] mhall119: that gives a lot of good information for a developer [14:31] dpm-afk, mhr3_: can you re-review https://code.launchpad.net/~dobey/unity-scope-click/translated/+merge/214182 please? [14:31] how can i set a new homepage on webbrowser-app in ubuntu touch? [14:32] dobey, did you talk to seb about it? [14:34] mhr3_: i looked at what system-settings is doing, and added usage of dh-translate. the translations will happen in "ubuntu" on launchpad, and not against lp:unity-scope-click. we'll have to pull the translations from ubuntu into the tree when the strings get translated. but i didn't talk to seb about it no [14:35] seb128, is that the blessed way? ^ [14:35] mhr3_: that's how we were doing it for ubuntuone packages as well [14:35] cjwatson, can you list some gsettings values for me? [14:35] gsettings list-recursively com.ubuntu.content.hub.default [14:36] gsettings list-recursively com.ubuntu.content.hub.destination [14:36] cjwatson, ^^ [14:36] dobey, mhr3_: it's one way which works fine, for most of the new touch apps and for system-settings we enabled translations with autocommit to trunk though, which means the project just needs to keep the pot updated [14:37] cjwatson, and what's in your ~/.local/share/content-hub/ dir? [14:37] seb128: i think that creates a weird situation with CI though [14:37] dobey, weird like? [14:37] seb128: weird like trunk != package [14:38] kenvandine: http://paste.ubuntu.com/7221853/ [14:38] seb128: since the CI train process is that trunk is what's in the archive [14:38] dobey, the real weirdness is that trunk == package :P [14:38] mhr3_: yes i agree, but double weirdness doesn't make it right :) [14:39] dobey, I don' think so, see e.g http://launchpadlibrarian.net/168483955/ubuntu-system-settings_0.1%2B14.04.20140228-0ubuntu1_0.1%2B14.04.20140303-0ubuntu1.diff.gz [14:39] dobey, mhr3_: the CI system takes the trunk commits and flag them as "Resync trunk" in the changelog [14:39] that works fine in practice [14:39] com.ubuntu.content.hub.default pictures ['com.ubuntu.gallery_gallery_2.9.1.922'] [14:40] cjwatson, grr... i wish i knew how that was getting set for some people... did you ever test the gallery-app switch to click? [14:40] but then you have the pot in trunk, which can be annoying for developers [14:40] bfiller, so i'm fiddling with the bootspeed on the phone, http://people.canonical.com/~ogra/ubuntu-touch-modified-280.png has a bootchart with all optimizations i can do ... if you look at the CPU graph at the top you notice some extra activity 3sec before the dashed red line (at 28sec) [14:40] dobey, right, no perfect system... [14:40] kenvandine: I don't believe so [14:40]