[00:17] can I tempt anyone to install the grooveshark webapp (the one by grooveshark!) from the clickstore? [00:17] as per https://bugs.launchpad.net/bugs/1389028 [00:17] Launchpad bug 1389028 in ubuntu-system-settings-online-accounts (Ubuntu) "u1 account crashes and gets removed when installing app" [Undecided,Incomplete] [00:17] I cannot believe I am the only one [00:18] sure [00:19] nah popey you tried already. you are of no use :) [00:19] lulz [00:19] in this particular case ;) [00:20] but how on earth does this make sense though [00:22] its consistent across different images and different channels even. [00:23] Starting to feel like my U1 account itself must be corrupt. But then again why can I install and remove any other app no problem at all ? [00:23] me brain explodes [00:45] Seems to work for me. [00:48] RAOF did you mean the Grooveshark thing? [00:48] matv1: Indeed. [00:48] ROAF wow [00:49] ROAF thanks very much for trying [01:07] Hello, Someone faced this problem before? [01:07] IOError: [Errno 2] No such file or directory: '/tmp/targetfiles-xxxxxx/BOOT/RAMDISK/file_contexts' [01:09] it seems it looks into the ubuntu ramdisk (ubuntu-root) not the kernel ramdisk (root folder) [01:10] so there is no file_contexts there and cause this error (make otapackage that use ota_from_target_files) [01:14] * matv1 afk === LarrySteeze|Away is now known as LarrySteeze [01:53] anyone? [01:59] they seems to all R.I.P here :) btw for the bug i wrote about someone report it when someone come's back to life === chihchun_afk is now known as chihchun === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === LarrySteeze is now known as LarrySteeze|Away === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [06:47] Hi, is there any documentation for audio system in ubuntu-touch ? [06:48] kiraank, nothing up-to-date AFAIK, is there anything specific you'd like to know? [06:49] I know that Pulseaudio is being used. I am interested to know, mixer control setting part. [06:49] kiraank, we were using UCM for a while, but these days the mixer control settings goes through the audio HAL [06:53] diwic: "audio HAL", do mean android audio HAL ? [06:54] kiraank, yes [06:58] diwic: How is policy and routing stuff is handled? [07:01] kiraank, the audio HAL's concepts map to pulseaudio's concepts [07:03] kiraank, you can read the code here: http://bazaar.launchpad.net/~ubuntu-branches/ubuntu/vivid/pulseaudio/vivid/view/head:/debian/patches/0207-Enable-pulseaudio-droid.patch [07:04] hi, is there a way to install not signed click package [07:07] diwic: thanks for patch. will go through to get better understanding. So can I assume that PA handles routing, rendering without help of AudioPolicyService of android? [07:16] kiraank, we talk directly to the Android HAL. I'm not familiar with AudioPolicyService so I think that it is not there on ubuntu touch [07:19] kiraank, out of curiousity, are you asking because you want to make some app, or...? [07:26] diwic: I am not making any app. I know that PA supports ucm. To use features like changing the profile/port, there is a need for any external client. Was curious to know how PA is doing same things in ubuntu-touch === mirco is now known as MacSlow [08:19] sil2100: morning. note that there is a new appmenu-qt5 bug #1394887 - should I assign it to you? [08:19] bug 1394887 in appmenu-qt5 (Ubuntu) "appmenu-qt5 fails to build against Qt 5.4.0 beta" [Undecided,New] https://launchpad.net/bugs/1394887 [08:22] good morning [08:44] bzoltan: good news, once lp:~mvo/oxide/multi-arch-fixes lands the 15.04 chroot should be good again [08:46] mvo: \o/ [08:46] bzoltan: there was more needed, but I uploaded the other stuff myself, but 320mb (or however big it is) needs a review first :P [08:47] mvo: if you need any help I am here [08:56] Mirv: o/ Let me take a lookie at that :) [08:59] sil2100: ok! [09:02] tsdgeos: morning to you too! um. about that dbus fix Qt build... [09:02] tsdgeos: ...so the patches were there, yeah. they just happened to be missing from debian/patches/series... [09:02] tsdgeos: so thanks for testing the network-manager fixes :( ubuntu5 in the same PPA has now built for all other archs than armhf. [09:03] armhf should be ready in maybe half an hour. [09:03] Mirv: so you mean i wasn't really testing the dbus changes? [09:03] tsdgeos: yes, unfortunately [09:03] Mirv: oh :D [09:04] tsdgeos: so, if you could look at the unity8 again after 1h and maybe plasma5 in the evening, and we'd relook at the situation on Monday? :) I'll rerun AP:s on the ubuntu5 version. [09:04] Mirv: sure, which silo? same? [09:04] tsdgeos: same, 027, https://launchpad.net/~ci-train-ppa-service/+archive/ubuntu/landing-027/+packages [09:05] I also combined Zoltan's changes in there now [09:05] ok [09:05] thanks, and sorry.. === duflu_ is now known as duflu [09:48] marcustomlinson: ping [09:49] mardy: pong [09:49] marcustomlinson: could it be that the youtube scope process has an opened connection to Mir? [09:50] pete-woods: ^ any idea? [09:51] marcustomlinson: so, the reason why the notification is not shown in vivid, is that in vivid we try to create a trusted session with the client, and if that works, we don't show the notification [09:52] mardy: no, splash screen is not even shown [09:52] marcustomlinson: but then it turns out that the trusted session does not work, so nothing is shown [09:52] marcustomlinson: I suspect it's a bug in Mir [09:52] mardy: yes I think it is. They are looking into it [09:54] mardy: greyback is aware of the issue [09:55] marcustomlinson: what issue, exactly? I still haven't a clear understanding of what's happening; my feeling is that the trusted session is created even when it should fail (because the initiator PID is not connected to Mir) [09:56] mardy: basically I just told him the splash screen doesn't show on vivid. He said they haven't made any fixes for that yet. Then he said he'd look into why its not showing up anymore [09:56] mardy: if you want more info you should probably ask greyback [09:56] marcustomlinson: but that's correct, the splash screen should not be shown [09:57] mardy: right, but only if they intended that right [09:57] mardy: nobody seems to have intentionally hidden it [09:57] dednick: hi! Could it be that now trusted prompts are creted even if the initiator pid is not connected to Mir? [09:57] marcustomlinson: ah :-) [09:58] mardy: so somebody needs to bring it back so they can hide it again (properly) ;) [09:59] marcustomlinson: so, the logic we have in OA is: we always try to create a prompt session on top of the client (and in this case there is never the splash screen); if that fails, than we fall back to showing a snap decision (and here there is the spash screen that we'd like to get removed) [10:00] marcustomlinson: so I wonder, if the reason why the splash screen no longer appears is that we never fall back into the second case [10:00] marcustomlinson: and that would happen if Mir agrees to create a prompt session even though the initiator PID is not connected to Mir [10:00] dednick: ^ [10:01] Good morning all; happy Friday, and happy World Hello Day! :-D [10:01] mardy: sure. So I have a set of steps to reproduce the splash screen every time on RTM. And it doesn't show the splash screen on vivid. But the process (that usually bring up the splash) is definitely running [10:01] mardy: so it works [10:02] mardy: its just strange, because focus is returned to the scope, then you wait staring at the scope for about 10s until it refreshes (whereas usually the splash screen is visible) [10:05] marcustomlinson: sorry for switching topic, can I ask you about the desktop files? [10:05] marcustomlinson: just saw your mail [10:06] where does one report bugs for the phone-app? here? https://launchpad.net/ubuntu/+source/phone-app [10:07] what are the usual reasons if the emulator does not work? [10:09] mardy: actually, just thought about that bug a bit. The only thing we put into the desktop file is the scope name [10:09] mardy: so that shouldn't need localisation. [10:10] mardy: the bug is talking about the description from the .application file [10:10] mardy: I how do we localise that? [10:11] marcustomlinson: wait a second, I was wondering how can live without a .desktop file :-) [10:11] marcustomlinson: don't you need that to provide an icon so that your scope can be displayed in the system settings, in the "memory usage" page? [10:12] mardy: then we need an icon entry in the .application file I guess [10:12] mardy: but lets not argue about the .desktop file then, we can buy more time on fixing that [10:13] mardy: the immediate bug is the translation of the service description in the .application file [10:13] marcustomlinson: the problem is that we don't read that field, if we have a valid desktop file [10:14] mardy: But we don't put anything other than icon and name into the desktop file at the moment [10:14] mardy: and NoDisplay=true [10:15] marcustomlinson: yes, so we read the name from there [10:15] (and the icon) [10:15] which is fine, no localisation required [10:17] marcustomlinson: I don't follow you. The bug is about the name not being translated, so why do you say that no localisation is required? [10:18] mardy: ok perhaps. I thought he was talking more about the description of the scope under the name (i.e. "Watch your favorite YouTube videos") [10:19] marcustomlinson: ah, that's not showed in the Ubuntu Touch UI; we show it only in Unity7, on the desktop [10:23] mardy: so is there a solution where the OA backend only uses the icon from the desktop file and the name from .application? [10:23] mardy: then explain to Kyle N how to localise the strings in /application === dholbach_ is now known as dholbach [10:25] marcustomlinson: no, we always read it from the .desktop file [10:27] mardy: ok maybe we actually already handle localisation of the name. Give me a minute to check [10:29] marcustomlinson: I see that the system settings plugin has some special code to get the name of the scope from the ini file, but if the .desktop file worked correctly that could be removed and make life simpler for everybody (except you :-p) [10:30] mardy: k, ignore the name for now, I think we can localise that in the deskop file ourselves. [10:30] mardy: is there a way to localise the description in the .application file? [10:32] marcustomlinson: yes, add a element with the name of the gettext domain (like youtubescope) [10:45] is system settings hanging when you go to battery -> screen brightness a known issue? [10:46] mardy: ok ignore that bug from the mail we actually already handle this on our side [10:46] mardy: but.. Please do keep it in the backlog to investigate removing the need for the .desktop file [10:47] Chipaca, bug 1337200i think [10:47] bug 1337200 in ubuntu-system-settings (Ubuntu) "High CPU due to excessive device changed signals from upower" [High,In progress] https://launchpad.net/bugs/1337200 [10:50] ogra_: dbus daemon is at 10% cpu [10:50] ogra_: pretty much the lowest it's been since starting system settings [10:50] then it is a new bug [10:50] file it [10:50] some of our IPC patterns are textbook "how not to IPC" :-( [10:50] will do [10:55] bug #1394944 [10:55] bug 1394944 in ubuntu-system-settings (Ubuntu) "blank screen in battery → screen brightness" [Undecided,New] https://launchpad.net/bugs/1394944 [10:56] (on a bit further testing, system settings does not hang past the initial dbus daemon spike, hence "blank screen" in the bug descr) [10:56] I have the same bug [10:56] zmaj: +1 it then please? [10:56] ie "affects me too" [10:57] I did it... [10:57] thanks [10:57] Though I myself have a bug that my touch emulator does not boot... [10:59] Chipaca, can you put image number and device name in there ? [11:01] ogra_: hey, i have a question about the preinstalled webapps [11:01] ogra_: i noticed that googleplus is pre-installed, but it's url dispatcher hook is not active it seems [11:01] its [11:02] googleplus isnt preinstalled [11:02] at least that would be news to me [11:02] hmm [11:03] then there's something odd [11:03] i'm talking about the vivid image, just to be sure [11:03] image 28 from this morning [11:03] well, would br news to me that we added it [11:04] hmm, then it was installed on my phone, but something changed in the way the url dispatcher db is managed may be [11:04] http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20141121/vivid-preinstalled-touch-armhf.manifest [11:04] scroll to the bottom [11:04] it lists the preinstalled ones [11:04] i will try to reinstall the app to see if that puts the hook back into place [11:04] ogra_: ok, checking; thanks [11:04] ogra_: done; krillin rtm-proposed #166 fwiw [11:07] I heard that they have added windowing into touch...now that is nice progress [11:10] mardy: hi, i shouldn't think so. I think we check for the app pid before creating the trust prompt [11:10] mardy: at least in trunk mir we do. [11:13] zmaj, can you also add info about which image version and device you use to the bug ? [11:25] dednick: OK, I'll investigate [11:26] mardy: what's happening with it? [11:30] dednick: the trust session is created (and is valid) even if the initiator is a scope (which shouldn't be connected to mir), but then when we actually try to show a window in the session, nothing is shown (and no error is returned either) [11:33] mardy: ah. i think there's something special going on with the scopes. i think they use the dash maybe? [11:34] Saviq: ^ you mentioned this yesterday [11:37] dednick: dunno, but I just hacked our trusted helper to use 99999 as initiator PID (and that pid does not exist), and the session is created and considered valid [11:38] dednick: this is on the RTM image [11:38] mardy: ok [11:39] mardy: what's the mir version on rtm? [11:39] greyback: thanks. I've been told that you are working on some bug about the splash screen being missing; is that about Online Accounts, or a general issue? [11:40] dednick: libmirclient8 is 0.8.0+14.10.20141010-0ubuntu1 [11:41] mardy: I'm aware of it, it's vivid only. I've not investigated it yet [11:41] mardy: ya, well that's the missing patch that alan_g did for non-existing application pids. [11:41] greyback: do you have the bug #? I'm trying to understand if it's caused by the issue I'm discussing with dednick, or if it's unrelated [11:42] dednick: ops, sorry, I forgot it was alan_g; but it's great that you found it out [11:42] mardy: I don't have a bug for it [11:43] but in my quick look, the OA process was started, but for some reason shell wasn't picking up it up as trust session [11:43] dednick: BTW, on vivid I've got the same version of mir [11:44] greyback: OK, let me describe you what I'm seeing, maybe it's the reason for the failure: [11:45] dednick: mardy but 1377968? [11:45] bug 1377968 [11:45] bug 1377968 in mir (Ubuntu) "a prompt session with an invalid application pid should be an error" [Medium,Triaged] https://launchpad.net/bugs/1377968 [11:45] mardy: http://bazaar.launchpad.net/~mir-team/mir/development-branch/revision/1967 [11:45] greyback: in vivid, OA always tries to create a trust prompt before showing any UI; only if that fails, it shows a snap decision (from a process which instantiates a QGuiApplication, which causes the splash screen to be shown) [11:46] alan_g: that one, excellent! :-) [11:46] yw [11:46] greyback: now, because of the bug just linked above by alan_g, the trust session always appears to have been created successfully [11:46] is there a proper texteditor in development? [11:47] something like libreoffice? [11:47] greyback: so OA opens its UI in the trust session; but that session is in fact non-working, so nothing is shown on screen [11:48] greyback: does it sound similar to what you've been told to investigate? [11:48] mardy: it's fixed in 0.9, just wasn't back-ported to 0.8. [11:48] mardy: it sounds like the bug that was mentioned to me yes [11:48] greyback: what is worse, is that OA doesn't know that its window is not shown, so in fact it continues running, hanging all future authentication requests [11:49] greyback: cool, then I guess you can forget about it [11:50] mardy: ok, thanks :) [11:51] mardy, FWIW, I just discussed this with tvoss yesterday and the longer-term plan would be to allow associating arbitrary PIDs of headless processes (e.g. scopes) with surfaces [11:52] mardy, this way, when OA opens a session for that PID, it will come up on surface(s) that's associated with that PID (be it dash, a separate scope view or whatnot) [11:52] at least that's the current plan [11:56] Saviq: ok, that hopefully won't cause problems [11:56] mardy, I think that should actually solve problems :) [11:57] mardy, as you'll be able to display a trusted prompt for a scope on UI that's actually associated with it [12:03] Saviq: well, the dash hosts many scopes, so if you refresh the youtube scope and immediately switch to another scope, and see the trusted prompt appear on top of the wrong scope, you might be confused [12:04] mardy, that's a corner case, and how is that worse than what you're doing now with the notifications? [12:06] Saviq: well, the scopes should ask the dash to get an authentication token on its behalf, and if the user swipes to another scope, the dash could cancel the authentication request (or prevent the swiping) [12:07] mardy, yeah, problem with that is dash would have to become a proxy for every trusted helper [12:07] Saviq: if the scope does everything itself under the hood, the dash has no idea of what's happening [12:07] Saviq: yep, but TBH I don't see a way around it [12:08] mardy, I don't think it's going to be too much of a problem (what you described) [12:08] mardy, the user did just see the youtube scope [12:08] mardy, and the prompt will say what's trying to access what [12:09] mardy, and if we pop the prompt up quickly enough, the user won't have time to swipe away ;) [12:10] Saviq: true, but imagine that the phone has just booted, and the user is swiping throught the scopes to access the last one [12:11] Saviq: this is more of a UI design problem than a technical one, I guess, but I think that the scopes should not cause any UI prompt at all, unless after an explicit action by the user [12:11] Saviq: maybe I'm mixing two different issues together :-) === MacSlow is now known as MacSlow|lunch === dandrader is now known as dandrader|afk [12:51] tsdgeos: I'm not sure if it's the non-bootstrapwiped mako I have, but when I now started running UITK AP:s on it with the ubuntu5 from silo 027, unity8 does restarts [12:52] :S [12:52] so it might be the dbus patches aren't that great [12:53] tsdgeos: not sure if the unity8 log tells anything http://pastebin.ubuntu.com/9147665/ [12:54] dbus log http://pastebin.ubuntu.com/9147694/ [12:55] can't see anything obviously wrong in there === alan_g is now known as alan_g|lunch === dandrader|afk is now known as dandrader === jhodapp is now known as jhodapp|sick [13:35] I'm going to land the NM fixes + qmake-armhf fix only, which accidentally I have... [13:36] and were tested === alan_g|lunch is now known as alan_g === MacSlow|lunch is now known as MacSlow === Spookley_ is now known as Spookley [14:21] bzoltan, what do you think of this https://code.launchpad.net/~ken-vandine/+junk/qtcreator-plugin-bacon2d [14:22] kenvandine, hey, does this work? http://paste.ubuntu.com/9148051/ [14:22] jgdx, for a quilt patch? [14:23] jgdx, do you have a bzr branch already that you used to build? [14:23] jgdx, yes [14:23] yes [14:23] you can push that and i can sponsor it [14:23] I'm building, so that's not done [14:23] kenvandine, right [14:24] the patch looks fine [14:24] jgdx, when you do use this in settings, make sure you bump the depends [14:24] so we don't forget to backport this to rtm when that feature gets backported [14:25] kenvandine, yup [14:26] jgdx, so without phonesim supporting the mtk stuff, we can't really test it :/ [14:27] jgdx, should we look at adding that to phonesim? [14:27] or is this something we would really only test with mock [14:28] kenvandine, both radiosettings (the whole interface) and mtksettings are untested in libqofono [14:28] i'm talking about testing in settings [14:28] when we use it [14:28] oh that's no problem [14:28] just mock? [14:28] yeah [14:28] great [14:29] makes me happy :) [14:29] sorry, thought you were talking about libqofono [14:29] we should sort out testing for libqofono though :) [14:29] ever figure out how to run those tests? [14:29] kenvandine, yeah, like you said :) thanks [14:30] so just installed? [14:30] kenvandine, yes [14:30] ok, i'll look at fixing that :) [14:30] if you don't mind... [14:31] kenvandine, fixing what? Adding radiosettings and mtksettings to phonesim? [14:31] fixing running the tests from the build tree [14:31] it bugs me that we can't do that [14:32] we rely on libqofono for quite a bit, i want it tested during package build at least [14:34] kenvandine, aah, cool. There probably is a way, but I did not look for the compiled tests in the build tree. [14:34] i tried briefly when i first packaged it, but didn't have time to figure it out [14:34] jgdx, but it really bothered me, i'd like to finish that :) [14:35] +1! [14:37] kenvandine, that qa email thread was really good reading this morning. [14:38] thanks for that [14:38] jgdx, feel free to add to it :) [14:40] jgdx, i feel pretty stongly on that point about where integration testing should be, the tests should be in the component that provides the API/backend not the consumer [14:40] if the indicators break actions, that should get caught in the indicator tests before they land [14:41] otherwise, we'll catch it in our tests after they are broken, which is too late [14:41] so no point in us testing the actions themselves, mocking should be good [14:41] and count on the api contract [14:41] kenvandine, totally agree. [14:42] if we have faith in the indicator tests [14:43] tsdgeos, Saviq, okay I think I have a branch that fixes bug 1394622 [14:43] bug 1394622 in ubuntu-app-launch (Ubuntu) "0.4+15.04.20141118~rtm-0ubuntu1 causes flickering on spread "alt+tab" gesture" [High,In progress] https://launchpad.net/bugs/1394622 [14:43] Not sure about landing and all that BS [14:43] cool! [14:43] But we can land it in Vivid :-) [14:44] tedg, glad [14:45] kenvandine, will thou sponsor? lp:~jonas-drange/+junk/libqofono-allow-changing-technologies-fixes-1373388 [14:49] dednick: I just filed bug 1395028, can you please tell me if I filed it on the right component, and whether it's something which could get fixed in the RTM? [14:49] bug 1395028 in QtMir "Dead processes are still shown in a trust session" [Undecided,New] https://launchpad.net/bugs/1395028 [14:49] jgdx, will do! [14:59] mardy: hm. not sure if it's qtmir or unity8 that is the problem there. probably qtmir [15:01] mardy: and "fixable in rtm" is only dependent on what priority it is given by the big shots. [15:03] mardy: i'm quite supprised that doesnt work though. i would think it should transistion back to A as soon as process B quits. [15:04] dednick: but that wouldn't be correct, if C appears in the session before B quits [15:05] mardy: but you said C is hosted on B [15:07] jgdx, what's the bug # for that libqofono feature? [15:07] dednick: no, they are independent processes; they just have the same MIR_SOCKET variable when they start [15:07] mardy: unless you're saying that your initiating 2 prompts (B & C) using pid of A. [15:07] mardy: in which case i think i'll throw my toys out of my pram. [15:08] kenvandine, bug 1373388 [15:08] bug 1373388 in libqofono (Ubuntu RTM) "Allow user to change SIM slot technologies ( ex. 2G vs. 3G )" [Critical,In progress] https://launchpad.net/bugs/1373388 [15:08] dednick: did you go and throw the toys out? ;-) [15:09] mardy: lol. apparently my network did it for me. [15:09] dednick: yes, that's what I'm doing -- I understood that it would be a supported case [15:09] mardy: can you detail exactly what the process is? every case is not covered. [15:10] you have one or 2 prompt helpers? [15:12] mardy: so: "client A" calls a trust helper ("helper B"), which creates a prompt session using "client A" pid, and creating a prompt "client C". ? [15:12] then ? [15:13] mardy: ABC not related to your ABC in the bug above [15:23] kenvandine: I like it :) [15:25] jgdx, https://launchpad.net/ubuntu/+source/libqofono/0.53-0ubuntu3 [15:25] bzoltan, cool... happy with the package name too? [15:26] bzoltan, when is that qmake build going to land in vivid? [15:27] kenvandine: it is not really a plugin as such ... more like qtcreator-template-ham2d [15:27] yeah, that's what i was thinking [15:27] kenvandine: or bacon it was [15:27] ham2d :) [15:27] hahah [15:27] almost food :) [15:27] bzoltan, or how about qtcreator-bacon2d ? [15:27] kenvandine: the qmake support depends on mvo's branch to land on click [15:27] so in case it ends up including more than just a template someday ? [15:28] bzoltan, so i was thinking my template could include some magic in deployment.pri to copy the right binary [15:28] what do you think? [15:28] kenvandine: you name it as you wish.. i would call it qtcreator-template-bacon2d and rename it with transition package when you extend it [15:28] i hate renaming packages :) [15:29] kenvandine: who does not :) [15:29] so deployment.pri could copy the right arch binary into a Bacon2D dir [15:29] or... we could statically link? [15:29] since it's compiled [15:33] kenvandine: statically linking is not possible with the open source license of the Qt :( [15:38] kenvandine, zomg thanks [15:38] bzoltan, what do you think of the idea of copying the binary in deployment.pri? [15:39] is that the right place? [15:39] my template could include that [15:39] anyone could solve me a doubt? [15:41] RMH-SIK, let's hear 'em! === chihchun is now known as chihchun_afk [15:54] kenvandine: sounds doable, why not. Thou I am not sure if it is custom to do so or if there is a known way to deal with such library distribution. [15:55] not that i know of... [15:55] bzoltan, mind if i stick my package in the same PPA as the ubuntu plugin? === dandrader is now known as dandrader|lunch [15:59] kenvandine: no, feel free [16:00] dobey mardy I updated the bug and added a video of what is happening [16:00] https://bugs.launchpad.net/ubuntu/+source/ubuntu-system-settings-online-accounts/+bug/1389028 [16:00] Launchpad bug 1389028 in ubuntu-system-settings-online-accounts (Ubuntu) "u1 account crashes and gets removed when installing app" [Undecided,Incomplete] [16:01] dobey mardy sorry about the quality though :) [16:03] kenvandine, them packages worked perfectly: http://pastebin.ubuntu.com/9150930/ [16:03] on the device, ril0 ril1 respectively [16:04] woot [16:04] matv1: go to https://login.ubuntu.com/+applications and delete the entry there for "Ubuntu One @ ubuntu-phablet" then try again, please [16:06] dobey ahah! you seem to be on the money [16:06] dobey there seem to be 2 account registrations for ubuntu-phablet [16:07] dobey that does not seem right, right? [16:10] matv1: indeed [16:10] although i don't know how you would have gotten two of them [16:10] but delete them, try again, and it should work [16:11] dobey right. how could that happen? [16:11] will do [16:11] no idea :) [16:11] by doing a "backup restore" of the homedir :) [16:11] you can only add one account on a device, and the server refreshes the existing token, when you log in from the same device [16:11] (whcih is one of the reasons we dont support that yet) [16:11] ogra_: nope, not even that. this is tokens on the server [16:12] ah [16:12] and the way the auth scheme works, this shouldn't ever happen. so seems like something was wacky with the account on the server side [16:13] it should happen when any install was attempted though, not just grooveshark [16:13] so i think that is just an oddity [16:13] renatu, bfiller: ok, i'm happy with the patch [16:13] just tell me when to pull the trigger to upload :) [16:13] kenvandine: running through the test plan, so far so good. will let you know soon [16:14] tedg, I hear you've manually created a fat click package? [16:14] dobey ogra as I said in the bug. I was able to reproduce this wth that specific app for the past 3 rtm images and also on proposed [16:15] just for grooveshark consistently [16:15] matv1: yes, but it doesn't have anything to do with that app [16:15] balloons, Sure, but more importantly I wrote the code that sets up running them :-) [16:16] balloons, What's up? [16:16] tedg, ahh, well let's talk then. I'm trying to create some fat packages for the core apps and put them in the store. Fingers crossed, they will magically appear then on the phone and desktop stores [16:17] matv1: anyway, i take it that after deleting the tokens on the server, the issue is gone now? [16:17] well, I should say.. I will be able to install the i386 version on my desktop [16:17] Sure [16:17] tedg, so my initial thought was to build traditionally an x86 click, and an armhf click. Then put the builds together and repackage it [16:18] cant you be more polite ... thats so discriminating, they only have heavy bones, they arent *FAT* [16:18] what should I be doing? [16:18] balloons, That's *basically* it, but you need to make sure everything that is arch specific is in "$(root)/lib/$(arch triplet)" if not already. [16:19] balloons, For instance there can be a bin directory there for the binary, and we'll choose the right one from the exec file. [16:19] If you're already installing them there, then it's just a merge. [16:19] But you might not be putting them there. [16:20] NOTE: It'll work fine even on non-*big boned* packages if they're in the arch triplet directory. [16:20] tedg, what do you mean big-boned? [16:20] balloons, Definitely not you ;-) [16:20] balloons, ogra_'s comment above ^ [16:21] :-) [16:21] * balloons should have suspected ogra_ [16:21] heh [16:22] dobey yes it works again now. Thanks very much [16:25] kenvandine: all the tests pass, good to push [16:29] woot [17:04] tedg, how did you specify multi-arch in the click manifest? [17:05] balloons, Hmm, I'm not sure on that one. [17:05] balloons, http://click.readthedocs.org/en/latest/file-format.html#manifest [17:06] balloons, "architecture" [17:07] tedg, ok that seems to ring with listing them out === psivaa-holiday is now known as psivaa [17:10] balloons, Yeah, we don't check that at run time. Probably just install and in the store. [17:11] ping jdstrand [17:18] could someone remind me how to upgrade to the latest version of ubuntu touch ... sorry for asking, I have read 2 of the links in the topic but have limited time and am preparing a demo [17:18] (running on nexus 4) [17:19] ubuntu-device-flash --channel=devel [17:19] popey: thank you! [17:19] np [17:43] do I need an RTM promotion before I'll start seeing paid apps? [17:45]