[00:56] is ubuntu touch ready yet for everyday users on nexus 7 2013 [01:10] good nigths. i have a question, Ubuntu phone will be have a initial phone setup with some email account and register the phone in the line? [01:16] bitgandtter: No, the phone isn't shipping with an email client and phone provisioning will presumably be done by your carrier in the normal manner. [01:16] (For example, for T-Mobile USA you just put your SIM card in and it autoprovisions.) [01:19] i imagine that because i wont read about this theme, but i want to be sure. I ask in the first place because i have an idea in the anti-theif scope [01:20] i leave in venezuela a country with a very higth index of thief [01:21] so if the phone could be register with the imei in some canonical cloud service under only one email owner, that will be provisioned in the first phone setup [01:21] that will prevent the activation of the phone by other user, unles the ifrst give him access or pass the ownership over to him [01:22] what do you think of this? [01:23] maybe this service will be something like android device manager and have some options like gps finder etc [01:39] by the way if the idea look nice to the developers i will fell enthusiastic to collaborate or develop this solution === jhodapp is now known as jhodapp|afk === vying is now known as Guest24224 [03:55] does ubuntu dekstop run smoother then ubuntu touch on a nexus7? [03:56] lotuspsychje: Well, since there are are 8 different DEs Ubuntu desktop can run... [03:57] SonikkuAmerica: well ive recently tested touch on my nexus7, but still not running smooth yet [03:57] SonikkuAmerica: so would be nice to know if ubuntu desktop works better on it? [03:58] Oh, like a chroot-and-VNC type of deal? I'd say if you want that use LXDE, but I'd say the responsiveness is roughly the same with LXDE-enabled chroot-and-VNC Ubuntu as Ubuntu Touch [04:00] SonikkuAmerica: i just want to know if everything works, touch support, automatic turn etc [04:00] lotuspsychje: In Touch? [04:00] SonikkuAmerica: no on desktop [04:00] SonikkuAmerica: i tested touch, and didnt work very well for me [04:01] So let me see if I got this right: You want me to tell you about chroot-and-VNC Ubuntu desktop running on a Nexus 7 (2013)? [04:02] SonikkuAmerica: not sure what you mean by the chroot and vnc, but its possible to run ubuntu-desktop on a nexus7 right (!nexus7 in #ubuntu) [04:02] SonikkuAmerica: and want to know if all apps work as it should [04:02] lotuspsychje: Oh! You must be talking about the now-deprecated ubuntu-nexus7-installer [04:03] it's dead [04:03] Jim [04:03] ah okay [04:03] the trigger still talks about 13.04 desktop for nexus7 [04:03] Well 13.04 is dead, I'd better go to -ops and have it checked out [04:03] so i guess ill have to wait the stable ubuntu-touch in future === zbenjamin_ is now known as zbenjamin [07:15] Anyone here running Ubuntu touch on a galaxy s3? === chriadam is now known as chriadam|away [08:19] Hi guys, I want to make a bootable usb or iso from the Ubuntu touch i386 image : trusty-preinstalled-touch-i386.zip, is there a way ? [09:09] Good morning all; happy Friday, and happy Barbershop Quartet Day! :-D === om26er is now known as om26er|lunch === vrruiz_ is now known as rvr === om26er|lunch is now known as om26er === MacSlow is now known as MacSlow|lunch === jhodapp|afk is now known as jhodapp [12:22] Anyone around? [12:22] I've flashed Ubuntu to an S3, and when I boot, I get the SIII logo, then it all goes black. Is this a good or bad sign? [12:22] It's been all black for a minute or two [12:44] how about user experience with latest ubuntu touch on Nexus 7 2012? === _salem is now known as salem_ [12:49] ramahoel: no longer supported I'm afraid. [12:52] I believe it is still supported [12:53] but I was wondering about functionality. What works and what doesn't [12:53] and how about stability [12:53] we dropped support for it, but there are still images being produced [12:54] we do not test on it anymore etc [12:54] Nexus 7 is still in active development [12:55] only for the 2013 model (flo) [12:55] there is not so much difference I guess [12:55] we dropped the 2012 one (grouper) [12:55] there's an enormous difference [12:55] it is completely different hardware [12:56] OK tks for the info [12:57] the images are funcional however? [12:57] on the 2013 version, yes. not sure on the old version [13:01] jodh, I'm told to ask you about using cgroups in my upstart jobs [13:02] tedg: here's something I prepared earlier :) ... https://wiki.ubuntu.com/TrustyTahr/ReleaseNotes/TechnicalOverviewUpstart#cgroups [13:02] tedg: alas, the cgroups upstart branch didn't make it into trusty. [13:03] what is the last functional image for Nexus 7 (2012)? [13:04] and where can I get it [13:04] jodh, So who do I get the PID of the exec'd process in post-start ? [13:05] tedg, did you read the bug btw, chrisccoulson added some summary [13:05] tedg: if you just run 'status' (no arg), you'll get the pid of the post-start and the main process [13:05] ogra_, No, just getting up :-) [13:05] morning :) [13:07] ogra_, perhaps jodh would be interested too ;) [13:07] oh right :) [13:07] bug 1303676 [13:07] bug 1303676 in webbrowser-app (Ubuntu) "[webapp-container] when using multiple webapps they crash randomly, if there is only one app remaining, this one starts being replaced when new ones start" [Critical,Confirmed] https://launchpad.net/bugs/1303676 [13:07] jodh, ^^ [13:10] jodh, So from that, it seems that we could just clean up the session in post-stop? [13:10] Or are folks thinking this is something Upstart should do itself? [13:11] tedg: still reading the bug. seems the apps are buggy though? [13:12] jodh, Just skip to the end :-) [13:12] We won't tell your high school English teacher ;-) [13:13] At some level, perhaps they're buggy, but we still need to protect against it. [13:16] there's nothing buggy about the app in this case - it already has the mechanisms to ensure it's children are all cleaned up. The issue here is that they're all stopped, so they either need waking up by the kernel (which happens when i group becomes orphaned), or someone else needs to clean up the mess ;) [13:16] *a group [13:18] I guess I'm more thinking about this as teaching us how to deal with apps that don't have those mechanisms in place as well. [13:26] hey all. probably a question a bot can answer.... I want to run Awesome-WM on my nexus7, but read that X11 is not supported. is this still the case? [13:28] yes, we don't support X on ubuntu touch [13:32] hi everybody [13:32] "How do we build popey-bot?" "We take beer and we make it watch Dr. Who, eventually it becomes sentient and we put it on IRC." ;-) [13:32] i have a problem [13:32] yamUbuntu: hello. [13:33] * popey rolls a newspaper up and raps tedg on the nose. Bad dog! [13:33] when i want to install ubuntu dual boot [13:34] an error message appears [13:34] like this [13:36] download falied need more storage:/cache need 5020331220 bytes for download and /data need 2.5g for system [13:36] popey: Can you elaborate? Is there a limited application store to provide alternatives to X11 applications? is there intention to support X in the future? [13:37] tel0s: yes and no. [13:37] Yes, we have a store which has apps which install and work. No, we don't plan to support X on the phone. [13:37] well, at some point in the far future we will support X apps in desktop mode [13:37] popey: thanks. can I see a list of supported applications somewhere without having to install the OS? [13:38] (some time on the way to 16.04) [13:38] tel0s: we dont have a web frontend to the store yet [13:39] I see. that's unfortunate. surely there is a repository of packages I can look through? It doesn't have to be a pretty web interface. [13:39] i would like to help me anybody please sorry my english is not good [13:39] tel0s: run this http://paste.ubuntu.com/7234943/ [13:41] popey: perfect! thanks! [13:41] np [13:42] dobey: ping === dpm is now known as dpm-afk [13:44] oSoMoN, ping [13:44] tsdgeos: hi [13:44] dobey: i'm trying to test the fix for https://bugs.launchpad.net/unity8/+bug/1301309 [13:44] Ubuntu bug 1301309 in Unity 8 ""reviews" widget not displaying properly" [High,In progress] [13:44] is that already live so i can test? [13:44] or? [13:45] tsdgeos: yes and no. the scope has support to show the reviews, but there are none on the server because we don't have submission fo reviews yet :) [13:45] i see [13:45] i'll see if i can just fake one and that's it then [13:46] tsdgeos: if you pull lp:~dobey/unity-scope-click/fake-reviews and build it locally, and run the scope from that branch, it should show a couple fake reviews [13:47] cwayne, plouf [13:47] let's try that, tx dobey [13:52] oSoMoN, is there a new settings api coming for webbrowser-app? specifically to set homepage, bookmarks, etc [13:53] Can i run standard console apps like vim? [13:53] and i assume the browser is firefox based/ [13:53] tel0s: yes, we have a terminal [13:53] tel0s: and you can ssh into it if you like... [13:53] cwayne, no, there’s still no settings API in the SDK afaik, so nothing in the browser [13:53] tel0s: no, not firefox [13:53] chrome? :D [13:53] oxide :) [13:53] (chromium under the hood) [13:54] well, parts from chromium :) [13:54] oSoMoN, ok, so for the time being, dropping files in ~/.config/webbrowser-app is still the right thing to do? [13:54] cwayne, yes, it hasn’t changed [13:54] oSoMoN, ack, thanks, just wanted to make sure the switch to oxide didn't change anything :) [13:55] hmm. It still seems that I'll be quite restricted on the platform. I'll have to give it a try and see I suppose. [13:55] cwayne, it changed many things, but not that one :) [13:55] :) [13:55] cwayne, dude ! [13:55] did you use the new browser yet ? [13:56] ogra_, i did [13:56] tsdgeos, just pushed a fix for the AP-failure with the modal-snap-decision AP-tests... currently running it through jenkins to see if everything's fine there too [13:56] it is snappier than androids [13:56] it really is :D [13:56] im just asking in terms of customization and whatnot [13:56] yeah ... [13:56] MacSlow: goodie [13:56] :) [13:56] * MacSlow keeps fingers crossed [13:57] Last question and I'll stop pestering you. Aside from ubuntu touch, I assume my only alternative is to run a chroot build of the desktop OS and access through VNC? or is there a beta/alpha port of the desktop for ARM? [13:57] nope [13:57] thats stuff we will start developing during the next two years [13:58] target for the desktop/convergence mode is 16.04 [13:58] ogra_: beautiful. :) [13:58] rsalveti, is adb no longer available on emulator? [13:58] first we needed a rock solid phone OS before starting to add shiny features ;) [13:59] that bit is close to be done ... [14:03] dobey: in which apps should i see the fake reviews? [14:04] tsdgeos: any click app should show them [14:04] ok [14:21] jodh, So, thoughts on how to clean up the processes? [14:25] dobey: i was missing a restart of the smart-scopes-proxy ^_^ [14:26] tedg: does the appmgr get notification that the process group leader has died? [14:26] tedg: well, been killed. [14:27] tsdgeos: ah, so the fix works then? :) [14:28] dobey: yes :) [14:28] jodh, Yeah, basically we pass along the stopped signal from Upstart. [14:28] jodh, it needs to know about that since it maintains the oom score for the apps [14:28] so yes [14:28] tsdgeos: great. thanks! [14:31] tedg: well couldn't you just react to that and kill (-main_pid, SIGHUP/SIGCONT)? [14:32] jodh, the appmanager doesnt kill anything [14:32] it only sets oom scores, the kernel does the killing then [14:33] jodh, Yeah, so it'd be nice that when we realize we're stopping the app, we could ensure that it is really stopped. Whether by crash or something else. [14:33] (but it knows the parent PID of the group) [14:33] ogra_: I gathered that, but that might be a solution to the problem you are facing :) [14:33] right [14:34] but i think there was a reason to leave the killing to the kernel [14:35] ogra_: I think I'm missing something. If these apps are not buggy, why are they getting killed by the OOM? [14:35] jodh, Because we're on a restricted memory device, the only way we know when to shutdown a background app is when we're out of memory. [14:36] jodh, Basically apps only close by being killed by the OOM. [14:36] jodh, apps in background get sigstopped and a low oom value ... the more apps you start the less ram you have ... at some point the bg app is killed when you have enough apps [14:37] the SIGSTOP/CONT bit is done by the appmanager ... as well as the rescoring [14:37] but the killing is left to the kernel ... [14:37] note that even if the app was killed by the kernel it will still be in your tasklist as backgrounded app [14:38] if you tap on it the app either gets SIGCONT or it gets newly started because the kernel had killed it [14:39] does the oom rescoring only affect the process group leader then? [14:39] so usually the longest backgrounded app will die if you start new ones [14:39] i think its only done for the leading pid, not sure though ... (i didnt work on the code, but know roughly how it works) [14:40] if all a jobs pids had the same oom score, wouldn't that solve the problem? [14:40] iirc the parent PID is the only thing that upstart-app-launch hands over ... [14:40] cwayne: it should be [14:41] Yeah, I think it only knows about the parent today. We do need to fix that. [14:41] But I don't think that the OOM would kill all of the things at the same score though. [14:41] I think it only kills what it needs to, which may only be the parent. [14:41] jodh, only if the kernel knows about that (since it is the killer) [14:42] Would be interesting to change the score so that the parent is always one less than the children. But that's probably getting a bit fancy :-) [14:42] jodh, but i dont know if the kernel would SIGCONT them before killing then [14:42] popey, https://bugs.launchpad.net/ubuntu-ui-toolkit/+bug/1306656 [14:42] Ubuntu bug 1306656 in ubuntu-keyboard (Ubuntu) "OSK does not get displayed after a suspend/resume" [Critical,New] [14:43] the prob is that they are stopped ... even if the kernel sends its KILL ... they wont recieve it ... [14:43] so when the appmgr sends SIGSTOP, why can't it first set the oom score for all pids in that process group to 1000 so that the kernel would consider the entire group for killing? I prolly need to read the oom code to see if that would work though (haven't looked at that for years... :-) [14:43] oh, and the special thing about the new webapps is that the forked renderer processes all run as root in a sandbox [14:44] * jodh curses self for starting a response with "so" [14:44] jodh, because it only gets told about the PPID [14:44] I think that, no matter, we need to handle the case of the parent magically disappearing (by whatever cause). [14:45] pmcgowan: confirmed [14:45] vg [14:45] well, for getting the release blocker out of our way even a hack would do [14:46] i agree this should be solved properly, but i'm wondering if the time isnt to short for this [14:47] jodh, can I get the list of processes in the session in post-stop ? [14:47] jodh, did you see my comment on that bug? any thoughts? [14:47] hi tedg :) [14:48] * ricmm eavesdrops on this convo [14:48] oh [14:48] i see everyone is talking about that already [14:48] sorry ;) [14:48] tedg, i'm a bit scared that any hack you do to u-a-l might impact the app startup time again ... we're still not shiny there [14:49] ogra_, Sure, that's why I'm trying to avoid building the cgroup, but if upstart is already building the session, and I can just use that in post-stop, it'd only delay when stopping apps. [14:49] Slower shutdown of apps should be fine :-) [14:50] tedg: not without groping around in /proc. using cgm would make life easier though since you can just cat the cgroup task file. [14:50] well, that means slower restart perhaps :) depends when the shutdown occurs :) [14:50] but yeah, less impact for sure [14:50] yeah, i'd like to see you say "slower shutdown is fine" to all the people who complain about firefox's "Firefox is already running" dialog ;) [14:50] heh [14:51] jodh, I'm not above groping /proc [14:51] tedg, jodh: not sure if you guys saw this: https://bugs.launchpad.net/ubuntu/+source/webbrowser-app/+bug/1303676/comments/31 [14:51] Ubuntu bug 1303676 in webbrowser-app (Ubuntu) "[webapp-container] when using multiple webapps they crash randomly, if there is only one app remaining, this one starts being replaced when new ones start" [Critical,Confirmed] [14:51] ogra_: back to this real quick: "the prob is that they are stopped ... even if the kernel sends its KILL ... they wont recieve it ..." [14:51] ogra_: the SIGKILL would get through [14:51] tyhicks, oh,, thanks ! [14:51] np [14:52] jodh, How does that work? Where are the sessions? [14:52] ogra_: SIGCONT and SIGKILL are the only signals that I'm aware of that get through to a SIGSTOP'ed process [14:53] tyhicks, right, i didnt know about KILL actually [14:54] * jdstrand is slightly surprised TERM doesn't go through [14:55] it might but I didn't try that yesterday bc KILL was sufficient [14:55] tedg: define session? [14:55] but I guess it makes sense. the process can catch TERM, but if it is STOPped it obviously can't do that [14:56] jdstrand: just tested it - TERM doesn't unSTOP a process [14:56] jodh, Honestly I'm a bit unclear on that, seems like a process group to me. But it seems from chrisccoulson's comment that setsid is the key here. [14:56] * jdstrand nods [14:57] tedg: ? I'm saying, why not consider using cgm(1) to cgroup-contain the apps, then you can query all the pids in that "app" by looking at the app-instance-unique cgroup task file. [14:58] jodh, I do like that, but we're worried that creating the cgroup on application start would delay it. [14:58] jodh, I think that makes sense longer term, to figure that out. But I was hoping for something quicker we could do in post-stop today, with what we have. [14:58] tedg: I'd suggest trying. [15:03] dpm-afk, ping when you're back [15:05] tedg: if not, you can get the session id from 'awk '{print $6}' /proc/[0-9][0-9]*/stat' [15:07] jodh, Does that stay around even after the parent PID is gone? i.e. for post-stop ? [15:17] mdeslaur, With the apparmor stanza in the upstart job, is the post-start job confined? [15:40] So, let me talk this through :-) [15:40] * ogra_ cleans his ears [15:40] If, in post-start, I got the process group of the parent pid [15:40] Then I stored that in the job's environment [15:40] Then in post-stop I sent KILL to that process group [15:40] We should be good, no? [15:41] try it ? [15:42] Heh, there is that :-) [15:42] the issue is easy enough to reproduce with a few new-API webapps [15:57] * tedg seems to see the pattern in this bug, clearly the problem is the BBC! [16:00] tedg, i rather think its the NSA intercepting the connection to the guardian [16:00] damn liberals [16:01] didrocks: I figured out what you're up to [16:01] giving good news? sure! :) [16:01] promotable image? no :p [16:01] waiting for #300? no! [16:01] didrocks: you're holding out [16:01] or maybe ;) [16:01] damn, that was my theory [16:01] holding out for #300 on release day [16:02] ha... #300 is going to be awesome :) === gatox is now known as gatox_lunch [17:11] robru, hey, are you around? [17:14] oSoMoN, hey [17:16] robru, hey, I have a landing request (silo 8) that was blocked by an FFe, but the FFe has just been acked, so I guess it’s good to publish now, can you confirm? [17:16] oSoMoN, confirmed [17:16] robru, thanks [17:17] mhall119, oooh, is there a new click for trojita ? [17:17] ogra_: there is [17:17] ~git140 [17:17] robru, would you mind publishing for me? [17:17] oSoMoN, published ;-) [17:17] that was fast, thanks :) [17:18] oSoMoN, well it had to be done right now ;-) [17:18] mhall119, got the full url again ? (i'll bookmark it now, promised) [17:20] ogra_: http://people.ubuntu.com/~mhall119/trojita/ [17:21] thx [17:21] I could use some help automating the click package creation [17:22] hint, hint [17:22] mzanetti: ^^ [17:25] OMG ! [17:25] i see mail ! === bfiller is now known as bfiller_afk [17:26] \o/ [17:26] ogra_: spoiler alert: Bruce Willis was mail all along === dpm-afk is now known as dpm [17:39] hey cwayne, I'm back [17:40] dpm, hey, where are we re: scopes translations? [17:42] cwayne, essentially, the click scope is now translatable, and that's pretty much it. So what is left is still to make remote scopes translatable and the phone to send the locale to the server, so that our default remote scopes are translatable. Any other local scopes still need to be made translatable in the same way as the click scope [17:42] Additionally, there needs to be an API accessible to read the translations of the app's metadata [17:43] essentially name and description [17:43] I've got a couple of bugs to track all this, let me dig them out [17:44] bug 1297889 [17:44] bug 1297889 in unity-scope-scopes (Ubuntu) "Scopes are not translatable" [Undecided,New] https://launchpad.net/bugs/1297889 [17:44] and bug 1302045 [17:44] bug 1302045 in Software Center Agent "Make click translations metadata available in the API " [Undecided,In progress] https://launchpad.net/bugs/1302045 [17:44] so even the scope titles themselves are still not translatable? [17:46] now, do I wait until Monday in the hopes of an image promotion, or just onto -proposed knowing that there are only 2 blockers keeping it from promotion [17:47] popey: davmor2: ^^ how safe is -proposed right now? [17:47] better than ever [17:47] davmor2 is on vacation [17:48] well [17:48] we just discovered a solid blocker [17:49] how solid? [17:49] mhall119, try doing two calls in a row [17:50] is that common? [17:50] :) [17:50] (pulse will crash) === gatox_lunch is now known as gatox [17:50] does it always crash on the second call, or only if it's made soon after the firsT? [17:50] * mhall119 doesn't make many calls, but also doesn't want to reboot after each one [17:51] right, especially not with the release image :P [17:58] cwayne, sorry for the delay. Exactly, scope titles are not translatable :( [18:04] If anyone is bored, this works for me, but I'd love it if other people could test it as well: https://code.launchpad.net/~ted/upstart-app-launch/process-group-kill/+merge/215475 [18:04] I'm going to test some more before calling it done. === bfiller_afk is now known as bfiller === LarrySteeze|Away is now known as LarrySteeze [18:30] alex-abreu, looks like tedg has a fix, care to test? [18:34] If anyone grabbed that branch already, refresh, forgot to move a -KILL over into the branch. [18:34] r148 [18:37] Copying is hard, this is why I could never cheat at school. [18:43] http://www.youtube.com/watch?v=7soFYjn_pQc [18:45] HOW DO I INSTALL UBUNTO TO MY ANDROID? [18:45] Cool, now tested with packages on a rebooted device. [18:49] how do i install ubuntu to my tablet? [18:49] android on that is older version [18:49] OTA never come [18:50] will it work on my device? [18:51] can i double boot with android? [18:56] tedg, well, it does kill the oxide renderers ... but it gets pretty chaotic if you have enough apps open [18:56] if they get killed they take way to long to respawn [18:56] and the app switcher starts re-ordering them ... [18:57] tedg, but for a hack that looks good :) [18:57] i have so many apps a i need a search funtion [18:58] i taped 7 times but nothing comes up [18:58] Step 1.5 - Optional Android Backup [18:58] If not enabled, enable developer mode, by tapping Settings -> About phone -> Build number (x 7 times) [18:59] do you have a supported device ? [18:59] what do you mean? [18:59] i come to irc for support [19:00] can i flash it using windows8? [19:00] do you have a device upbuntu touch can run on ... [19:00] bc the thing isnt booting anymore [19:00] we only support three devices by default ... and there are about 50 other community ports [19:00] i dont know im going to try it [19:01] if yours is not one of these models you wont be able to run ubuntu on it without porting it yourself [19:02] its got mali400 grafics [19:02] what device is it ? [19:02] and its android [19:02] that doesnt help [19:03] its got 2 cams [19:03] its 7" [19:03] what manufacturer or model [19:03] mdeslaur, So if an app is confined, can it get out of the process group that upstart starts it in? [19:04] If using process groups would solve that problem too, I'd be happy. [19:04] :-) [19:04] model? [19:05] you dont need modle to instsll ubuntu on computers [19:05] ubuntu for phones is different [19:06] isnt android open source? [19:06] so its even easyer [19:06] parts of it are [19:06] so you mean its not open? [19:06] ive been had [19:06] oarts of it are not [19:06] *parts [19:07] i.e drivers [19:07] if parts arent that means its not open [19:07] the google apps or the playstore ... [19:07] is it open YES or NO? [19:07] if its not completelly open the naser is a resounding NO [19:08] tedg, so i can confirm your hack "fixes" the issue but it really points out how slow the webapp container restart is ... gives a slightly weird experiance [19:09] ogra_, hmm, haven't noticed that, I was starting other apps. What are you doing for that? [19:10] ogra_, Oh, you mean when you go back to the app after it's been killed? Yeah, that kinda sucks. [19:10] i have 9 apps running ... and i randomly pick one from the app switcher in a rotating manner [19:10] at some point one app that was longer backgrounded than others dies [19:11] OOM Russian Roulette [19:11] if i now select that app i see the surface of the open app *underneath* that app ... until the webapp-container restarted and loads the app that i actually tapped [19:11] there is quite some deelay until that has settled .. and like 5 sec where you only see the app underneath [19:11] Okay, yeah. I see that. I can't fix that. :-) [19:12] right [19:12] its an app startup speed thing [19:12] i wonder if mir couldnt show the surface it uses in the app switcher until it has loaded properly again [19:12] ricmm, ^^^ [19:13] we have talked about that for a while, showing a snapshot [19:13] right [19:13] * tedg assigns bug to pmcgowan, he spoke first [19:13] :-) [19:13] heh there is one open somewhere [19:13] even a non responsive pic of the right app is better than seeing the wrong app in front of you [19:13] Seems like worst case at least dimming the app behind would be better. [19:13] and we already have the surface stored somewhere [19:14] cant be so hard to shuffle that on screen [19:14] (though effectively we should simply start faster) [19:14] 5 sec is way to long [19:15] yeah why it so long [19:15] oxide starts a tad slower (it is bigger) [19:15] currently the webapp container still loads webkit alongside, the fix is stuck in proposed (final freeze) === frecel_ is now known as frecel [19:16] dropping that should speed us up a little [19:16] well, and teds hack to fix the bug is kind of contributing to the restart time again [19:17] pmcgowan, I think that ricmm was working on getting some numbers on that. Not sure how far he got. More looking at the qtubuntu side of things. [19:18] Seems we have more control to fix the QML apps than we probably do the webapps. [19:20] I find the weather app funny, it goes red at 80°F. That's really only hot for someone that lives in London :-) [19:20] Should take into account averages for the area. [19:25] * ogra_ tries with the webkit dropping patch [19:33] nope ... doesnt improve [19:36] hmm, ok [19:36] if i restrict myself to 5 webapps nothing crashes [19:36] s/crashes/gets killed/ [19:39] yup, even starting a few native apps doesnt get in the way [19:40] can i install firefox OS on my android? [19:40] ask in a firefox OS channel perhaps [19:41] firefox OS is going to rule [19:41] tedg, ugh, why is u-a-l seeded in the desktop ? for what do we use it there ? [19:41] its coded in html5 [19:41] its lightning fast [19:41]