=== iveand is now known as rikshaw76 [00:20] <_polto_> hi === chihchun_afk is now known as chihchun [00:25] <_polto_> I apt-get upgraded my nexus7 and wifi is no more available. If I do "ifconfig wlan0 up" I get an "Operation not possible due to RF-kill" so I did "rfkill unblock all", after that ifconfig & iwconfig works until the next reboot and wifi manager do not show up. I have to configure wifi manually from the terminal. === sporkeee is now known as wilee-nilee === wilee-nilee is now known as 17SADBGVJ === freeflying is now known as freeflying_away === 17SADBGVJ is now known as wilee-nilee [01:10] HI all [01:10] How much space does Ubuntu touch need? [01:10] I'm thinking of buying a Nexus 4 to run Ubuntu Touch, wondering if the 8GB version is enough? [01:14] Anyone? [01:14] <_polto_> Xeddy, Nexus4 is officially supported. so yes it's enough. [01:15] _polto_: Thanks! But do you know how much space the OS itself uses, or what the partition structure is like === claydoh_ is now known as claydoh [01:33] <_polto_> Xeddy, I do not remember. but I guess < 2G [01:40] is there a repo with the mali drivers for the Nexus 10? [02:01] hi, does anyone knows if ubuntu touch works smoothly on the last Nexus 4 4.3 build JWR66Y? [02:02] dangpzanco: it's not a matter of android build, that gets replaced [02:02] wiped clean [02:03] dangpzanco, the touch is in development, it is a bit rough. [02:03] where can I get the drivers for the Mali T604 of the Nexus 10? [02:03] rsalveti: indicator works for me right now [02:04] and I changed nothing besides where I was sitting [02:04] cyphermox: clean flash with 27.2? [02:04] nmcli still works... so I can't explain what's up [02:04] clean flash an hour or so ago? [02:04] yeah, that should be 27.2 [02:04] well, seems quite a few folks are able to reproduce the issue [02:04] wonder if that could be a race somehow [02:05] can I see what image from the phone itself? [02:06] this is starting to be really ridiculous, I haven't landed anything on the phone in two weeks, so it has to be external to NM [02:06] thanks for the tips guys! [02:07] rsalveti: is it possible that GFileMonitors don't work on the phone? [02:07] as in, inotify watches? [02:07] cyphermox: /etc/media-info [02:07] well, but then it'd always fail [02:08] Ubuntu Saucy Salamander (development branch) - armhf (20130827.2) [02:08] you mean /var/log/installer/media-info certainly [02:08] hmm, weird, I have an image from the 22 [02:08] cdimage-touch no? [02:09] anyway, I uploaded NM the last time prior to the 22 [02:10] right, it was on the 20th [02:10] so if it's something I changed, you would have to have started seeing it on the 21th at least, and the image 20130822 seems to work [02:10] I'll reflash later with --pending, once I get some more mtp work done [02:11] cyphermox: flash with --pending [02:11] yeah [02:11] that's why it works for you [02:11] 23.1 was the one I noticed that was broken [02:12] well, what changed then? [02:13] cyphermox, I have one phone that works, one that doesn't [02:13] I looked at all the obvious stuff ( polkit config, dbus conf file ) and couldn't find any diffs [02:14] yeah, not sure [02:15] dbus send is failing w/access_denied when run as 'phablet' [02:15] works fine if run as root [02:15] this has got to be something in logind then, could it have grown proper session knowledge on touch? [02:15] maybe [02:15] has logind landed? [02:16] don't see it [02:16] now I do [02:16] doesn't seem to be the issue, it's too recently changed [02:17] bug 1217650 [02:17] bug 1217650 in ubuntu-ui-toolkit (Ubuntu) "Unity8 does not start with the latest ubuntu-ui-toolkit packages" [Critical,New] https://launchpad.net/bugs/1217650 [02:17] what else would cause access_denied for dbus_send? [02:17] is that already known or am I the first to encounter this bug ? [02:17] awe: nothing if the config is correct [02:18] assuming dbus itself hasn't changed, and it doesn't seem to have [02:18] https://launchpad.net/ubuntu/+source/dbus [02:18] tyhicks uploaded one earlier, and there is a ubuntu3 missing, but I think that's just a +1 error [02:19] again, too new to be the issue [02:19] same version of both [02:19] 1.6.12-0ubuntu2 [02:19] already discussed with tyhicks, his stuff hasn't landed yet [02:19] is this issue specific to maguro or to mako, or happens on both? [02:19] yeah I know [02:20] both [02:20] my maguro is working [02:20] did a fresh install on mako earlier, and it's broken [02:20] how can I get the mali drivers package for the nexus 10? [02:21] mijk: rsalveti might know ^^^ [02:22] mijk: you can download the binaries from google directly, let me find the link [02:22] for Android or Ubuntu? [02:22] https://developers.google.com/android/nexus/drivers#mako [02:22] for android [02:22] which is what we use in ubuntu touch [02:23] oooh, okay so it does work in ubuntu [02:26] so you have to manually move stuff around and edit xorg.conf I gather? [02:27] nops, we don't use x11 with touch [02:27] we're currently using surface flinger still, but will move to use mir soon [02:27] which uses the android libraries over hybris [02:30] k gotcha [02:30] thanks [02:31] hell yes! I got some images showing over MTP :D [02:33] sizes are good, image names are good, just need to fix up the actual file transfer fun === freeflying_away is now known as freeflying === Namidairo`bnc is now known as Namidairo === piscodig is now known as discopig [03:42] so has pulse been changed.. etc? [03:55] hello [03:55] sudo-update [03:56] hello anyone here? === nhaines_ is now known as nhaines [05:47] * snwh is away: Away [05:52] * snwh is back (gone 00:04:41) [06:07] anybody know how to add some udev rules to the ubuntu-root directory as a part of the build process? [07:24] rsalveti: hi ... anything found on maguro? [07:27] Mirv: morning [07:32] hello asac [07:37] Mirv: see -unity :) [07:37] nik90: bug 1217743 is fun ☻ [07:37] bug 1217743 in Ubuntu Clock App "Stopwatch runs backwards" [Undecided,New] https://launchpad.net/bugs/1217743 === schwuk_away is now known as schwuk [07:57] ogra_: we got shot :) [07:57] stop the image presses :) [07:57] shot ? === Namidairo is now known as Namidairo`bnc === Namidairo`bnc is now known as Namidairo [08:18] hi diwic, good morning. It seems that on Nexus 4 headphones don't work (that's on an image from yesterday evening). Do you know if this is this a known bug, or is there is a workaround? [08:26] diwic: !! [08:26] diwic: my hero :) [08:26] diwic: i can test whatever is needed to fix pulse on maguro :) [08:26] i want to get to a point where i can ship an image without breaking basic call ability :) [08:26] diwic: so i am here, maguro is running .2 [08:27] whatever it takes, i am in :) [08:28] asac, you have no SIM though :) [08:28] yeah... [08:28] diwic, where are the fixes ? [08:28] but i can at least try playuing sound in settings menu :) [08:28] ogra is better to test the whole thing for sure [08:28] yep [08:28] he also is more professional and less distracted :) [08:29] lol [08:33] ogra_, the new UCM files for Maguro? Can I send them in an email to you, or do you prefer seomthing else? [08:33] * diwic reads scrollback [08:33] asac, I can send the files to you too if you like :-) [08:34] diwic: if ogra is avail, try to use him :) [08:34] but happy to do the doulbe check [08:34] dpm, I'm looking at it now [08:38] oh, i didnt see them on the mail, sorry [08:38] ogra_, asac I just sent an email to both of you, feel free to coordinate testing efforts :-) [08:38] yeah [08:38] there are changes both on the HiFi and VoiceCall sides [08:39] dpm, but completely non-working I can't reproduce. Here, the left side headphone is working but not the right side [08:41] diwic: i think ogra_ coordinates himself :) [08:41] i will cehck in a bit [08:41] diwic: with that things are supposed to work somewhat? [08:41] * ogra_ reboots with the new files in place [08:41] asac, hopefully even better than "somewhat" [08:42] but let's see [08:42] i'm not sure how much i messed up that image already though .... i'll do a test with a freschly flashed image afterwards [08:42] * asac reboots [08:42] after copying files into that dir [08:43] yay, looks good [08:43] sounds in system-settings finally work [08:43] paplay output comes from speaker [08:44] diwic: hmm. doesnt work for me in system-settings [08:44] diwic: do i need to do something else? [08:44] maybe i need a fresh install? [08:44] anyway if it works for ogra [08:44] then i did something wrong [08:45] let me reflash [08:45] ringtone, call mic and earpiece work, speakerphone works [08:45] sintel plays but gets choppy after a while [08:45] sintel, is that a video? [08:45] volume control through the volume rockers doesnt work [08:45] yeah [08:45] the first from the three we ship === arunprasadr_away is now known as arunprasadr [08:46] ogra_: nice... seems we want that :) [08:46] i guess it now uses SW decoding ... until the gstreamer patches are in [08:46] ogra_: so lets land a) these ucm files [08:46] yep [08:46] b) ui-toolkit [08:46] asac, gimme 30min to test with a freshly flashed image [08:46] cool [08:47] ogra_: i think its safe to already upload though, no? [08:47] doesnt waste much time and in case its good we would be ready [08:47] earlier [08:47] in case its bad it surely isnt worse than now [08:47] asac, yeah, its definitely an improvement, but i messed around a lot with the current image last night to try to debug sound [08:48] so i'd feel safer to test with a fresh flash [08:48] ogra_, are you getting the same exception stack at boot time as Ricardo reported? [08:49] diwic, yes, every 100st boot or so [08:49] we have seen driver issues before that looked very similar [08:49] there i had a similar stack and a reboot when the driver loaded [08:51] Behavior on listView | http://askubuntu.com/q/338276 [08:55] <_polto_> guys, any chance to get webgl working in the browser ? You can test here : http://www3.elphel.com/eyesis4pi_samples [08:56] which parameter should I use with phablet-flash? [08:56] what's the difference between the builds? [08:56] Laney: hey [08:56] you get the std flipped image with cdimage-touch [08:56] Laney: you are about to land gstreamer major bump? [08:56] asac: yes [08:56] Laney: can we coordinate that for a sec? :) [08:57] Laney: what i need to be aware of is the known impact this has on image builds and our tests [08:57] It'll be in proposed for a bit [08:57] can you do a build with proposed to find out? [08:57] Laney: no [08:57] ogra_: ok, thanks. I'll use that one now. [08:57] Laney: is it already in proposed? [08:57] no [08:57] e.g. damage done? [08:57] so, how then?' [08:57] Laney: ok ... lets talk in /msg :) [08:58] proposed isn't damage done either [08:58] ogra_: I guess I should be switching to ubuntu-system in the near future? or is that not done yet? [08:58] timp, should be done next week i think [08:58] until then go with the normal flipped one [08:58] ok, thanks [09:01] diwic, asac, thumbs up ... all working [09:01] oh, i heavet tested headset ... one sec === chihchun is now known as chihchun_afk [09:02] nice [09:02] ogra_: so we would like to try gstreamer staging packages [09:02] ogra_: to understand the impact of this landing in the images [09:02] before pulling the trigger [09:02] ok, headset only semi-works [09:03] asac, i think rsalveti has done all that already, hasnt he ? [09:03] ogra_, can you define "semi-works" ? [09:03] ogra_: well, i feel like i cant get enough insurance that something we want to land works [09:03] given that we want to get an image :) [09:04] diwic: can you help test the gstreamer stack on on maguro? [09:04] diwic: the staged packages are here: https://launchpad.net/~ubuntu-desktop/+archive/gstreamer-1.1/+packages [09:04] the idea is that whatever usese gstreamer now will start working :) [09:04] asac, well, unless you prefer me to work on the music hub... [09:04] ok [09:04] diwic: lets see if ogra can help. i couldnt get the sound workig [09:04] and fix up all other audio bugs [09:05] so i am a bad tester [09:05] diwic: yeah go for that for now [09:05] we will find someone else :) [09:05] ogra_: so if this gstreamer goes up, will we automatically start using it? [09:05] or is there something else to this transition? [09:05] diwic, headphones work if i plug them in during the call (or re-plug) ... mic stays on the phone and doesnt get switched to headset at all [09:06] asac, no idea, rsalveti has the details [09:06] i assume we will need to change the seeds and some deps [09:06] ogra_: right now mediaplayer links against 0.10 [09:06] i dont think that will pick it up automatically :) [09:06] right [09:06] * asac hopes that was not forgotten [09:06] lol [09:07] ogra_: so landing this might have zero impact? aka is super safe to just land? [09:07] jhodapp has a complete multimediastack ready afaik [09:07] ogra_, can you check what the active port is in these cases (pactl list sinks, pactl list sources) to see if it flips correctly or not? [09:07] including the player :) [09:07] asac, again, i cant judge, rsalveti and Chicken worked with jhodapp on it [09:08] ogra_: jhodapp is US? [09:08] yep [09:08] ok i guess we can wait another two hours [09:08] asac, not sure what's the question, but the current gstreamer version don't work on the device === chihchun_afk is now known as chihchun [09:09] asac, those guys asked us to update so we can get support for touch [09:09] seb128, asac is scared it might work even less i guess :) [09:10] seb128: i am hoping for a straight ack from someone knowledge that pushing the staged gstreamer package [09:10] will have zero impact on our images [09:10] and ability to update the image and do new builds :)? [09:10] asac, it's going to have an impact, it might make things actually work [09:10] when they don't [09:10] i couldnt get that, so i am in conservative mode given that we have other hot things baking :) [09:11] seb128: from the look, we dont even depend on anything in there, we also dont pull in the -dev package [09:11] so i think it has zero impact until someone flipps the switch somewhere in qtmultimedia build-depends [09:11] asac, well, lesss than "not working" is still "not working" ... [09:11] but i am just not sure :) [09:11] there is nothing we can break more [09:11] seb128: so we currently only have 0.10 from what i know in th eimage [09:11] asac, correct, and that switch can't be flipped before we update gst [09:11] hmm. no we have 1.0 [09:11] not sure what epende on it [09:11] * asac checks [09:11] we have 0.10 becausee telepathy-ofono depends on it [09:12] but only one plugin [09:12] asac, libqt5-multimedia-plugins depends on 0.10 currently [09:12] right [09:12] so that one is safe afaik [09:12] and that one is in the new stack too afaik [09:12] ogra_: so what depends on 1.0? [09:12] on the image? [09:12] qt multimedia depends on 0.10 [09:12] telepathy-ofono as well [09:12] whats left? [09:13] nothing i think [09:14] ogra_: i see libgstreamer1.0-0 on the image [09:14] ogra_: do you see whats pulling that in? [09:14] mediascanner uses it i think [09:14] ogra_: whats that doing? [09:14] is that critical? [09:14] ah, no, not installed [09:14] does it work right now :)? [09:14] hmm [09:15] so maybe we just seed gstreamer1.0 for the sake of it? [09:15] isnt pulse using it? [09:15] ok so Laney says webkit probably uses it? but then i thought we use qtwebkit [09:15] webkitgtk [09:15] Can I Flash Ubuntu Touch to Wiko Cink Slim | http://askubuntu.com/q/338289 [09:15] if its just webkit then we probably can go ahead after quickly checking that stuff still works somewhat [09:16] i dont think video ever worked well in browser etc. [09:16] asac, PulseAudio isn't using gstreamer. Gstreamer is using pulseaudio. [09:16] asac, we have tons of crap we don't use on the touch image [09:16] so someone can confirm that the only place we use gstreamer 1.0 for right now is the browser? [09:16] including gtk gnome-control-center webkitgtk [09:16] oh, crap [09:17] if so we can just go ahead i feel :) ... after checking that the browser starts still :) [09:17] diwic, i testeed the wrong stuff ... [09:17] ogra_, is that good or bad news? [09:17] bad [09:17] seb128: interesting. do you know how that stuff ends up be there? [09:17] install the PPA and see if anything breaks [09:17] there aren't even the plugins so I seriously doubt anything is using gstreamer 1.0 [09:17] asac, some of this is https://code.launchpad.net/~ted/indicator-datetime/split-out-control-center/+merge/182414 [09:17] yeah [09:17] which is a bit disconcerting but there we are [09:17] diwic, the image i test still has qtubuntu-media installed ... removing it [09:18] Laney: well, i cant test all ... becaause sound doesnt work, so i wanted to understand how i could observe potential breakage [09:18] asac, some of it is https://bugs.launchpad.net/ubuntu/+source/evolution-data-server/+bug/1193018 [09:18] Launchpad bug 1193018 in evolution-data-server (Ubuntu) "GOA support not completely split" [High,Confirmed] [09:18] ah [09:18] what i hear so far its like: if you can start it and use browser and everything its probably ok [09:18] so i will just do that now :) [09:18] asac, that second bug is on my list for this week/before ff [09:18] ogra_, ok...at least it does not seem to be on the image now, does it? [09:18] asac, basically eds is built with support for the gnome online account, and that brings in gtk and webkitgtk [09:19] seb128: so i could observe potential breakage somewhere in online accounts stuff? [09:19] diwic, no, but lxc-android-config was also old ... so i was using the android stack still ... [09:19] how can i even use that code path on the phone right now? [09:19] e.g. where can i register any account [09:19] asac, GNOME online account, and no you can't [09:19] * ogra_ reboots the phone [09:19] seb128: ok so its just the backend for now? e.g. no app/frontend uses online account? [09:19] asac, that's GTK ui based and GTK doesn't run on surfacefinger/mir [09:20] asac, right, it's the backend used by the Ubuntu GNOME flavor [09:20] doesnt friends use online accounts ? [09:20] asac, it's just not properly split out [09:20] what does friends do? :) [09:20] never used it [09:20] so hard to say if it broke [09:20] * asac tries before upgrading [09:20] ogra_, asac: friends used Ubuntu Online Account [09:20] err ... where is friends :)? [09:20] in the apps lens [09:20] is there an app icon? [09:20] yes [09:20] oh got it :) [09:21] ok it sends me to system settings [09:21] lets see [09:21] you need to add a e.g facebook account in system settings first [09:21] then it lists your wall basically [09:21] and let you post [09:21] <_polto_> guys, any chance to get WebGL working in the browser ? You can test here : http://www3.elphel.com/eyesis4pi_samples [09:22] seb128: ok interesting ...that opens an embedded browser thing for auth [09:22] i assume that would be what mnight break with the4 new gstreamer then [09:22] * asac goes and tries [09:23] asac, I doubt gstreamer would break web auth :p [09:23] ogra_: where do we stand with the pulse landing? [09:23] diwic, phew, all working, even the headset ... one thing i still see is that in headset mode both mics are on [09:23] ogra_: the data files :) [09:23] are they in? :) [09:23] asac, no, still testing, see above, i just noticed that i had tested with qtubuntu-media installed [09:23] seb128: yeah. well more thinking that if something is really screwed the whole process would get busted [09:23] works wtill fine without it it seems [09:23] dont really anticipate to see a real gstreeam code triggering on this phone :) [09:24] gstreamer1.0 [09:24] ogra_: can you just upload and we finish testing while that builds? [09:24] asac, the browser on the phone use qtwebkit, that doesn't use gstreamer1 [09:24] in case its good we just saved time [09:24] otherwise we wont get a big problem either [09:24] asac, so there should be 0 chance it creates issues [09:24] diwic, ^^ got a tree for me > [09:24] ? [09:24] ogra_: i will also do a clean install then [09:25] seb128: i think its slightly greater than 0 ... but yeah. let me just do my duty and then we push this stuff [09:25] asac, thanks [09:25] ogra_, you mean to sponsor? [09:25] diwic, yeah [09:25] it'll be in proposed for a while anyway [09:25] so the touch image surely doesnt pull in the stuf from the ppa [09:25] thats odd [09:25] needs libav 9 first, see ubuntu-devel earlier [09:25] Laney: nothing gtes upgraded with dist-upgrade [09:25] ogra_, all right, one moment [09:26] oh wait :) [09:26] erm! [09:26] i didnt run update :) [09:26] just dist-upgrade [09:27] Laney, define "a while" ? [09:27] we should have this dione by end of the week [09:27] until infinity gets libav 9 uploaded [09:27] * asac thinks [09:27] Laney: do we need it in proposed if its stuck anyway? :) [09:27] maybe we should wait until we know the path is cleared? [09:28] thats what proposed is for [09:28] then it's all ready to go automatically [09:28] <_polto_> WebGL on ubuntu--touch ? [09:32] does proposed loose its memory about version numbres after we kick stuff out? e.g. can we still go back in case we find out that we badly need a minor security patch in gstreamer today? [09:32] you just can't use the same one again [09:32] Laney: but a lower versionm is ok? [09:33] ya [09:33] nice [09:33] thats good news [09:33] :) [09:35] hmm [09:35] not sure, but my phone doesnt come up after dist-upgrading to ppa [09:35] i also got other stuff from archive though i think [09:35] let me give this phone a few more minutes :) [09:38] ogra_: http://paste.ubuntu.com/6035990/ [09:38] thats what got updated [09:38] ogra_, hmm, need to sync up the branch with your upload yesterday [09:38] hmm wait [09:38] asac, upgrade or dist-upgrade ? [09:38] and are you on .1 or .2 [09:39] ogra_: dist-upgrade [09:39] on top of .2 [09:39] ogra_: http://paste.ubuntu.com/6035992/ [09:39] maybe i am wrong though [09:39] let me fix that by flashing from scratch [09:40] what are you trying to achieve ? [09:40] ogra_: upgrade to gstreamer staging ppa [09:40] and check that stuff and online accounts still works :) [09:40] e.g. doesnt crash badly [09:40] * ogra_ wonders if thats worth it without knowing the full patchset [09:41] ogra_: well, i just want to forsee the future: basically trying what laney wants to land :) [09:41] this helps me understand the impact of that landing [09:41] the hope ws that impact is zero :) [09:41] right, but you test something that might need patches in other places of the stack [09:41] so i just wanted to double check before going ahead and push gstrteamer [09:41] ogra_: well, then its not ready :) [09:41] i would wait for jhodapp or rsalveti, really [09:41] if our uploads busts the whole phone it cant go in [09:41] if it doesnt it can go in [09:41] well, what you double check will be incomplete [09:42] then we shouldnt upload :) [09:42] so thats a waste of time if there are app patches you miss [09:42] well, if it has zero effect [09:42] then its good to have it in already [09:42] why not [09:42] its a whole stack, three people work on to get in until friday [09:42] less to do later :) [09:42] right. and laney wanted to land this stuff [09:42] you try to test one small piece of that stack out of context [09:42] so that rsalveti can pick it up when he gets up [09:42] so i just want to see if we can land this stuff without bricking our image ... simple as that :) [09:43] land == upload to archive [09:43] yes, and i say we might need Qt patches, or app patches that you will miss [09:43] how can you judge by that if the transition is good or bad [09:43] ogra_: right. but Laney wanted to upload and they said it will help other folks so i try to make that happen :) [09:43] mardy, hey [09:43] ogra_: if the phone doesnt regress i dont care if its complete [09:44] ogra_: so you can do everything in archive and at best at the end upload a final, atomic switch [09:44] the phone will most likely regress for a moment ... [09:44] mardy, is there a chance you could update the shotwell uoa patch for http://www.yorba.org/download/shotwell/0.15/shotwell-0.15.0pr1.tar.xz ? [09:44] i dont think so [09:44] if it does the laney upload cant go in before rsalveti is back up [09:44] simply because stuff needs to be in place first to depend on it with the patched packages [09:44] but i think it can go in from what i heard :) [09:45] err... thats upside down. you can also make the packages avawilable first [09:45] and then land the dependencies :) [09:45] sure [09:45] from what i undersatnd thats what we try to do here :) [09:45] my device also doesn't get unity after dist-upgrading [09:45] so lets see if just this upload does any harm [09:45] but thats not how it is done atm [09:45] otherwise its green light and wont torpedo our dashboard [09:45] there's some suspicious qml errors in .cache/upstart/unity8.log [09:46] ogra_: so you want me to hold back the gstreamer upload as it will bust the imge? [09:46] thats fine with me :) [09:46] just wanted to be nice [09:46] wtf [09:46] asac, no, i want you to wait for someone who knows the transition before you judge it by piecemeal [09:46] Laney, that's likely the ubuntu-ui-toolkit update from yesterday [09:46] seems likely [09:46] seb128: i thought that is now in? [09:46] the revert [09:46] i'm downgrading it to check [09:46] Laney, they reverted the commit this morning [09:47] ogra_, ok, pushed to lp:~ubuntu-audio-dev/alsa-lib/ubuntu [09:47] seb128: hi! Sure, I'll try updating the patch [09:47] seb128: can you check if the package made it through proposed? [09:47] diwic, great, pulling and uploading then [09:47] mardy, thanks [09:47] 3 minutes ago [09:47] https://launchpad.net/ubuntu/+source/ubuntu-ui-toolkit/0.1.46+13.10.20130828-0ubuntu1 [09:47] that upload has no changelog! [09:47] asac, just made it [09:47] the little horrors [09:47] ok [09:47] so if i now dist-upgrade i will get it? [09:47] :) [09:47] seb128: ? [09:47] Laney, they pushed directly to trunk to revert [09:47] asac, you should yes [09:48] not likely [09:48] Laney, I guess that confused things [09:48] it probably has to publish [09:48] ok cool. /me waaits for his phone to be fresh and shiny before dist-upgrading a) distro only first and b) then gstreamer [09:48] and definitely has to be pushed to mirrors [09:48] :) [09:48] seb128: ui-toolkit is still on http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html [09:48] guess that is trailing? [09:49] asac, that was the previous run and it has "Valid candidate " [09:49] asac, it got cleared in that run [09:49] ah so the page could say: Valid Candidation: now copying ... [09:49] ? [09:49] asac, no, because then you have http://people.canonical.com/~ubuntu-archive/proposed-migration/update_output.txt [09:50] asac, details are written in https://wiki.ubuntu.com/ProposedMigration [09:50] ok that is getting too cryptic for now [09:50] ok manually installing that works [09:50] will check that later [09:50] gstreamer didn't break it for me [09:50] diwic, hmm, that tree looks messed up :( [09:50] Laney, good [09:50] Laney: have you tried online acccounts login? [09:50] in settings? [09:51] e.g. the webbrowser there? [09:51] ogra_, in what way? [09:51] diwic, i thought you pulled in my commit for the upload [09:51] if that doesnt crash i am happy :) [09:51] I loaded that [09:51] and got a facebook login page [09:51] ogra_, you never pushed that commit [09:51] i didn't actually add the account [09:51] Laney: are you mako or maguro? [09:51] diwic, i got merge conflicts all over the place here with bzr pull [09:51] none of those [09:51] Laney: yeah guess thats fine [09:51] grouper [09:51] Laney: oh. [09:51] diwic, eeek, then it should have been pushed first [09:51] ogra_, so all I did was to change UNRELEASED to saucy [09:52] Laney: ok go ahead then i guess... we can still kick it out of proposed as its waiting :) [09:52] just getting the last package together [09:53] hehe [09:53] i assume thats the risky package you tried to hide all times :) [09:53] diwic, well, we need to clean that up i guess or luke will get a heart attack [09:53] no it's already in the ppa [09:53] it's -bad but I need to remove jhodapp's patches [09:54] ogra_, are you a bzr wizard? I'm not. [09:54] * diwic deals with more git than bzr usually. [09:54] diwic, not exactly a wizard but i guess with some forced overwrites i can bring that back in sync [09:55] ogra_, then you know more than I do, I don't know how to do "some forced overwrites" [09:55] that will take a while though [09:55] no new build today? [09:55] uhm [09:55] Error: nmcli (0.9.8.0) and NetworkManager (unknown) versions don't match. Force execution using --nocheck, but the results are unpredictable. [09:56] bah, why is the UDD branch out of sync too [09:56] hahahaha [09:56] phablet@ubuntu-phablet:~/Music$ apt-cache policy network-manager [09:56] network-manager: Installed: 0.9.8.0-0ubuntu19 [09:56] sorry, your surprise amused me :P [09:56] Ubuntu Saucy Salamander (development branch) - armhf (20130827.2) [09:56] seems a bit broken [09:57] just on the surface :P [09:58] wtf [09:59] ogra_: so we cant even upload three bloody data files to pulse :)? [09:59] lol [09:59] so https://launchpad.net/ubuntu/+source/alsa-lib clearly shows 1.0.27.2-1ubuntu2 is in saucy ... and thats what i get from the archive ... yet the UDD branch is on -1ubuntu1 [09:59] asac, alsa, not pulse [09:59] yeah that :) [09:59] ogra_, because you never pushed that commit [09:59] the branch is messed up ... but i cant find a clean branch to overwrite from [10:00] ogra_, I don't think it's that messed up [10:00] Laney: so you said you got it back to working after installing the new compoents? [10:00] diwic, UDD branches are aut-created from uploaded packages [10:00] i did a dist-upgrade and see the same issue [10:00] ogra_, what's wrong about it, except that you don't have a commit in it? [10:00] diwic, the upload should automatically have created a commit of my upload that i should just be able to easily merge [10:01] Laney: never mind... guess mirrosr still need time here [10:01] ogra@anubis:~/Devel/branches$ bzr branch lp:ubuntu/alsa-lib alsa-udd [10:01] Most recent Ubuntu version: 1.0.27.2-1ubuntu2 [10:01] Packaging branch version: 1.0.27.2-1ubuntu1 [10:01] Packaging branch status: OUT-OF-DATE [10:01] diwic, that theoretically cant happen ^^^ [10:01] ogra_, well, in theory, anubis is dead ;-) [10:02] * ogra_ tries to create a commit manually from the package then [10:02] ogra_, maybe the version has ascended to a higher plane of existence [10:02] diwic, just a few 100 years though [10:02] err [10:02] 1000 [10:02] haha [10:03] asac: I just wgetted it from lp [10:07] geez. whats up with this branch [10:26] want to load on bionic (targa), but seems to be just a black screen... [10:27] so mostly anything related to sound hangs on the HTC vision, without the possibility to ctrl-c, or even kill -9... even "cat /dev/snd/pcmC0D0p" hangs. so no aplay, pulseaudio, etc.. how is that possible? kernel driver bug? [10:28] can i load the img in VBox? [10:46] diwic, i cant even manually pull your patches into the alsa-lib package, something is seriously borked :( [10:47] ogra_, okay, so what do you suggest we do? [10:47] http://paste.ubuntu.com/6036155/ [10:47] this is just after copying the content of debian/patches and the changelog over [10:48] i dont really get why it does that since only the patches changed, not the actual files it lists [10:48] ERROR:phablet-flash:Unsupported device, autodetect fails device [10:48] ogra_, try cleaning the build-area directory and try again [10:48] ^ phablet-flash doesn't tell me anymore how to pass the device type [10:48] is the option still there? [10:48] and what was it? [10:49] diwic, which build area ? [10:49] alex-abreu: unity-webapps needs to be rebuilt against the new packagekit in saucy-proposed. will it cause a problem if I just upload a no-change rebuild directly to the Ubuntu archive? [10:49] alex-abreu: sorry, libunity-webapps I mean [10:49] diwic, i did apt-get source alsa-lib and copied the 4 files into the debian dir ... i didnt do anything more ... that gets me such an output [10:50] ogra_, after apt-get source alsa-lib, unapply (quilt pop -a), then copy the files [10:50] uh, why are they applied on unpack ? [10:50] ok, rebooting my device helped. after reboot it is detected as maguro [10:50] that seems rather wrong [10:51] that's standard for 3.0 (quilt) [10:52] timp: pass it in _after_ cdimage-touch or whatever you are using [10:52] oh [10:52] diwic, uploaded === sergiuse1s is now known as sergiusens [10:53] sergiusens: phablet-flash --help doesn't show the options anymore [10:53] sergiusens: ahh... nevermind I got it now :) [10:54] sergiusens: --help needs to be passed after cdimage-touch also [10:54] "Use -h or --help after each command to learn about their provisioning options." [10:54] is what it says at the bottom of "phablet-flash --help" [10:54] yes I see now. [10:55] should have an additional newline so it isnt glued to the prompt ... [10:55] that would make it more visible [10:55] anyone else having trouble launching click apps (from the store)? [10:55] or list something like usage: phablet-flash [image] [options] [10:55] ogra_, okay, thanks. Did you end up making any changes to the branch? [10:56] diwic, no, i just blindly pulled your changes into the package [10:56] ogra_, cool. [10:56] i still would like to knwo why UDD is broken though [10:56] * diwic goes for lunch now [11:00] ogra_: is the stuff up? [11:00] can we kick off a new image once its in? [11:00] pulse/also fix that is :) [11:00] dunno, how about the other breakage [11:00] and yes, it is uploaded [11:00] ogra_: which one? [11:00] ogra_: ui-toolkit is fixed [11:00] i verified that [11:00] that one :) [11:00] good [11:01] sp in ~1hi should be able to re-spin [11:01] s/sp/so/ [11:01] ah, I guess there's no EU vanguard for webapps ... [11:03] * cjwatson tries #webapps internal [11:05] cjwatson, what's the issue with webapps? [11:05] needing to do a no-change rebuild [11:05] seb128: I'm asking vrruiz on #webapps - but as above, I need to get libunity-webapps rebuilt against the new packagekit in -proposed [11:06] ah [11:06] if there'll be an autorelease today, I'm fine with that, otherwise I'd like to bypass it [11:06] cjwatson, there should be one: http://10.97.0.1:8080/view/cu2d/view/Head/view/WebApps/ [11:06] the stack's currently blocked on a couple of things by the looks of it [11:07] sil2100, ^ do you plan to push that once I preNEW review the cordova packages? [11:07] cjwatson, it's yellow, which means a coredev should hack the packaging changes [11:07] ah [11:07] cjwatson, give me 5 minutes to review the new sources in there [11:07] cool [11:07] Yes, but not only - we need those 2 new packages pre-NEWed [11:07] alecu: I get the same error with xda app [11:07] alecu: (no manifest found) [11:08] cjwatson: so, a rebuild is needed, yes? [11:08] cjwatson: of libunity-webapps? [11:09] sil2100, shrugh, those are non trivial to review, can't we just publish without those new packages? [11:09] sil2100, why did they get added to the config before review? [11:09] seb128: this is something robru did without knowing ;/ When he's vanguarding there's usually no one to check if he's not doing anything strange [11:09] seb128: we already pointed that out to him... I'll remove those from the config for now [11:10] Redeploy and rebuild [11:10] sil2100, thanks [11:10] seb128, cjwatson: I'll also make sure libunity-webapps is rebuilt [11:12] sil2100: yep. thanks [11:13] rvr: ah, didn't realise you were here too [11:13] :) [11:13] so if sil2100 will take care of this then that's fine [11:13] Doing that now [11:15] alecu: actually, I can see the click doesn't get installed upon download, even if the UI seems to be happy [11:15] doens't show up in click list === alan_g is now known as alan_g|afk [11:19] ogra_: stuff ready>? [11:20] nope [11:20] as i said, 1h or more [11:20] stop being so impationet :P [11:20] ogra_: well, one can hope for speedup :) [11:20] ogra_: well, just scared about the next bomb landing [11:20] proposed migration takes its time [11:20] :) [11:22] https://launchpad.net/ubuntu/saucy/+source/alsa-lib/1.0.27.2-1ubuntu3 .... built already [11:23] sil2100: Mirv: can we put daily-release in semi-manual mode? [11:23] sil2100: Mirv: until we have gotten back to green image? i just want to know about stuff before landing while we are producing this golden image [11:23] but maybe its too lte [11:23] lets do that next time [11:23] and wait and prey on this one :/ [11:24] sil2100: Mirv: maybe it would be a good idea to put evertyhing in manual until we are past the FF rush :) [11:24] well, thats ewhat you get having a vUDS the same week you have FF [11:25] asac: we've enough breakage that it's in semimanual mode. but sil2100 is handling this tick now. [11:25] everyone rushes their stuff in ... to be free for UDS stuff [11:25] Golden image you say? [11:25] asac: hmmm [11:25] sil2100: Mirv: can you run through me stuff you push? [11:25] sil2100, yeah, we thought we'd do the final release earlier this cycle [11:25] :P [11:26] asac: let's think about that after this tick, since it's already running and we can't change much in how we release - but I'll be poking you whenever there's a manual publication, so that you can also ACK it [11:26] asac: I need to release the new webapps now for instance, since I have been asked to [11:26] sil2100: whats the potential impact on the image? [11:27] webapps could start working ? [11:27] thats the best case [11:27] perhaps ... [11:27] sil2100: yeah. so we have a broken image. i think its valid to say to your stakeholders that due to that we have to coordinate every landing [11:28] and double cehck that it doesnt make things worse [11:28] :) === alan_g|afk is now known as alan_g [11:28] but since its started [11:28] we can only prey and firefight in case nwo :) === tricia is now known as Guest69198 [11:29] asac: I think it's safe to say in my session Qt 5.1 isn't going to be uploaded before FF? ;) [11:29] When will Ubuntu Touch be available for Xperia Tipo release | http://askubuntu.com/q/338342 [11:29] Mirv: why do you want to say that? [11:30] Mirv: is that important :)? [11:30] Mirv, did you notice that Qt was excluded from the general feature freeze exception ? [11:30] Mirv: i think its safe to say that we plan to land qt 5.1 [11:30] so better do it now or you will have to do paperwork [11:30] before 13.10 and that we are working hard to make that happen as a non event] [11:30] for users and developers :) [11:30] you can say that we put quality high and that we are happy to accept paperwork [11:30] :) [11:30] lool: is this with ubuntu-system or cdimage-touch? === gusch is now known as gusch|lunch [11:31] asac, well, that doesnt mean kubuntu will grant the exception :) [11:31] asac: how *badly* broken is the image right now and why ;) ? [11:31] yeah, but you can say that we will work hard to eliminate doubts and concerns [11:31] * sil2100 doesn't want to find out himself [11:31] asac: well there is a couple extra problems that come if it'd be uploaded, but generally that's news to me if we plan to land Qt 5.1, and nice to know. I haven't heard much on actual decisions. [11:31] sil2100, no calls on maguro (witing for alsa-lib to come out of proposed) [11:32] in fact no sound at all on maguro [11:32] ogra_: there are regressions, and there is a no regressions policy. no, I haven't seen any FFe actually, where'd it be? [11:32] sil2100: its not suitable for giving the image to our dogfooders. call is broken [11:32] is the biggest problem [11:32] ogra_: but I guess it's not because of the packages we push out, right? [11:32] sil2100: byut you can boot it, use it etc. [11:32] sil2100: pulse was not coming completely through daily-release [11:32] asac: right, I'll look into the config and see how to enable 'manual mode' [11:33] so yes, daily release manual doesnt protect us from everything [11:34] lool: I mean, ubuntu-system still can't do useful things with packagekit due to the lxc-android-config thing you fixed yesterday not being on the image yet [11:34] sil2100, no, but we want to make sure you dont land any regressions from daily while we knoe that only the alsa fix is missing to make the image good [11:34] asac: if, after this tick, we enable the manual-mode for all stacks (i.e. manual publishing forced every time for everything), be sure to escalate this to all the management [11:34] sil2100: yeah. will do. do you have insight which landings might get delayed? [11:34] asac: so that everyone is aware why things aren't landing automatically [11:35] asac: what do you have in mind? [11:35] Mirv, bug 1208989 [11:35] bug 1208989 in Ubuntu "[FFe] standing freeze exception for Ubuntu Touch-specific packages" [Undecided,Confirmed] https://launchpad.net/bugs/1208989 [11:36] ogra_: ok, interesting. well, I hope asac or someone can join the session in 2.5h to bring some insight into it [11:37] ie. if the known regressions are allowed or if it's continued to be waited that they get fixed [11:37] well, that will depend on the release team [11:37] @ogra_ Does today have a change with alsa\pulse? [11:37] OrokuSaki, a pending one, yeah ... waiting for it to come out of -proposed [11:38] sweet [11:38] note that we didnt have a morning image today [11:38] I always use the pending... hmmm =) [11:38] oh [11:38] gracias [11:38] * ogra_ holds it back fro a manual build once the alsa fix is there [11:38] *nod* [11:43] sil2100: just wonde3r if you have good insight in our feature pipeline ... e.g. what do we expect to land in which daily stack === alan_g is now known as alan_g|lunch [12:04] cjwatson: this is with ubuntu-system + apt-get updated to latest packages [12:04] cjwatson: so PK is read-write [12:04] How do I set that up here? Enter developer mode, upgrade, exit developer mode? [12:05] cjwatson: I personally mount -o remount,rw /, apt-get update, shutdown -h now [12:05] (apt-get dist-upgrade) [12:05] obviously, it might break in interesting ways when I get a system image update later on :-) [12:05] maybe I should just edit that one file ... [12:05] lool: does the shutdown remount / as ro ? [12:06] cjwatson: but touching writable_image is as valid a way to do it, except it's applied only on boot [12:06] alecu: / is always mounted ro unless you have set the system builder / writable image flag (used to be "developer mode" but that was confusing for app develoeprs) [12:06] Hello [12:07] cjwatson: yup, that's what I proposed yesterday; today I'll check wiht barry whether we could get new images by switching to daily-proposed [12:07] stgraber didn't want to push new images to daily if these aren't promoted to /current on cdimage [12:07] which is fair [12:07] -rw-r----- 1 syslog adm 232M Aug 28 12:04 /var/log/syslog [12:07] looks like this will eventually become problematic [12:08] lool: OK, so having rebooted with that, what do I need to do to reproduce your bug? [12:08] cjwatson: install some appstore package [12:08] cjwatson: e.g. xda develoeprs app [12:08] cjwatson: it's probably in the suggestions, or just search for "xda" in the app lense [12:08] cjwatson: the other test package I use is the "hello world" package [12:08] any developers here? [12:09] lool: yes, some partitions are very small. I used to host my build environ on the phone, but it's not fitting anymore since I started using ubuntu-system === gusch|lunch is now known as gusch [12:09] alecu: just keeping a 200 M syslog isn't acceptable on a mobile device [12:10] right [12:10] it seems we miss logrotate to gzip + trash old data [12:10] lool: just in passing, does '/var/lib/sudo auto persistent none' seem sensible to you? it's annoying having to enter my password every time when sudoing [12:10] it also seems it's getting spammed quite a bit [12:10] cjwatson, lool: the hello world package seems to be broken: the app name in the manifest has a "_" which is not allowed [12:10] cjwatson: would temporary be enough? enter it once after the reboot [12:11] cjwatson: problem with persistent stuff is that we commit to backwards compatibility "for ever" [12:11] (the package name in the manifest is ok, though. It's only the app name the one that's broken) [12:11] cjwatson: as in, no postinst to fix things up [12:11] cjwatson: "temporary" will be trashed on each boot [12:12] lool: temporary should be fine, I'll do that [12:12] thanks [12:12] alecu: Ok [12:12] jdstrand: could we change the app review scripts to block manifests with "_" in the app name? (see alecu's comment above) [12:12] beuno: would you reupload hello world with a - instead of a _ in the name? [12:12] beuno: do we also need to change the SDK template? [12:13] lool: we already do that [12:13] wait, let me read [12:13] jdstrand: this is on the app name, not the package name [12:13] jdstrand: line 5 here: http://pastebin.ubuntu.com/6034427/ [12:14] alecu: that is an error [12:14] because the appname forms part of the APPID [12:14] $pkgname_$appname_$version [12:14] it is an error, but is it enforced? [12:14] none of those individually are allowed to have an underscore [12:14] the review scripts catch that [12:14] click refuses it, but it shouldn't have got into the app store [12:15] jdstrand: exactly [12:15] maybe beuno forced it in the early testing [12:15] jdstrand: if it's already covered, we're all good there; thanks for confirming! [12:15] it was preapproved into the appstore before there were review scripts :) [12:15] right [12:15] I just wrote the review scripst last week [12:15] stgraber: hmm - "sudo: /var/lib/sudo writable by non-owner (041777), should be mode 0700" [12:15] that was uploaded to the appstore at IoM [12:16] stgraber: it would be nice if initramfs-tools-ubuntu-touch had a way to mount temporary writable-paths with a selected mode / arbitrary mount options [12:16] jdstrand: I think not, this one is more recent [12:16] jdstrand: in the IoM we were still using the "staging" app store [12:16] lool: ^- I think I may need to make it persistent to avoid this problem [12:16] alecu: can you give me a url to the package? [12:16] jdstrand: but this is on one of the two apps in the "prod" one [12:16] cjwatson: hmm that lets me think that we have to be careful with permissions because these are bind-mounts and the bind-mouted path might have different permissions [12:16] jdstrand: sure [12:17] alecu: yes, but xda and hello_world were uploaded before the review scripts were written [12:17] I iterated with folks and on the review scripts with them, but if you point me at the package, I'll run the review scripts again [12:17] lool: the bind-mounted path is right; the temporary one is a fresh tmpfs with default perms which is wrong (and in fact a security hole here) [12:17] jdstrand: https://public.apps.ubuntu.com/download/ar.com.beuno/hello-world/ar.com.beuno.hello-world-0.4.click?noauth=1 [12:17] also note, the review scripts are not yet automated [12:18] lool: so I really think I need to make this persistent [12:18] jdstrand: yes, running the review script on both apps seems very reasonable [12:18] jdstrand: the url for the xda app is: https://public.apps.ubuntu.com/download/com.ubuntu.developer.mhall119/xda-developers-app/com.ubuntu.developer.mhall119.xda-developers-app-0.1.5.click?noauth=1 [12:19] cjwatson: like /android/ is +t [12:19] alecu: xda-developers should be fixed. hello-world is not.. There are a number of problems: http://paste.ubuntu.com/6036415/ [12:19] jdstrand: just in case you need to get those latter, they are logged in .cache/ubuntu-download-manager/ubuntu-download-manager.log after you try to install. [12:19] cjwatson: /android/data is system:system 751 [12:20] lool: /android/data is persistent ... [12:20] cjwatson: sorry looking at wrong tree [12:20] alecu: it looks like beuno didn't update the package yet. beuno, can you take care of that ^ [12:21] cjwatson: /userdata is system:system with system-data under it [12:21] lool: /userdata isn't handled through writable-paths ... I'm not sure what you're trying to say here [12:21] jdstrand: ah, there it is: "malformed application name: 'hello_world'". Thanks! [12:22] cjwatson: what I mean is that security wise there are two ways to get to the sudo data [12:22] cjwatson: one is /var/lib/sudo which gets bind-mounted with wront permission [12:22] jdstrand: I can certainly use that reviewing script, is it in lp? [12:22] lool: when I mount /var/lib/sudo persistent, it's root:root 700 [12:22] and another way, that sudo can't test for, is via /userdata [12:22] which is correct [12:22] /userdata/system-data/var/lib/sudo is also root:root 700 [12:22] I don't see the problem here [12:23] bind mounts share the permissions of the bound-from location - checking any one of them is sufficient [12:23] cjwatson, no you can upload a no change to libunity-webapps ... thx for the heads up :) [12:23] lool: to be clear, when I bind mount, it's fine, but "temporary" mounts a fresh tmpfs which has wrong permissions [12:23] beuno: the new sdk will create a proper manifest, but 2.7.1-0ubuntu7 isn't enough (what is in saucy) [12:23] so this needs to be persistent [12:23] cjwatson: I understand why tmpfs doesn't work in terms of permissions [12:24] dholbach: did bzoltan talk to you about uploading a 2.7.1-0ubuntu8 to do the final touches on the manifest (ie, fix the underscore bug)? [12:24] cjwatson: so if I'm in system, I can change /userdata/system-data to be another hierarchy of my chosing [12:24] jdstrand:as far as I know it is happening [12:24] lool: realistically I think system is root-equivalent [12:24] and on next boot this will be bind-mounted to /var/lib/sudo [12:25] bzoltan: thanks [12:25] cjwatson: ok, then I think we're good; that's the only thing I feared [12:25] dholbach: can you make sure that happens? :) (if it is someone else I should poke, let me know) [12:25] cjwatson: that other user/groups were involved in the place getting bind-mounted, introducing a weak path, but if system is as strong as root, I don't see a way [12:25] lool: to name just the first thing I notice, you could also use it to replace /userdata/system-data/etc/default/ufw, which is sourced by a root shell [12:26] cjwatson: right, not specific to sudo, it's just that it made me wonder about this [12:26] dholbach: also, would it make sense to have alecu added to canonical-click-reviewers? [12:26] especially since some directories have funky permissions [12:27] lool: I assume that the system uid/gid exists for some reason, since from the Ubuntu point of view it clearly should just be root [12:28] alecu: fyi http://paste.ubuntu.com/6036434/ [12:29] alecu: that one still has blockable issues, but it will install correctly === dednick is now known as dednick|lunch [12:29] alecu: there are sdk bugs filed for mseveral of those [12:30] great, thanks! [12:30] * jdstrand ran through those last week === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun [12:32] asac: I'll be slowly switching to manual mode now [12:32] sil2100: so, just to be clear, did libunity-webapps get released, or is it now frozen and I should upload a no-source-change rebuild independently of daily release? [12:32] cjwatson: it will be out in a moment, the queue was piled up and it's all prepared now for take-off [12:32] ok, brilliant, less work for me [12:33] asac: I'm switching to manual mode slowly [12:33] jdstrand, yes - test-building now [12:33] jdstrand, sure adding him [12:34] dholbach: thanks! :) === _salem is now known as salem_ [12:35] jdstrand, on it [12:35] jdstrand, I can't wait for the ubuntu bits to be split out of the qtcreator build :) [12:36] I had the build and a hangout running at the same time and cpu temperature went up to 96°C [12:36] :) [12:37] dholbach: heh, I know that situation very well! :) [12:39] * ogra_ blames the hangout plugin [12:39] cjwatson: thinking of something like http://paste.ubuntu.com/6036460/ + http://paste.ubuntu.com/6036475/ [12:39] sil2100: cool. [12:39] ogra_, I expect to be able to upload in 30-45m [12:39] err [12:39] cjwatson: but would want stgraber to +1 it [12:39] jdstrand, ^ [12:40] * ogra_ pokes alsa-lib with a long stick ... [12:40] lool: Yeah, something like that [12:40] come on, move out of proposed [12:41] lool: maybe make some arrangements to avoid unnecessarily postfixing none === alan_g|lunch is now known as alan_g [12:42] cjwatson: you mean allow for 5th column to be empty? this might be uglier if we ever need a 6th one [12:42] albeit I see no reason we would [12:42] lool: no, I meant special-case "none" in the script [12:42] so none => bind, anything-else => bind,anything-else [12:42] just a nicety though [12:44] oh, right, it felt uglier in the code, but I can do that so that we have a nicer fstab [12:48] lool: I'm happy with the change, though I don't see how that'll fix /var/lib/sudo since it's not even in the writable-paths you pasted [12:48] ogra_, alsa-lib has migrated to release now [12:49] diwic, lol, i pinged #ubuntu-release this minute abouot it [12:50] stgraber: we'd be adding it, but it needs a special mode [12:50] * ogra_ was waiting for the confirmation mail ... forgetting that i only signed it but it has your mail address in the changelog [12:50] stgraber, cjwatson: LP #1217866 BTW; thought I had to wait some more for stgraber to review [12:50] Launchpad bug 1217866 in lxc-android-config (Ubuntu) "Allow specifying arbitrary mount options in writable-paths" [Undecided,New] https://launchpad.net/bugs/1217866 [12:50] stgraber: ok with second patch? [12:53] lool: yep, looks good. But please also add /var/lib/sudo (with the custom mode) in the same change so we don't need a second upload for that [12:53] asac, image build running [12:53] nice [12:53] lool,stgraber: I already uploaded for /var/lib/sudo - you'll need to rebase [12:53] lool: I can take care of the uploads (since we need 4 uploads to get that done with proper timing, ...) [12:53] stgraber: happy if you do :-) === chihchun is now known as chihchun_afk [12:54] stgraber: don't think this is urgent now that cjwatson uploaded it as persistent path [12:54] stgraber: so if we want to make it back to a temporary path, would we have to add some upgrade logic? [12:55] yep, we'd need something to clean it up from /userdata [12:55] but the boot time hooks which I still need to implement will cover that [12:56] morning [12:56] * rsalveti reads backlog [12:56] stgraber: ok [12:56] rsalveti: morning [12:56] new to this... installed copy of ubuntu . but it looks like its all synced to someone else . how do I perform a wipe [12:57] Nexus 4 on sale $100 off FYI [12:57] https://play.google.com/store/devices/details?id=nexus_4_8gb [12:59] ?any ideas [13:00] gfjbbcgh, probably just dummy data [13:00] hello [13:00] can it be wiped from the os ? [13:00] you have like the video's and stuff right? [13:00] gfjbbcgh, remove the demo-asset- packages with apt [13:01] I've read that all applications currently available in ubuntu are not available in ubuntu-touch. Is that true ? === hikiko is now known as hikiko|lunch [13:01] serge1, they are all available ... wether they are useful or can run without X is different :) [13:01] dholbach: \o/ [13:02] without X, so what about libreoffice for example ? [13:02] ogra_: what about gimp ? [13:02] why in the world would you run either of them on a phone ? [13:03] can the apk removal be ran from terminal from the phone itself [13:03] ogra_: if you want to use your tablet as a pc... It's useful [13:03] (they can run, but not display on the phone ... you would need VNC or ssh -X to exec them remotely) [13:03] convergence ogra_ CONVERGENCE! [13:04] popey, not on 13.10 :P [13:04] well yes [13:04] ☻ [13:04] lool: well, if you're quick, there won't actually be any real images with the persistent path [13:04] ogra_: No matter if it is not possible for 13.10 but is it planned ? [13:04] asac: ogra_: the new gst packages will not solve the hardware decode issue [13:04] we'll need to get green with software decoding [13:04] which is fine [13:04] oki ... sonds good [13:05] rsalveti: do we need to do anything for that? [13:05] like uploading a small fix? [13:05] ogra_: so could we have a vnc client on ubuntu touch [13:05] is this being ported to the sch-r760 version of USCC? [13:05] ogra_: run vnc server and vnc client locally [13:06] stgraber: ^ [13:06] for the hardware decode support, we need the new gst 1.1 packages, changes in gst to make it use libstagefright, and then change qtmultimedia [13:06] not happening this week [13:06] only pending issue from our team at this point is getting maguro to work properly with pulse [13:07] serge1, no, vnc client needs X [13:09] cjwatson: it takes about 2 hours to land an initrd change so I doubt we'll make it before any image is published, though I can certainly choose not to publish any image with that path as persistent to system-image [13:09] cjwatson: (we need to get initramfs-tools-ubuntu-touch uploaded, built and moved out of proposed, then same thing for ubuntu-touch-generic-initramfs and finally the same thing for android) [13:10] stgraber, wow, thats highly optimistic [13:10] (2h) [13:10] diwic: so, the kernel exception stack is happening every time I start pulseaudio here [13:10] ogra_: well, I can rescore things and I can upload android in about 20s, that helps :) [13:10] diwic: sometimes it also reboots my phone if I start pulseaudio twice in a row [13:10] stgraber, i'm actually pondering to move the bootimg generation back into livecd-rootfs [13:10] diwic: I wonder if that's because it's trying to initialize the hdmi "card" [13:10] rsalveti, okay, that sounds promising w r t nailing it down [13:11] ogra_: Not sure why it's optimistic [13:11] diwic: is there a way for us to disable that card specifically? [13:11] the result wont differ but it would save us from having to upload android all the time [13:11] rsalveti, we can certainly disable it, but I want to know if that actually fixes things [13:11] cjwatson, usually it takes around 45min for each of the packages... plus a 500M downloacd for a no change rebuild of android [13:12] (plus the android build and promotion indeed) [13:12] diwic: yeah, just wanted a quick way to disable that to see if the exception stack will happen [13:13] diwic: so we can at least know it's because of the hdmi card [13:13] I know that one is not used by android [13:13] cant you disable it in the kernel ? [13:15] whatever is the easier way to do it :-) [13:15] rsalveti, the quickest way is to do the pulse stuff I suggested [13:15] rsalveti, or strace possibly [13:15] right, let me get some more logs for it [13:16] rsalveti, the stacktrace looks pretty much like the one i see when the device reboots on boot [13:16] whgich happens before even the rootfs is mounted [13:17] ogra_: that happens when pulse is started [13:17] just link /usr/bin/pulseaudio to /bin/true and the stack is gone [13:17] rsalveti, for me it happens if the device is initialized [13:17] and then if you start pulseaudio manually, the kernel trace shows again [13:17] while in initrd [13:22] jdstrand, are http://paste.ubuntu.com/6036584/ the changes you were expecting? [13:25] jdstrand, uploaded [13:26] dholbach: hmm, I don't see the manifest shanges there. let me look at lp [13:26] Mirv, ^ [13:27] I uploaded the change and it looked to me it like it was just the mimetype and depends [13:27] I'm happy to do another upload later on [13:27] but that's just what was there [13:27] the diff isn't available yet [13:27] lool: so, for me, the click package does show up in click list after I install it from the app lens [13:27] jdstrand, I can't seem to run the lint tool with the latest rev: https://pastebin.canonical.com/96510/ [13:27] barry, stgraber: could we use daily-proposed for testing new client upgrade code? I'd install against daily-proposed -1 + upgrade? === dednick|lunch is now known as dednick [13:28] dholbach: what I was looking for was qtcreator - 2.7.1-ubuntu7~saucy1~test2 from https://launchpad.net/~ubuntu-sdk-team/+archive/staging/+packages [13:28] cjwatson: I should probably wipe my device to try this [13:28] jdstrand, s/absname/abspath, I'll file the merge proposal [13:28] beuno: whoops :) [13:29] cjwatson: does it show on the app panel of the home screen too? [13:29] lool: I can't open the app after installation, though, and it doesn't seem to remember that it's installed [13:29] jdstrand, or should I just push to it? [13:29] Mirv, is the change jdstrand mentioned ^ scheduled to go into saucy too? [13:29] dholbach: there was something new in lp:qtcreator-plugin-ubuntu , maybe that? "The Cordova porject type in the template definition file had a typo. The fileName was wrongly filename. So the .desktop file was not created." [13:29] lool: not sure about the server side, but on the client side, you'd need to edit /etc/system-image/channel.ini until i get the feature in to specify the channel on the cli [13:29] cjwatson: right [13:29] lool: only under "More suggestions" [13:29] Mirv, I uploaded that last night [13:29] beuno: what revision are you on? [13:29] anyway, I need coffee etc. before UDS [13:29] cjwatson: do you get the manifest error when trying to open it? [13:29] jdstrand, was it 1215133? [13:29] dholbach: yes [13:29] cjwatson: see you there [13:29] jdstrand, 36 [13:29] lool: no error, it just does nothing [13:29] Mirv, bug 1215133 [13:29] bug 1215133 in qtcreator (Ubuntu Saucy) "click manifest name should not contain underscores" [High,Fix committed] https://launchpad.net/bugs/1215133 [13:30] cjwatson: just as an update: https://launchpad.net/ubuntu/+source/libunity-webapps [13:30] cjwatson: sorry it took so long, I had issues with the architecture without documentation, had to hack my way through it [13:30] dholbach: that was in the previous upload [13:30] dholbach: https://lists.ubuntu.com/archives/saucy-changes/2013-August/008837.html [13:30] Mirv, no - that was a partial fix [13:30] dholbach: right, it lacks proper "LP:#" [13:30] Mirv: but ubuntu7 only had the partial fix [13:30] Mirv, I changed the changelog entry to say that [13:31] "Some initial work on #1215133: Replace "_" with "-" in the name of the [13:31] sil2100: thanks! [13:31] security manifest file" vs "Replace "_" with "-" in the name of the security manifest file (LP: #1215133)" [13:31] Launchpad bug 1215133 in qtcreator (Ubuntu Saucy) "click manifest name should not contain underscores" [High,Fix committed] https://launchpad.net/bugs/1215133 [13:31] Ubuntu SDK on Ubuntu 13.04 | http://askubuntu.com/q/338387 [13:32] beuno: did you see the request to reupload hello world without _ in the name? [13:32] lool: yes, he has just uploaded a new version, and asked me to test it. [13:32] lool, I have and I did. Want to test now? [13:32] beuno: I'm reflashing a clean system and will test this first thing :) [13:32] dholbach: aha, another upload needed then, ok [13:33] Mirv: let me double check that [13:33] Mirv, let me know when the kubuntu-packagers branch is updated and I'll take a look [13:33] jdstrand, with the fix, it tells me something's wrong, but I can't quite tell what from the output: https://pastebin.canonical.com/96511/ [13:33] lool: with a clean install it still breaks, because of the click bug on r/o filesystems [13:33] alecu: which bug is that? [13:33] Mirv: right, ubuntu7 is confirmed incomplete. [13:34] maybe I should set the bug back to "in progress" [13:34] cjwatson: https://bugs.launchpad.net/click/+bug/1217472 [13:34] Launchpad bug 1217472 in click "pkcon failure on the R/O system image" [Undecided,New] [13:34] dholbach: makes sense [13:34] done [13:34] oh. can y'all please just file bugs on the Ubuntu click package rather than on the upstream project, so that I don't lose them? [13:34] ok, taking the dog for a walk real quick before sessions kick off [13:35] alecu: I think that's just the unwritable /var/lib/PackageKit thing, though, fixed in lxc-android-config recently [13:35] jhodapp, diwic: With the GStreamer wrapping of Android side codecs provided by DSP, do we get accelerated MP3 decoding? [13:36] cjwatson: great. It's not on today's RO image, it seems. [13:36] beuno: ack. at the end of the script it says 'Warnings found' as opposed to 'Errors found'. so look up in the warn dictionaries and you'll see: [13:36] "warn": { [13:36] "desktop_file_hello_world.desktop": "TODO" [13:36] } [13:36] jhodapp, diwic: Or do we have to add a gstreamer plugin for MP3 support? [13:36] beuno: so, it is clean [13:36] alecu: it's not yet, no [13:36] lool, we cant [13:36] jdstrand, ah :) thanks [13:36] beuno: but we don't have tests for validating the desktop file yet [13:36] * beuno nods [13:36] jdstrand, want me to push this tiny fix to the script? [13:36] alecu: it won't make it into the image until pending is promoted to current [13:37] beuno: I actually did already [13:37] sergiusens: can I use --pending with the RO image somehow? [13:37] * beuno pulls [13:37] you all are heroes! [13:37] jdstrand, thanks [13:37] lool: barry stgraber I asked this a lot already, but do we have the 'pending'/'devel' channel ready? [13:37] np [13:37] dholbach: the super kind? or just regular? [13:37] the super kind :) [13:38] :-) [13:38] lool, no accelerated MP3 to my knowledge, should be software decoding even on the android side [13:38] dholbach: do you want the quick reproducer on the manifest verification? [13:38] bfiller: is renato on vacation? [13:38] popey: yes [13:39] jdstrand, that'd be great [13:39] sergiusens: there's a daily-proposed channel, but it's not something you'd be able to track every day [13:39] bfiller: when's he back? [13:39] dholbach: a) open qtcreator and create a new project of type Simple Touch UI [13:39] sergiusens: or actually you might, but you'd at least lack deltas [13:39] popey: monday [13:39] great, thanks [13:39] dholbach: b) give it a name with an underscoe. eg, test_underscore1 [13:39] sergiusens: testing it right now and will report back how it works :-) [13:40] dholbach: c) after clicking through the wizard, click 'Packaging' on the left [13:40] bfiller: do we have anyone (other than renato) who can profile qml apps to identify performance issues? [13:41] dholbach: d) examine the manifest tab. if any "name" and the key to the hooks dictionary have an underscore, the bug is not fixed [13:41] s/any// [13:42] dholbach: that's it [13:42] popey: not at the moment, we can discuss more. renato is totally booked as well [13:42] dholbach: pushed qtcreator_2.7 [13:43] asac, image up [13:43] bfiller: ok. clock app has some performance issues which needs someone to take a look. it does a lot of animation which differs from most of our other apps [13:43] popey: can someone from sdk help take a look? [13:44] katie, poke [13:44] bfiller: ok, will do [13:44] mterry, hello [13:44] mterry, i was watching for you to come online === chihchun_afk is now known as chihchun [13:45] katie, :) I have a few questions! Let me see... I wrote them down [13:45] popey, is that the reason why it is exactly 9min behind on all my devices ? [13:45] katie, do we have information on what indicators' "greeter mode" is like? [13:46] ogra_: not sure about that. I'm more thinking of the animation of the hands and numbers [13:47] mterry, i'll have to get back to you on that on [13:48] one [13:48] ogra_: cross your fingers man [13:48] asac, nah, then i cant roll ciggarettes [13:49] i know it is good ... no need to pray [13:49] ;) [13:49] ogra_: the ogra gut cannary at work? [13:49] would love if that would be a bullet proof method [13:50] katie, OK. I also had some questions about the emergency dialer in the greeter [13:50] it totally is :P [13:50] mterry, i can probably answer those :) [13:51] katie, so the only way to get into the emergency dialer is the button on the lockscreens, right? Like, we don't insert a phone icon on the launcher? === dandrader is now known as dandrader|afk [13:51] mterry, correct [13:51] Mirv, did you see jdstrand's instructions to reproduce above? [13:52] Mirv, jdstrand: I'm going to be busy with sessions now - I guess I can start a build in the background but might be a bit too tied up to play around with QtC for a while :/ [13:52] katie, OK. When it's open, do we show it on the launcher then? (if user swipes launcher over) [13:52] mterry, if you launch the phone icon in the launcher, it should take you to the normal telephony app [13:52] katie, OK [13:52] mterry, which if the device is locked, will take you to the pin code entry [13:52] *sorry, meantpasscode [13:52] katie, :) [13:53] dholbach: I'm busy with Qt 5.1 right now, Zoltan would know more regarding the fix I pulled from his commits to that latest commit [13:53] mterry, gotta be carefull with my pins and passcodes! [13:54] katie, but when the emergency dialer is open, do we show it on the launcher then? (if user swipes launcher over during use) [13:54] bzoltan, can you verify that with the latest commits http://paste.ubuntu.com/6036681/ works now? (re: 1215133) [13:55] ogra_: diwic: with current ucm files I can hear the sound, but it's still quite low I'd say [13:55] lool: should I add migration code for .developer_mode to .writable_image while I'm at it? [13:55] rsalveti, depends how you play it it seems ... [13:55] sergiusens, ev, lool: I won't be around for the error click session, I have to start another community session - sorry === hikiko|lunch is now known as hikiko [13:55] with alsamixer handsfree is in '21' [13:55] ogra_: with aplay [13:55] ricmm, tedg: Hey, I can't launch dropping letters from the unity home screen, but if I launch terminal and cd /opt/click.u.c/*dropping*/current and run qmlscene dropping-letters.qml, it works; could you help me debug this? [13:55] rsalveti, the ringtone is loud and clear if i use system-settings [13:55] stgraber: if you dont mind, that'd be nice [13:55] lool: I don't believe we've promoted an image since, so I could do that and save us the trouble of having people do it by hand [13:55] ogra_: aplay /usr/share/sounds/alsa/Front_Left.wav [13:55] stgraber: especially since nobody got an upated image in the mean time [13:56] rsalveti, and its quiet and faint from the phone app [13:56] stgraber: right [13:56] dholbach: no worries [13:56] lool: I did mind having to upload all the stack of packages just for that, but since I've got to do it anyway, might as well :) [13:56] ogra_: not for me [13:56] weird [13:56] ogra_: must be the crash, let me get some logs [13:56] yep [13:56] I know yesterday I was able to get the proper volume sometimes [13:56] right after boot, but wasn't happening all the time [13:56] stgraber: I'd vote for supporting .developer_mode for a month on top of moving it; we should update docs to use the new path [13:57] lool: I cant right now, smack in the middle of something else [13:57] but I'll ping you when im done [13:57] ricmm: is someone else able to help with this? [13:57] lool: I wonder if the qtubuntu changes to support Path in .desktop files ever landed [13:57] cjwatson: hmmm that could be it [13:58] bug 1204596 [13:58] bug 1204596 in unity-mir "Unity 8 does not honor Path= in desktop files" [High,Fix committed] https://launchpad.net/bugs/1204596 [13:58] says fix-released in the distro tasks though [13:58] mterry, just talked it over with JohnLea .. and the answer is no, Idon't think the emergency caller should be on the launcher [13:58] oh, but the unity-mir one was only 12 hours ago. maybe that? [13:58] (I find it super annoying to leave terminal; I can never hide the keyboard, so can't reach the home screen button on the left bar whic is displayed under the keyboard; is there a workaround for this?) [13:58] greyback: do you know about this one^ ? that lool mentioned [13:59] greyback: context is that click apps don't seem to launch anymore, unless you run qmlscene by hand [13:59] katie, yay, that makes it simpler :) [13:59] * asac is getting 28 [13:59] * asac hopes [13:59] asac: you're much older I'm afraid [13:59] mterry, great! [14:00] lool: which image, I can debug [14:00] ? [14:00] lool, thats why he hopes :) [14:00] katie, what about transition? Do we want the same transition apps get when appearing? Or like a fade in or something? Also, is it dismissed by turning of screen or is there a button to dismiss? [14:00] * lool need to run to vUDS [14:00] ogra_: so you still believe its a good image? :) [14:01] lool: hmm, that's strange. Usual phablet image I suppose (i.e. not using Mir)? [14:01] mterry, yes, same transition, and it is dismissed by turning off the screen [14:01] ogra_: i am installing on maguro :/ [14:01] mterry, later we should add a button to dismiss too [14:01] hmm /me is stupidly running the image from 22 [14:01] which is not really helpful [14:01] lool: lol [14:01] well i am not su sure, but i am surely much wiser :) [14:01] greyback: sorry need to update, had forgotten I couldn't set to --pending with system image [14:01] beuno: guess what? now the app manifest seems right, but the package name in the click index is different to the one in the manifest: "ar.com.beuno.hello-world" in the click index, "com.ubuntu.developer.beuno.hello-world" in the manifest. [14:02] asac, yes, and ? [14:03] hearing any issues ? [14:03] alecu, GAH! [14:03] lool: ok, please let me know if it is still broken [14:03] alecu, I'll re-upload after this session [14:03] ogra_: its pushing :) not yet booting [14:04] i hope i hear something yes :) [14:04] barry: ah stupid me, I thought I could test with new channel.ini, but it isn't supported in /current image [14:04] * lool goes back to flashing .27.2 [14:05] lool, cjwatson: so defaults,mode=0700 should be fine for sudo, right? [14:05] lool: with older system-image you can probably edit /etc/system-image/client.ini [14:05] lool, Catching up, is there something in ~/.cache/upstart/*log ? [14:05] stgraber: seems correct to me [14:05] Kaleo: are you available for the Qt 5.1 session? https://plus.google.com/hangouts/_/368836c93bc010c273ce53686164a50dd2e47904?authuser=0&hl=en [14:05] oh wow, why is /var/lib/sudo root:lool on my system [14:06] janv. 1 1985 [14:06] it even speaks french ! [14:06] very realistic [14:06] * lool blames btrfs === dandrader|afk is now known as dandrader [14:07] barry: ok, trying that now [14:07] barry: well, it's still relatively useless if I don't updte system-image before hand anyway [14:07] ;) [14:08] lool: both packages uploaded, I'll do the two others as things publish in the archive and will do an image rebuild once it's all done [14:08] stgraber: great, thanks [14:09] stgraber: Yeah [14:12] ogra_: diwic: i dont hear anything in settings -> ringtone with latest [14:13] but i never could get it to work with the ucm files as well, so i assume i might have a bad setting? [14:13] what files do i need to delete? [14:13] none, it shoudl just work [14:13] my rsync is still running ... i'll test in ~30min [14:14] alecu, re-uploaded 0.6 [14:15] ogra_: i can tell that it doesnt play any sound [14:15] let me see what i have in /home/phablet/.* [14:16] not sure what to remove [14:16] let me try wiping everything [14:18] super. pulseaudio only takes like 5-7% cpu and previously mediaserver used to consume 25-30% on mako while playing a normal mp3 #success [14:19] katie, I lost connectivity there. "what about transition? Do we want the same transition apps get when appearing? Or like a fade in or something? Also, is it dismissed by turning of screen or is there a button to dismiss?" [14:19] sergiusens: I wanted to try bootstraping from daily-proposed and did this: https://code.launchpad.net/~lool/phablet-tools/different-channels/+merge/182654 [14:19] diwic, hey! [14:19] sergiusens: BTW the phablet-tools code is much nicer to read since a couple of months ago; feels much better structured [14:19] sergiusens: I can tell you spent quite some time improving it :) [14:20] om26er, asac I'm on a UDS session [14:20] diwic, ah, ok. will ping you later [14:20] diwic: ok ... whenever its good for you to answer :) [14:21] mterry, yes, same transition [14:21] mterry, and yes, it is dismissed by turning off screen [14:21] katie, OK, thanks! [14:21] mterry, later there may be a button to dismiss, but not for now [14:21] mterry, np :) [14:22] asac, is that mako or maguro ? (works for me on mako) [14:22] ls === schwuk is now known as schwuk_away [14:24] lool: yeah, I broke some backwards compat someone was using.. thanks [14:24] lool: so I want to improve your MR, barry told me there's a channels.json we can use [14:25] sergiusens: there is indeed [14:25] lool: I can get this in if urgent and work on that on top [14:25] sergiusens: it's not critically urgent, but I'd rather we land this as to enable the QA team to use it [14:25] sergiusens: we can improve it to use channels.json afterwards; seems like a good idea indeed [14:26] sergiusens: http://system-image.ubuntu.com/channels.json which would avoid hardcoding some of the internal structure of system-image.u.c in phablet-tools too [14:29] lool: ack, I'll give this a test and happrove [14:31] ogra_: diwic: pulse logs with maguro, using syslog: http://paste.ubuntu.com/6036785/ not sure if it's useful [14:31] because it seems something is affecting the omap interrupt handler === Namidairo is now known as Namidairo`bnc [14:34] om26er: maguro [14:34] rsalveti: can you try maguro build? [14:34] for me sound doesnt work [14:34] but it never worked (also with UCM fixes), so it might be me [14:34] asac: that's the one I'm testing [14:35] rsalveti: so you see the same? [14:35] sound works sometimes with latest image, but there are still issues [14:35] oh it works [14:35] which is what we're currently debugging [14:35] rsalveti: but its ok for dogfooding? [14:35] or just super flaki compared to before? [14:36] not ok for dogfooding it seems [14:36] not sure i wanted to hear that answer [14:36] rsalveti: mayube debug for a bit and then describe the impact :) [14:36] behaviour [14:36] that's what I'm trying to do :-) [14:36] I'll get back to you once I know more [14:38] barry: I've flashed 22 (-b reports 20130833) and upgraded system-image-* + ubuntu-system-settings, I don't get hangs when leaving / reentering settings, but after pointing at daily-proposed in client.ini it doesn't see the new image [14:38] lool: which device? [14:38] barry: grouper [14:38] sound works pretty ok ... just a bit low on volume [14:39] barry: this was after flashing with ./phablet-flash ubuntu-system --channel daily-proposed --revision -1 [14:39] barry: (with new branch submitted above which just points phablet-flash at daily-proposed index.json) === alan_g is now known as alan_g|tea [14:39] barry: could it be it's checking daily and not daily-proposed? [14:40] lool: what does `system-image-cli -c` say? [14:40] barry: system-image-cli --dry-run -v -v [14:40] gets me to a traceback [14:40] yay [14:41] barry: http://paste.ubuntu.com/6036810/ [14:41] hello [14:41] sergiusens: ah forgot to update tests [14:42] lool: that's interesting. i managed to unbrick my n7 so let me try to reproduce both problems locally [14:43] lool: no worries [14:43] asac, so the network indicator issue was our fault ... [14:43] * ogra_ just uploaded a fix [14:44] sergiusens: fixed === schwuk_away is now known as schwuk [14:45] Do I need to change something on the MP to trigger another test run? I forgot again [14:48] lool: nope, it will run eventually [14:48] barry: doing an apt-get dist-upgrade for now, will retest once there's a workaround/fix for using daily-proposed [14:48] sergiusens: thanks [14:50] sergiusens: latest runs I tested the new backup stuff, seemed to work fine; I don't remember why you list specific things to backup rather than backing up all of /userdata? [14:50] Mirv: sorry, missed it [14:50] lool: [systemimage] Aug 28 14:40:52 2013 (1923) no matching channel/device: daily-proposed/grouper [14:50] barry: yes, saw this in the log, but it exists, yes? [14:51] lool: seems to! [14:51] lool: all of userdata is too much [14:52] sergiusens: because of? demo contents? [14:52] lool: i.e.; if you move from a cdimage image to a system image you end up backing up all of ubuntu [14:53] jdstrand, bzoltan1, Mirv: local test-build running - can anyone check http://paste.ubuntu.com/6036681/ with the latest package Mirv built? [14:53] sergiusens: ah right; I guess we could fine tune handling when backing up system-image installs [14:53] sergiusens: seems /userdata/system-data and /userdata/user-data would be good targets in ths case === alan_g|tea is now known as alan_g [14:53] lool: ah, sure, I can look into adding that [14:54] sergiusens: no hurry, just thinking of the next transition (partitioning changes) and fuller backups on the long-ter,m [14:55] ogra_: what did we do? [14:56] asac, https://bugs.launchpad.net/ubuntu/+source/livecd-rootfs/+bug/1217596/comments/6 [14:56] Launchpad bug 1217596 in livecd-rootfs (Ubuntu) "[regression] empty network-indicator since 23.1 with clean flash" [Medium,In progress] [14:56] lool: well, I'll add it as a WI even though we had the session yesterday :-) [14:56] ogra_: see i somehow had this sense [14:56] when we did that [14:56] asac, peripherial hackery we missed when fixcing autopilot [14:57] right,, but remember how i wondered about what else might be using it :)? [14:57] well theoretically nothing should [14:57] practically someone wasnt up to date when creating all that hackish stuff in the first place :) [14:57] well, i wouldnt need to wonder if practice was close to theory :) [14:57] ogra_: well, we did a loosy stuff grepping and consulting folks when rolling the real fix... [14:58] in any case it will be fine the next build [14:58] i thought we grepped for admin [14:58] yeah, i definitely did [14:58] i guess thats why i prefer having all the script magic in one place rather than all maintainer scripts [14:58] well [14:58] sergiusens: thanks :) [14:58] that whole subdir in livecd-rootfs has to go [14:58] if all that stuff would be in one bzr tree ... we could be ... almost perfect :) [14:58] it cant [14:59] ogra_: so timeline for this fix? [14:59] different things happen in different places [14:59] asac, uploaded a while ago [14:59] ogra_: can you try call in and out for the GN as well? [14:59] asac, but its a build time fix [14:59] let me flash first ... the bugfix got in my way [14:59] ogra_: that means we have to create a new image you mean? [15:00] yeah cool. [15:00] happy we caught this as well now [15:00] right [15:00] that file gets mangled at build time [15:00] as a workaround for the still missing lightdm [15:00] ogra_: ok go test it as usual and let me know [15:00] jdstrand: hmm so I've pinged various folks about launching apps not working with latest packages on read-only image [15:00] togethe with this bugfix we now did :) [15:01] jdstrand: I had this with latest /current and also after updating to latest packages [15:01] jdstrand: launching uncontained works [15:01] jdstrand: but not under aa-exec [15:01] jdstrand: I suspect something regressed in terms of permissions [15:01] lool: possibly-- is this with mir? [15:01] jdstrand: I can't tell how unity starts stuff, but launching qmlscene Stock_Ticker.qml works, while launching aa-exec -p com.ubuntu.stock-ticker-mobile_stock-ticker-mobile_0.3.7ubuntu1 -- qmlscene Stock_Ticker.qml doesn't [15:01] jdstrand: I think so? [15:02] lool: how are you launching with aa-exec, via the terminal app? [15:02] I mean, I have libmirserver1:armhf and stuff [15:02] jdstrand: yes [15:02] but I think term app is unconfined [15:02] since it's not clickified? [15:02] lool: can you paste: grep DEN /var/log/kern.log ; grep DEN /var/log/syslog ? [15:03] lool: the term app doesn't matter, I was trying to rule out another bug [15:03] Kaleo: https://blueprints.launchpad.net/ubuntu/+spec/client-1308-qt51 whiteboard updated, the DPR/font issue assigned to you (as soon as work items can be put to work items field) [15:04] jdstrand: kern.log: http://paste.ubuntu.com/6036881/ [15:04] jdstrand: syslog http://paste.ubuntu.com/6036882/ [15:05] lool: ok, I can fix that, but heading into a meeting [15:05] jdstrand: thanks; I think this breaks running all clicks (preinstalled and downloaded) albeit other bugs might get in the way too [15:06] might I mention that having vUDS during FF week is rather difficult? [15:06] eh [15:06] +1 [15:06] lool: yes, I will fix it after this session [15:06] whoever decided on vUDS this week... [15:08] jdstrand: can confirm that if I strip aa-exec from .desktop files, I can run preinstalled clicks again [15:08] now trying installation [15:08] lool: hahahaha. the dash in "daily-proposed" is what's breaking things. :( i'll have a fix (and better error reporting) shortly. [15:08] lool: yes. I know the problem [15:08] we need to fix those denials [15:09] w00t, downloaded app also works [15:09] dholbach: ^ [15:10] ricmm, tedg, greyback: So app launching issue was just apparmor regression [15:10] ogra_, diwic, 0828 image, fresh install, apps with sound now don't freeze, but no sound [15:10] (caused by unrelated uploads) [15:10] lool, Ah, okay [15:10] lool: ok, thanks for the update [15:11] well, I wouldn't say it is a regression-- nothing changed with apparmor, the system around apparmor changed [15:11] lool, I'll upload qtcreator 0ubuntu9 once it passed a local build and test [15:11] but that is semantics. apparmor-easyprof-ubuntu needs to be updated [15:11] alecu, dholbach, jono_: With latest packages I can install and run clicks from appstore and run preinstalled clicks iif I disable the apparmor stuff, Jamie will fix this last remaining bug soon [15:11] jdstrand: sorry, tried to clarify with "unrelated uploads" [15:11] mamenyaka, sounds good, that means the sound stack works but most likely your volume settings are wrong [15:12] dholbach, and then we launch the store? :) [15:12] jdstrand: I'm just super happy that we didn't face other click / click-scope / packagekit issue with r/o images in the appstore scenario [15:12] ogra_, when I tap on play in music: msm_pm_wait_cpu_shutdown(): Timed out waiting for CPU 1 SPM to enter sleep stateCPU1: unable to kill === chihchun is now known as chihchun_afk [15:12] jono_, and with the next qtcreator upload (currently test-building), the submitted apps should stand a better chance of getting accepted [15:12] beuno, I'd have to defer to mhall119 and dpm for that one :/ [15:13] dholbach, mhall119, dpm, launch! launch! launch! [15:13] lool: the problem is that /android/... seems to be a new path for libs [15:14] lool: indeed, it's great that no other big thing is needed for r/o images. I'd like to try it in tomorrow's image, where all of this fixes have landed. [15:14] its interesting that dropping letters does not work, but hello-world does. but regardless, the fix is easy [15:14] ogra_, volume is at max, unmuted in the indicator [15:14] jdstrand: could this be due to surface flinger? there might also be device specific stuff since IIUC Mir has per-GPU support [15:14] alecu, could you verify 0.6 works well? [15:14] mamenyaka, i'm talking about your UCM files [15:14] (btw, I keep getting it wrong because the SDK resets the fields) [15:15] beuno, alecu: So I do face a bug with installing click apps; not sure where it belongs [15:15] lool: I've tested with mir and surfaceflinger, and didn't see that [15:15] ogra_, that was my other guess :) [15:15] beuno, alecu: I can intsall beuno's hello world, I fixup the .desktop to disable aa manually, then hit "open" [15:15] beuno, alecu: But then screen flashes for a sec and I land back on the XDA app [15:15] beuno, alecu: Going back to the home screen, I see both are running, and I can go to the hello world app and it works [15:16] so it seems like an unity launch thing; perhaps not waiting enough for the app to render and considering it dead? [15:16] jdstrand: this is on N7 BTW [15:16] cool [15:16] jdstrand: might relate to the new images? mounts might be slightly different [15:16] that's what I figure [15:17] beuno: (this was with 0.6) [15:17] ogra_, I have this in speaker, enablesequence: cset "name='RX3 Digital Volume' 68" [15:17] cset "name='LINEOUT1 Volume' 100" [15:17] cset "name='LINEOUT3 Volume' 100" [15:17] but why xda doesn't also fail is a slight mystery [15:17] lool, perfect, thanks. I have a few more bugs to file against the SDK :) [15:17] jdstrand: ? I also patched the xda .desktop file [15:17] mamenyaka, probably diwic can find some time to debug with you ... or tell you waht logs and info to attach to a bug [15:17] oh, I missed that. I thought you meant that they worked without modification [15:18] jdstrand: basically all preinstalled or downloaded apps failed, when I patch the .desktop file to drop aa-exec they always work, except for a display glitch when running an app just after downloading when one was already running [15:18] ogra_, okay, thank you; can you give me diwic ? [15:18] yep, got it [15:18] lool, dholbach sounds great, when will these fixes be in the images? [15:19] jono_, qtcreator is not on the images - the other fix I don't know [15:19] jono_: images aren't getting promoted anymore due to various regressions in basic features reqiured for dogfooding [15:19] lool, right [15:19] jono_: outside of this, I think jdstrand will upload quite soon -- when he has a free second basically [15:19] jono_, QtC build will likely take another hour here, plus some testing, then I'll upload, then some more time to build in LP and land [15:20] jono_: can't tell how long it takes to fix all regressions, but with jamie's fix and a pending image or apt-get updating to latest packages, we should get this working by end of day today I'd think [15:20] tvoss__: ricmm: did you guys have any time to look at the linking problems of the url-dispatcher thing in the platform-api? [15:20] lool, sounds great [15:20] thanks for the focus on this [15:21] dholbach, when will ratings and reviews land? [15:21] jono_, not sure, ralsina: ^ do you know? [15:22] lool, dholbach also, any more resolution on apps that need to build C++ and how they fit into click packages? [15:24] jono_, this is going to be discussed at UDS tomorrow (http://summit.ubuntu.com/uds-1308/meeting/21954/community-1308-qml-extensions/ and http://summit.ubuntu.com/uds-1308/meeting/21908/foundations-1308-arch-specific-click-pkgs/) [15:24] jono_: there's an app ^W^W a session for that! [15:24] dholbach, awesome [15:24] lool, :-) [15:24] jono_: tomorrow 16:05 [15:24] lool, I was talking to dholbach about this the other day - time is of the essence on this as some app showdown apps use C++ [15:24] jono_: Architecture-specific Click packages [15:25] jono_, I sent out a mail about this: https://lists.launchpad.net/ubuntu-appstore-developers/msg00518.html [15:25] diwic: ogra_: still giving the exception stack even with the hdmi card disabled in the kernel http://paste.ubuntu.com/6036955/ [15:25] seems it fails when starting the playback somehow [15:26] jono_: I doubt we have something ready in under a week, as it means getting cross stuff setup right, updating SDK rules and such; it's not a lot, but it's probably best to tell folks to switch back to rw images for this (as much as we hate this) [15:26] jono_: it would be good to see where they had to write C++ too [15:26] lool, take me as an example, I use c++ to talk to libupnp [15:26] rsalveti, odd, i still balme the kernel [15:27] *blame [15:27] ogra_: right, it's probably something in the kernel for sure, but pulseaudio is making that issue to show [15:27] lool: apparmor-easyprof-ubuntu version 1.0.20 uploaded [15:27] asac, so seeing that lool and stgraber just uploaded a lot of changes for the system image i think we should wait for that to land before i do another respin [15:28] (primarily that will require an android rebuild still) [15:28] w-flo: seems completely valid to me, and might or might not make it to builtin features of the platform some day, I doubt iOS or Android offer this in their base platforms for instance; I guess one can't write an arch-independent uPNP implementation in QML/JS [15:29] lool: I would say install that and rerun the hooks, but, well, you have an ro image :) [15:29] jono_: late september [15:29] I can see how we have some kind of weakness over java / dalvik where a lot of low-level things can be programmed in java [15:29] like UDP/TCP connections [15:29] lool, that's fine, I just bundle the .so with the click package. Though that makes it armhf dependent [15:29] jdstrand: I can mount -o remount,rw it ;-) [15:29] lool, w-flo: that is what I did with evilapp [15:29] jdstrand, so is that considered evil? :) [15:30] w-flo: Yes; what I think is interesting to fix is when people are using C++ to e.g. set an environment variable for their app or something like that [15:30] (to tweak Qt's behavior) [15:30] I've head this is pretty common [15:30] it tests confinement. as of today, it doesn't actually do anything evil. it just tries to see if it can do evil [15:30] there might be other cases where we could provide support mechanisms to avoid the need of C++ [15:31] oh, i see. yeah, upnp support in the sdk is probably overkill [15:31] lool: we can wait for the session, but I think we need to allow for C++ [15:31] jdstrand, well, we are on a phone now ... it could do realyl evil stuff ... like calling your mom :) [15:31] or texting her that you drop by tomorrow [15:31] ogra_, # /system/bin/logcat Unable to open log device '/dev/alog/main': No such file or directory [15:31] yes, it could-- it just happens to only check if it can do stuff like that, instead of actually doing it :) [15:31] Well I do hope c++ will be allowed, otherwise I'm doomed :D [15:32] mamenyaka, sounds like a udev problem [15:32] second boot [15:32] jdstrand: my goal never was to prevent C++; I mainly disagree with promotion of .debs [15:32] * jdstrand nods [15:32] I think app authors should build their deps as part of their app project and bundle the .so [15:32] me too [15:32] yep [15:32] and we should enable them to cross-build for all target arches rather than building on device [15:33] I actually build a deb file using pbuilder-dist, then extract the .so file and put that into the click packaging dir.. seems to work fine :) [15:33] lool: also yep :) [15:34] w-flo: it does, it just is not easy enough [15:34] * w-flo looks forward to easier builds [15:37] ogra_, resolved, weird things happen [15:37] hah [15:39] jdstrand: how do you rerun the hooks? and who's supposed to do this? [15:39] jdstrand: seems we will want to trigger this during session startup somehow [15:40] lool: correct. stgraber is creating a way for us to hook into something on first boot after upgrades [15:40] ok [15:40] there is a work item for that [15:42] lool: hold on [15:43] jdstrand: I ran ls /var/lib/apparmor/clicks/*.json | xargs -n 1 /usr/bin/aa-clickhook [15:44] didn't work though [15:45] I haven't been paying attention most of today, but is this known? (from apt-get update && apt-get dist-upgrade) click : Depends: python3-click (< 0.3.1.1) but 0.3.2 is installed [15:46] ogra_, I have this in dmesg, this is good, right?: ALSA device list: #0: Midas_WM1811 [15:46] y okay, because my other device hasn't got this [15:46] bigcalm: I have python3-click and click at 0.3.2 on my armhf touch device [15:46] lool: This is a Galaxy Nexus [15:47] lool: nothing special installed, just apt-get update and dist-upgrade ended up giving me that [15:47] bigcalm: could just be a symptom of a partial upgrade? dpkg -C to check, or perhaps apt-get -f install? [15:48] bigcalm: you can try apt-get install click python3-click to see what is the next issue apt would face [15:48] I can force it (as suggested by apt), but don't like to do so if it's a known problem [15:48] Will do [15:49] lool: http://paste.ubuntu.com/6037049/ doesn't look happy :( [15:50] cjwatson: ^ [15:50] albeit that seems like a desktophook issue [15:51] Yay for breaking things. Maybe I could take davmor2's job [15:51] bigcalm: I had an issue with desktophook recently myself (https://bugs.launchpad.net/ubuntu/+source/upstart-app-launch/+bug/1215478) [15:51] Launchpad bug 1215478 in upstart-app-launch (Ubuntu) "desktophook fails when ~/.local/share/applications doesn't exist" [High,New] [15:51] but might have been broken manifest [15:53] lool: your bug wasn't a desktophook issue, as I noted in my reassignment comment [15:53] it was a bug in Ted's thing, which is different [15:53] I'll have a look at bigcalm's traceback once I get a break between sessions [15:53] bigcalm: I'm guessing /home/phablet/.local/share/applications/ubuntu-ui-toolkit-gallery.desktop is a dangling symlink? [15:53] bigcalm: have you been removing click apps by hand? [15:54] lool: ok, got asked a question in a session [15:54] coX^>|VVXSS+[BcP$W XCb[B[APNC [15:54] =c/-cv-VHHHHHHHHHHHHHHHH_+_^^ [15:54] cjohnston: yes it's a dead symlink. No I haven't been removing apps [15:54] lool: so, you need to run the system hooks and the user hooks [15:54] (I'm cjwatson not cjohnston) [15:54] Sorry :) [15:54] bigcalm: what is the target of the symlink? [15:54] Happens to me all the time [15:54] lool: so: sudo click hook remove apparmor && sudo click hook install apparmor [15:55] lrwxrwxrwx 1 phablet phablet 81 Aug 26 16:26 /home/phablet/.local/share/applications/ubuntu-ui-toolkit-gallery.desktop -> /home/phablet/dev_tmp/ubuntu-ui-toolkit-gallery/ubuntu-ui-toolkit-gallery.desktop [15:55] oh, so that's a symlink not created by click [15:55] lool: click hook remove click-desktop && click hook install click-desktop [15:55] presumably by hand [15:55] Ah, that was when I tried to run the gallery on the phone - which it didn't do [15:55] lool: click hook remove upstart-app-launch-desktop && click hook install upstart-app-launch-desktop [15:55] cjwatson, i'd put my bets on the SDK [15:56] lool: those three commands should reset your desktop files and update your apparmor policy [15:56] bigcalm: ok, so I'll just make it ignore dangling symlinks [15:56] cjwatson: not by my hand, by qtcreator [15:58] jdstrand: works! [15:58] \o/ [15:59] lool: and 'sudo aa-status' shows they are running confined? [15:59] * jdstrand is sure it does, but it is always nice to see :) [15:59] 16 profiles are in enforce mode. [16:00] ar.com.beuno.hello-world_hello-world_0.6 [16:00] com.ubuntu.developer.mhall119.xda-developers-app_xda-developers_0.1.5 [16:00] com.ubuntu.dropping-letters_dropping-letters_0.1.2.2 [16:00] com.ubuntu.stock-ticker-mobile_stock-ticker-mobile_0.3.7ubuntu1 [16:00] so yes! [16:00] :) [16:00] jdstrand: cool, thanks for uploading the updated profile [16:00] np [16:00] * beuno feels trapped [16:00] jdstrand: is there some test that will prevent promotion of an image if this regresses? [16:00] lool: note, that aa-status will show you the pid that is running under the profile when the app is running [16:00] jdstrand: not being able to launch any app because e.g. platform changed, qt was updated or something, is pretty bad [16:00] ogra_, after upgrading, the other device shows too: ALSA device list: apq8064-tabla-snd-card [16:00] great [16:01] cjwatson: having removed the dead symlink, the packages have installed okay [16:01] bigcalm: fixed in click 0.3.4, hopefully making its way in soon [16:01] I guess integration tests trying to run apps should catch this [16:01] so the alsa driver initializes [16:01] lool: there is another discussion on that that came up wrt autopilot [16:01] but still no sound [16:01] lool: I'll CC you [16:01] jdstrand: I guess evilapp's autopilot tests would be the best way to test this [16:02] jdstrand: I mean from an image QA point of view -- not for all app authors (useful too) but just for us not to call /pending a good image unless this works [16:02] jdstrand: I haven't really been following the aa-exec discussion, but is it a problem that the generated code uses "aa-exec" without full path, aa-exec is in /usr/sbin/, and /usr/sbin apparently isn't on the phablet user's default path? [16:02] (which incidentally is a regression from the OneTruePath work we did way back in dapper or whatever it was ...) [16:02] cjwatson: the discussion lool and I are having now? [16:02] cjwatson: things work now, it seemed to be that libs appeared out of /android/something instead of some other path [16:03] yeah [16:03] hm, ok [16:03] right, what lool said. something changed in the image that moved paths around and apparmor-easyprof-ubuntu wasn't updated [16:03] lool: so, are autopilot tests part of image based tests? [16:04] jdstrand: yes, a bunch of them [16:04] which nexus device has apq8064-tabla-snd-card? [16:04] lool: right, so currently autopilot tests are using upsatrt-app-launch [16:04] jdstrand: http://reports.qa.ubuntu.com/smokeng/saucy/image/3805/ [16:04] which means that autopilot tests run under apparmor [16:05] which means that if something changed in the image like this, the tests would fail [16:05] jdstrand: i don't think they've been converted over to upstart yet? [16:05] doanac`: perhaps not yet, but this bug was filed last night: bug #1217700 [16:05] bug 1217700 in AppArmor "apparmor prevents autopilot introspection from working on phablet devices" [Undecided,New] https://launchpad.net/bugs/1217700 [16:05] correct. thomi is trying to convert them [16:05] jdstrand: the problem is that none of the tests we have covers a click package, so no test against an apparmor confined app [16:05] ok, that's fine [16:06] lool: to answer your queston, when ^ hits, then we have what you asked for [16:06] jdstrand: ok cool [16:07] mardy, mpt: if you want to join the settings's vUDS session: https://plus.google.com/hangouts/_/5423f0843ea0c49351f4b1f771400ebf2d0e7cc5?authuser=0 [16:07] ogra_: no please [16:07] not [16:07] before plz [16:07] the problem currently is that when autopilot is used, qmlscene wants to start its own dbus-daemon by using dbus-launch [16:07] lool, ^ (I think you subscribed to the blueprint) [16:07] ogra_ we dont have time all day long [16:07] charles, ^ [16:07] asac, was that "no before" or was that "before plz" ? [16:08] s/no/not/ [16:08] do it now [16:08] no time to loose [16:08] ok [16:08] maybe already too late [16:08] and they busted us [16:08] and apparmor blocks that. I have an idea I am playing around with to that work without widening the permission set or disabling apparmor during the tests [16:08] asac, that might end up badly for the system image but firing it off now [16:08] we'll at least have a good zip [16:08] seb128: Ok; might join but mostly wanted to follow the video [16:09] asac, running [16:09] ogra_: yeah. next time dont wait [16:09] :) [16:09] never wait if there is no bad bug [16:09] lool, ok, works for me [16:09] :) [16:09] lool, broadcast is going to start when we get more people [16:09] asac, well, i had the impression you wanted all this for the system image people ... [16:09] ogra_: they didnt give me requirements at all [16:09] asac, which will likely be unbootable now [16:09] also didnt approach me about coordinating landing [16:10] ok [16:10] I feel we're speaking of people as if they aren't in the room [16:10] hehe [16:10] hi [16:10] asac, i heard that someone thinks we are speaking about people like they arent in the room [16:10] thanks [16:10] yeah lets stay focussed on our mission to get automation back on [16:10] I dont think any big system-image changes are pending; the main things were adding back support for .developer_mode and avoiding a sudo warning; no big deals [16:11] kk [16:11] sounds good [16:11] and not worth delaying a good image superseding the one from the 22nd ;-) [16:11] lool, well, there were lxc-android-config changes as well as initrd changes [16:11] lool, the former will be in, the latter wont [16:11] yes, these are all nice to have but no big deals if they make it to an image [16:11] hope we didnt pick up any dirt [16:11] :( [16:11] it might break developer_mode for some people, but that's ok [16:11] asac, if you dont want that, freeze the archive [16:12] lool, ok, i was just cautious in case the lxc changes rely on the initrd changes or some such [16:12] right. we would want to do that if we could keep up with the pace [16:12] it looked like a set [16:12] but we cant so we can only control one side [16:12] and hope for the other === dandrader is now known as dandrader|lunch [16:12] and lool made a good assessment [16:13] it's sad if we didn't land all of it in the image, but it's ok; I don't know how far stgraber managed to upload things (long chain of deps apparently) [16:13] no big deal [16:14] lool, the final bit (android package rebuild) is missing [16:14] so the initrd changes wont be there [16:14] lool: I'm about to upload android (waiting for the publisher to finish) [16:15] lool: not waiting for android will cause a security issue btw [16:15] stgraber, ouch [16:15] lool: as lxc-android-config will be there, mount /var/lib/sudo without support for the custom flags [16:15] stgraber, i just started an image build [16:15] so that'll make /var/lib/sudo world writable [16:16] not an issue on the non system image [16:16] (/ is world writable there) [16:16] sure, the path won't be affected on non-system iamges [16:16] ogra_: I'll kick another rebuild once android is built and published and I'll manually promote that one to system-image [16:17] stgraber: let's just kick another build once android package is ready [16:17] I'd like Jamie apparmor update to be picked up too [16:17] it's not like this image will be immediately published from current to pending anyway :-) === dholbach_ is now known as dholbach [16:20] Who manages https://wiki.ubuntu.com/Touch/Install ? I got stuck in looping boot loader so in the process of going back to Android and then reflash. [16:20] In Device factory reset "In Recovery (Android robot on his back with a red triangle)... tap the volume up button and the power button simultaneously which will bring you into stock recovery. Again.. Don't hold the buttons, just tap them simultaneously. Also make sure you're holding the correct volume button. Up will be the volume key on the right." didn't work for me. Had to use volume down [16:25] Now I can're reproduce it. Damn fickle hardware :( [16:26] diwic, got a minute now ? :) [16:26] om26er, maybe [16:26] diwic, i get very low sound on nexus 4 [16:27] diwic, where to report bug for that ? [16:27] om26er, which device, speaker, earpiece or headphone? === dholbach_ is now known as dholbach [16:27] om26er, and which image are you running? [16:27] diwic, 20130827.2 with apt-get dist-upgrade afterwards [16:28] rsalveti, what is an L3 error anyway? [16:29] diwic, mako, on speaker and earphone [16:29] diddledan, headphone volume is fine btw [16:29] diwic: not sure yet [16:29] diwic: but at the bottom there's a dma transaction error, not sure if we should be concerned with it [16:29] rsalveti, "Custom errors in L3: Target like DMM/FW/EMIF generates SRESP=ERR error" [16:29] diwic, the one where you call the level3 support of the OEM perhaps [16:31] ogra_, heh [16:31] diwic: http://kernel.ubuntu.com/git?p=ubuntu/ubuntu-saucy.git;a=blob;f=arch/arm/mach-omap2/omap_l3_noc.c;h=0d6aaae95e8628d8d476cb3d505b309a03473054;hb=refs/heads/maguro [16:31] om26er, so what you're saying is that speaker (not earpiece or headphone) volume is low? [16:32] rsalveti, yes, that's how far I am as well [16:32] diwic, no, I mean speaker and earpiece have low volume but when I connect a headphone sound is pretty acceptable there [16:33] I had a call a few minutes ago and had to hold a finger on my other ear so I could hear the call clearly. [16:33] om26er, okay. and this was not the case a few days ago? [16:34] diwic, it was working till yesterday [16:34] om26er, ok, please file a bug for it ( touch-preview-images ) and include the output of "pacmd list" thanks [16:34] sure [16:35] family calling, sigh [16:36] hey guys, i've managed to get to Step 4 of the install. But i'm stuck on this command phablet-flash (cdimage-touch|cdimage-legacy|ubuntu-system|community) -b .... what am I doing with the bracketed parameters? [16:38] Theo, pick cdimage-touch from it :) [16:38] it means you need to pick one [16:38] yeah just didn't know which. I'm on my way now! [16:38] Thanks! [16:39] good luck [16:42] diwic, have a moment? [16:44] lool: LP: #1217932 will be fixed in system-image 1.3 [16:44] Launchpad bug 1217932 in Ubuntu system image "Dashes in channel name or group name breaks resolution" [Critical,Fix committed] https://launchpad.net/bugs/1217932 [16:45] rsalveti, I have to quit for the day and take care of my family instead. [16:45] mamenyaka, not today, sorry [16:46] diwic, no problem, maybe tomorrow === alex_abreu is now known as alex-abreu [16:54] barry: col [16:54] barry: cool [16:55] lool: anyway, i'll upload a 1.3 before my eod. we'll see what else i get in there by then ;) [16:55] Just recovered from a boot loader loop by doing a manual reinstall. Glad I'm not put off by such things :) [16:55] bigcalm: did you find what was causing it? [16:56] bigcalm: I was bitten by such a thing a while ago, barry had something ressembling recently [16:56] asac, ogra_: I see 20130828.1/ [16:56] lool: I can't pin-point it. Did an update and reboot this afternoon [16:56] jdstrand, "desktop": "test_underscore1.desktop" is the only bit containing an underscore - the rest has a hyphen [16:56] jdstrand, good to go? [16:56] lool: went into a loop from that point onwards [16:56] bigcalm: yeah, i nearly bricked my n7 yesterday after doing a phablet-flash. i got it working again though [16:57] dholbach: can you paste the complete manifest? [16:57] and latest image has the right apparmor-easyprof-ubuntu [16:57] jdstrand, http://paste.ubuntu.com/6037303/ [16:58] dholbach: yeah, that will work [16:58] jdstrand, the .qml and .desktop file have underscores, so I guess that's all right [16:58] ok, uploading [16:58] Mirv, ^ [16:58] stgraber: would you want me to push a no change android upload? [16:59] since it's big [16:59] lool, yep [16:59] bzoltan: note, dholbach's upload still creates: '"apparmor": "test-underscore1.json",' which doesn't need the transmogrification. Ie '"apparmor": "test_underscore1.json",' is fine. I know you know this and it is not a blocker [16:59] lool: I've got gigabit internet for that [16:59] lool, feel free to take over the builder [16:59] * lool bows [17:00] I have gigabit ... ethernet .... [17:00] to a cable modem :-) [17:00] jdstrand: I know, you know that I know it :) [17:00] lool: I'm just waiting for ubuntu-touch-generic-initrd to publish completely, it's currently stuck half way through (source moved to release but binary still in proposed) [17:00] dholbach: yes it is. it is only the "name", the "version" and the application name (ie, the keys to the hooks dictionary), that should not have underscores [17:00] bzoltan, fix uploaded to saucy [17:00] bzoltan: :) [17:01] stgraber, its already 3h btw :) === schwuk is now known as schwuk_away [17:01] stgraber: ack; I was furiously refreshing the android pacakge's page [17:01] jdstrand: Yes, I left this change there to reduce the number of lines the second fix attempt needed. Next week when JP is back we will do a little refactoring of the Click packaging and manifest handling. Visual version increaser and smarter policy group editor is to be expected [17:02] lool: don't trust LP for that, you need rmadison [17:02] bzoltan: sounds great. thanks again for all your work on this :) [17:02] stgraber: rmadison is slooooooow [17:02] lool: (LP thinks it's published at the beginning of the publisher cycle, so if you trust it, you upload and you change doesn't get picked up) [17:02] ack [17:03] jdstrand: :) I actually enjoyed that Click/Manifest solo sprint :) I touched code I have never seen before. Now I can fx things much faster if needed. [17:03] * lool walks out for dinner [17:04] cjwatson, since you thought i'm wrong thinking 2h is optimistic ... it is 3h now and android wasnt uploade yet :) [17:04] *uploaded [17:04] ogra_: my comment related to the time from upload to image generation [17:05] lool: I hope it will get better once we get the new archive offload box (snakefruit) [17:05] cjwatson: what does that do? [17:05] lool: move a bunch of jobs off the overloaded lillypilly [17:05] ok, cool === alan_g is now known as alan_g|EOD [17:06] ogra_: but ok, I hadn't realised that this required multiple passes through the publisher cycle [17:06] cjwatson, right, but it is three uploads for the whole chain before you can build images ... its closer to 4-5h [17:06] yeah [17:06] stgraber: maybe whatever this is that's waiting for ubuntu-touch-generic-initrd should be pulling it from -proposed [17:07] i'm really considering to move the bootimg generation back to livecd-rootfs [17:07] that would save the last upload ... which is the most annoying one === dandrader|lunch is now known as dandrader [17:16] asac, 28.1 is ther in case you didnt notice [17:17] cjwatson: ack, i think it's more for me, rather than stgraber. [17:19] xnox: hmm, my no change rebuild just FTBFSed, any idea? https://launchpadlibrarian.net/148638383/buildlog_ubuntu-saucy-i386.android_0.20130801.1-0ubuntu8_FAILEDTOBUILD.txt.gz [17:20] ogra_: does it work :)? /me installs [17:21] stgraber, xnox, likely missing upstream changes that rsalveti did for the camera fixes [17:21] asac, still rsyncing [17:21] ogra_: that's already pushed (to phablet-saucy) [17:22] sergiusens, doesnt help the package if nobody synced it :) === bschaefer_ is now known as bschaefer [17:22] stgraber, so there is your answer ... :) [17:23] sergiusens: ogra_: i'll sort it out today then. [17:23] ogra_: that buffer error seems more of a ricmm thing [17:23] xnox: let me know once you've uploaded it so I can trigger an image rebuild for system-image [17:23] the NativeBufferAlloc that is [17:24] sergiusens, well, after all there are phablet.u.c changes that need to go into the package ... if it still fails then .... [17:24] oh, but if xnox didn't filter it in, that explains it [17:24] ogra_: yeah, if it's in, this shouldn't fail [17:24] yep :) [17:28] stgraber: ack. [17:31] seb128: contact picker ui component exists already [17:34] I get boot loops now, is this actually correct in the new initramfs "touch" script? The ' in line 138 should be a ; I guess: if [ "$5" = "none" ]' then === bfiller is now known as bfiller_ === bfiller_ is now known as bfiller [17:35] bfiller, thanks for the pointer [17:37] stgraber, ^ (I think the initramfs script has a syntax error) [17:37] Hi everyone. I just got a nexus 4. Are you instructions in https://wiki.ubuntu.com/Touch/Install up to date? (as in no additional tweaks necessary) [17:41] w-flo: there's indeed a typo there... [17:41] ogra_: for me its good [17:41] lets wait for dashboard and mako confirm [17:41] and then push the trigger on this one [17:41] w-flo: didn't spot it when I reviewed lool's diff. I'll fix and re-upload all the mess. [17:41] stgraber, thanks :) [17:42] (and do a test boot before uploading just to be safe) [17:42] stgraber, hmm, that explains why tsdgeos had a reboot loop ... i guess the dist-upgrade updated the initrd [17:42] ogra_: hmm, I don't see how a dist-upgrade would update the initrd [17:42] ogra_: you need to generate a new boot image for that... [17:44] stgraber, nope [17:44] stgraber, update=initramfs just does the right thing [17:44] stgraber, w-flo: Sorry about that [17:44] tedg, btw, katie said she'd have to get back to us about indicators-in-greeter [17:44] jdstrand, Reading your e-mail about the autopilot stuff... I don't *think* that autopilot starts it's own dbus... that would be odd. [17:44] lool, it's actually good news, because it means the boot loop is *not* caused by an unsuccessful attempt to fix the alsa driver :) [17:45] mterry, Ah, okay, thanks for asking! [17:45] tedg, I see still no upstart update [17:45] w-flo: boot loop with which image? [17:45] lool, self made image (for ported device) [17:45] ok [17:45] tedg, how do we enable auto-pushing u-g-s-b to saucy once ready? [17:45] tedg: it does via dbus-launch. see the denial [17:45] mterry, Perhaps we should get together a call with her and mpt, it could be fun! :-) [17:46] mterry, We just need to update the config file. [17:46] mterry, I guess I need someone on didrocks' team to sign off on the packaging. [17:46] tedg: I don't know where, but somewhere in the stack dbus-launch is getting called under the qmlscene process [17:47] cyphermox, can you do preNEW stuff for auto-uploading? [17:47] jdstrand, Hmm, yeah. Seems odd that autopilot would do it, itself. [17:47] tedg: well, the is libautopilot-qt [17:47] s/the/there/ [17:48] tedg: istr a discussion from long ago about not running autopilot tests under the user's session [17:48] tedg: which is consistent with the denial [17:48] jdstrand, Just to be curious, why can't I start my own confined dbus if it's under my apparmor profile? [17:49] tedg: we could use '/usr/bin/dbus-launch ix,' I suppose, but we are trying to keep execs down to a minimum [17:50] jdstrand, I guess I'm more asking, why aren't we doing "/usr/bin/* ix" ? [17:50] jdstrand, They'd still be contained, no? [17:50] tedg: and I don't really like the precedent of adding extra accesses to policy for automated tests [17:54] mterry, cyphermox, https://code.launchpad.net/~ted/cupstream2distro-config/daily-session-broadcast/+merge/182721 [17:55] lool: code tested, uploading the fixed version now [17:55] tedg: they would be, yes, but that is more access than is required for SDK apps and that butts up against application lifecycle [17:55] stgraber: thanks; I'll add a sh -n + checkbashism test to the build afterwards [17:56] jdstrand, Eh, okay. I was more thinking that I might want to use "cal" in my program or something like that. [17:56] tedg: click packages don't have a concept of dependencies, so you can't rely on cal being present [17:56] stgraber: ogra_: sergiusens: that ftbfs is because the package is not yet reflecting the changes done in phablet.u.c [17:57] thats what i said :) [17:57] sorry, didn't read the entire backlog [17:57] but seems xnox will do the sync today [17:57] xnox: let me know if you need any other patch in phablet.u.c, but that should be good already [17:57] we only need to rebase your package patches [17:57] jdstrand, Sure, but I imagine your definition of "rely" and most devs are different :-) If it shows up on the phone they have and kinda works... ;-) [17:58] tedg: for the converged device, we are probably going to have a different template that might have those types of accesses [18:00] tedg: for now, this is the spec: https://wiki.ubuntu.com/SecurityTeam/Specifications/ApplicationConfinement#Launching_other_applications [18:01] rsalveti: yup [18:02] asac, all working (i dont have a headset handy, but dont think thats a prio) [18:03] ogra_: is 28.1 out already? [18:03] rsalveti, yep, works well for me [18:04] cool, let me flash it [18:09] Is the U.S. Cellular - Model SCH-R760 model supported? | http://askubuntu.com/q/338517 [18:11] hey guys, could do with some opinions. I'm trying to flash touch to a nexus 4. I got as far as phablet-flash which stalled in the terminal (still respondive but no return code and not really doing anything) on the device it outputted something along the lines of "cannot locate autodeploy.zip" i would assume the push failed. I've tried sideloading and adb push ing a zip i made outof the image but to no avail. The zip is viewable bu [18:12] Any help would be humbly appreciated [18:15] Theo: well it's strange the script has failed. Try rebooting phone, and running script again [18:15] Theo: note it doesn't need to re-downlaod everything [18:16] yeah i did that, i'll try it again and get the full output. [18:16] Theo: else, if it sticks there again, in the android recovery, choose the "sideload" option [18:16] Theo: then on your PC, change to hte download directory that phablet-flash indicated above [18:16] Theo: then run "adb sideload " to push file to device [18:17] Theo: push the file with "mako" in its name first [18:17] (i.e. the smaller zip file) [18:18] Theo: you just caught me going out the door, I'll be back online in 30+ [18:18] good luck! [18:18] i only see one zip in the 20130822 folder and its armhf.zip [18:18] kk thanks anyway! I'll still be on then [18:18] Theo: there should be 2. Hmm [18:18] presumably a little lost still. :) [18:19] please make sure your phablet-tools is up to date, then try again [18:19] i did both aptgetupdate and get upgrade [18:19] already there :P [18:19] Theo: you're running Saucy? [18:20] * greyback gone [18:21] rsalveti: not happy with the tarball, it's bloated, but i've rebased patches and can build off it. I'll send you the tree diff of "extra" not-used stuff in the tarball [18:22] xnox: why bloated? [18:22] yeah [18:22] we just need to remove whatever you want in there :-) [18:22] I got some pieces removed, but not all of them [18:22] Theo, http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20130822/saucy-preinstalled-touch-armel+mako.zip [18:23] thats the one greyback referred to [18:23] rsalveti: well the android-*.tar.xz is 142MB and the one used in the archive is 92MB at the moment =) [18:23] rsalveti: anyway running a test build at the moment. [18:23] xnox: right, have the diff for the stuff that needs to be removed in hands? [18:24] rsalveti: not at the moment, following vUDS whilst doing a test build. I'll upload with tarball as it is, and send you a diff of paths later. [18:24] it's not blocking rebase =) [18:24] xnox: right, but which tarball, the latest one available in there? [18:25] rsalveti: I used android-20130828.tar.xz , is that the right one? [18:25] yup, cool [18:25] =) === iveand is now known as rik-shaw [18:30] sideloading results in "error: closed" [18:30] but the devices is showing in adb device -l === porchao is now known as porcha [18:32] popey: may i beg for a smoke test? :) [18:32] of .1? [18:33] asac: I'm away from home for the evening, sorry [18:42] rsalveti: so in the packaging i was updating the .zip with ubuntu boot image, i guess I don't need to do that, and instead i can just take: zip, boot.img and android-boot.img as it is? [18:42] Anyone available to give a rookie a hand, a little stuck on the installation. [18:43] Theo: we are in the middle of vUDS =) uds.ubuntu.com so everyone is on google hangouts at the moment =) [18:43] xnox: the final zip from the tarball you're using should already contain the right boot.img by default [18:43] why do you need android-boot.img? [18:46] rsalveti: i used to ship it to support all types of images, but i guess there is no legacy images any more.... [18:47] yeah, all we care about now is the zip and boot.img [18:47] xnox: but I know you're not using the same logic to download the generic initrd in there [18:47] but if you get that available before the build (instead of downloading it with pull-lp...), it should be good already [18:49] mterry, Do you the upstart-dbus-bridge? [18:50] tedg: parse i fail [18:51] tedg: are you talking in en_RPN locale? [18:51] xnox, en_yoda [18:51] Insert "have" ;-) [18:52] rsalveti: i am not allowed to use pull-lp, so I use internal ftp mirror on the buildd. So it should be only ever-so-slightly delayed than pull-lp... [18:53] rsalveti: and yeah i pull it before build starts. [18:54] yeah, then it should be good [19:00] tedg, let me check [19:00] tedg, that's a package? Doesn't show up in my apt cache [19:02] ogra_: the camera app failure on mako is a known regression, right? [19:02] and expected? [19:02] yes [19:02] mterry, No, it should be in upstart [19:02] WIP [19:02] mterry, And provide the dbus event that we need [19:03] tedg, I have /sbin/upstart-dbus-bridge installed [19:03] mterry, Ah, cool and I'm guessing it's running? [19:03] You should actually have two [19:03] (per session) [19:03] anyone here has mako?: [19:03] :) [19:03] and can test .1 build so we can push it out? [19:04] tedg, none running [19:04] Hmmm [19:04] mterry, Are there jobs for them in /usr/share/upstart/sessions ? [19:05] Wonder if the PPA does different packaging. [19:05] tedg, no... /usr/share/upstart/sessions only has jobs for event and file bridges [19:07] mterry, Hmm, so that's the issue, but I'm not sure why. Let's see if we can find jodh :-) [19:07] rsalveti: sergiusens: https://plus.google.com/hangouts/_/775b5f9d919a2624105bafd80c1846c84238b409?authuser=1&hl=en to talk about app developer mode for read-only system images [19:08] /join #ubuntu-uds-community-1 [19:09] mhall119: thanks === bschaefer_ is now known as bschaefer === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [19:42] should you be messing with a development then? [19:43] wilee-nilee, who exactly ? [19:44] wrong channel my bad [19:44] smart phone fumbling [19:46] :) [19:47] smart phone dumb user, lol [19:55] Hi [19:55] I have problem in the configuring the SIP account on the Ubuntu 12.04 [20:05] xnox: did you find the source of the android FTBFS issue? [20:06] I'm running in the Double NAT Network.Does't cause any issue? [20:06] lool: he's working on updating the package with the proper tarball [20:06] lool: yeah, i've rebased patches and local build finishes. Will test that grouper still boots & will upload. [20:06] that will fix the ftbfs [20:06] cool [20:10] does anyone know which log file in /cache/recovery/[log | last_log] is the newer one? [20:13] achiang: when recovery exists it's supposed to dump it in last_log, so log should be the newer one [20:14] sergiusens: hm... i'm not sure i actually understand that clarification. ;) [20:14] sergiusens: recovery exists, dump info into last_log... what is in log then? [20:15] achiang: let me double check, I always have to double check anyways [20:20] stgraber: uploaded a sh -n check to avoid this happening again [20:20] a bit surprizing that lintian doesn't do this; it seems to have a hard time picking up interesting shell scripts to test [20:21] lool: would also be nice to get this tested once before uploading :-) [20:21] rsalveti: bah you know, did one last change before uploading [20:22] haha, yeah, I know [20:22] purely cosmetic [20:22] lool, btw, update-initramfs works on all supported devices [20:22] still [20:22] and does the right thing [20:22] stgraber: ping, i'm trying out the customized image tarball now, and still not passing the signature check. maybe i don't have a proper keyring from you? i have archive-master.tar.xz but do not have any other keyrings [20:22] (for testing) [20:22] achiang: ok, on recovery exit: copy_log_file(LOG_FILE, true); copy_log_file(LAST_LOG_FILE, false); ... 2nd arg is append, and it copies the TEMPORARY_LOG_FILE that lives in /tmp/recovery/log [20:22] spent a few minutes in panic mode because my device was in a reboot loop mode [20:23] but thanks for pushing the fix quickly, everyone :-) [20:24] i could disable the automatic flashing if someone thinks thats better [20:24] sergiusens: ok, so log is cumulative recovery attempts and last_log is simply the most recent attempt? [20:24] achiang: that's what the code shows [20:24] sergiusens: thanks [20:24] achiang: I see you are trying system images [20:24] ssweeny: ^^ [20:25] achiang: there's another log for that I think, let me check [20:25] sergiusens: well, i'm trying to get our customization tarballs to work [20:26] achiang: you need the 3 same keyrings as for a standard update, otherwise it indeed won't validate [20:26] achiang: (you need to load archive-master.tar.xz, then image-master.tar.xz and finally image-signing.tar.xz) [20:26] achiang: http://phablet.ubuntu.com/gitweb?p=CyanogenMod/android_bootable_recovery.git;a=blob;f=system-image-upgrader;h=b5bce1aa2cdf54933b0a66397e5585cb739e3d95;hb=refs/heads/phablet-saucy [20:27] stgraber: I'm just noticing that we still have archive-master.tar.xz.asc archive-master.tar.xz in the repo, we are good to get rid of that, right? [20:28] sergiusens: stgraber: thanks, i'll keep poking then [20:28] ogra_: lxc-android-config uploaded [20:28] sergiusens: why would we be able to remove it? we need to have those two files in the recovery image. [20:29] rsalveti, rightm, waiting for android then [20:29] achiang: the .asc I gave you was signed with image-signing, so you need to load the two others first to get a trust path from archive-master down to image-signing [20:29] stgraber: ack, merci [20:30] stgraber: hmm, sorry, I must of confused the names, thought we were downloading it [20:36] stgraber: dumb question but where do i find those keyrings (and signatures)? [20:37] achiang: http://system-image.ubuntu.com/gpg/ [20:37] stgraber: thanks. and they get pushed into /cache/recovery too? [20:37] achiang: system-image-cli grabs them and pushes them to /cache/recovery when preparing an update, yes [20:38] stgraber: hm, ok. [20:38] Hi folks, after upgrading the image on my nexus 7 today, unity8 crashes repeatedly. Is this a known bug I should wait out or do I need to do a clean reflash? I put some unity8 log info in bug #1218065 ... [20:38] bug 1218065 in Unity 8 "unity8 crashing repeatedly on nexus 7" [Undecided,New] https://launchpad.net/bugs/1218065 [20:46] stgraber: adding those keyrings, i achieved success with our custom tarball! thanks! [20:46] yay! === bschaefer is now known as bschaefer|lunch [21:08] sergiusens: where can I find the click package for com.ubuntu.dropping-letters? [21:09] kenvandine: ping [21:09] mhall119, pong [21:09] kenvandine: hey, a question came up during a session today about whether online-accounts providers can run under app isolation and be installable as click packages === wilee-nilee is now known as Guest91671 [21:09] do you know if that's possible? [21:10] I know there was some work going to so allow Scopes to be packaged as Click and run under isolation [21:10] jdstrand: ^^ would probably interest/affect you too [21:10] mhall119: currently no [21:11] jdstrand: https://jenkins.qa.ubuntu.com/job/dropping-letters-click/ [21:11] jdstrand: anything wrong with it? [21:11] mhall119: also, click and scopes is not for 13.10. see thread on ubuntu-devel@ from yesterday [21:11] kenvandine: jdstrand: probably not something that needs a solution before 13.10, but afterwards we might want to think about it [21:11] jdstrand: ok, understood [21:12] sergiusens: no, thomi had some autopilot issues using that package with them using upstart-app-launch and I'm trying to reproduce [21:12] mhall119: jdstrand I think it's in the image customization target, but not in click [21:12] jdstrand: ack [21:12] sergiusens: thanks! :) [21:12] jdstrand, thanks :) [21:13] heh, sure thing :) === bschaefer|lunch is now known as bschaefer [21:19] anyone able to help my phone install? [21:21] I have a nexus 4, everytime I try to run phablet-flash -b it comes up with error:too few arguments [21:21] anyone able to help me? [21:24] SeanHosie54: https://wiki.ubuntu.com/Touch/Install Step4, the parameters have changed [21:24] Your missing the cdimage-touch and board name [21:24] phablet-flash cdimage-touch -d mako -b (i think is it) [21:27] thankyou [21:28] how do I reply to 1 person :) [21:28] rsalveti: sergiusens: so building package from android.tar.xz makes grouper go into reboot loop. [21:28] shall I upload into (a) ppa or (b) archive [21:28] ogra_: ^ [21:30] xnox, hmm, are you sure you have the fixed initrd ? [21:30] yeah [21:30] you probably got the broken initrd === Fyodorovna is now known as wilee-nilee [21:30] xnox: try rebuilding [21:30] if it works, just push it [21:32] rsalveti: i'll try reflashing, somehow adb/fastboot keep disconnecting from the device for me. [21:32] * xnox twiddles thumbs [21:33] xnox, check your build log, you want ubuntu-touch-generic-initrd 0.24 or newer [21:33] that should have the fix [21:34] ogra_: right I caught 0.23 [21:34] once it has downloaded what command do I run to install the image please iBotpeaches [21:35] bfiller: so not sure, but think that landing now might be a good time :) [21:36] at least prep for landing [21:36] check with rsalveti and folks whats currently going on in archive etc. though [21:36] SeanHosie54: I wouldn't recommend flashing this, if you can't read the wiki and its detailed steps [21:38] iBotPeaches: I dont know what you mean? === jhodapp is now known as jhodapp|afk [21:40] * xnox current eta 20m for new build to finish locally 5m to flash test again + 2m upload + 40m build on buildd/publish [21:40] SeanHosie54: There is a well maintained wiki, which I linked you to that answers anything you are gonna ask me. [21:42] wow, ogra time travelled to commit something in the future [21:53] Access microphone/record sound from QML? | http://askubuntu.com/q/338610 [21:56] fginther: are autopilot tests disabled again for calculator app on CI jobs? [21:58] boiko, jenkins did something unexpected. I'll have it fixed in a moment [21:58] boiko, thanks for notifyingme [21:59] fginther: no problems, not urgent either [22:00] Hi, anyone in here who has working bluetooth keyboard for mako? I only get " "Connect" with signature "" on interface "org.bluez.Input" doesn't exist" when I try to connect with bluez [22:01] pairing devices works without problems since 2 weeks... [22:04] asac: ack, rsalveti are things good to go? if so 2 mr's need to go in [22:04] ChickenCutlass: https://code.launchpad.net/~bfiller/ubuntu-seeds/ubuntu-touch.saucy-add-dialer-messaging/+merge/182487 [22:05] ChickenCutlass: and this https://code.launchpad.net/~bfiller/unity8/add-new-telephony/+merge/182184 [22:05] bfiller: yup [22:05] bfiller, happroved [22:05] bfiller: will updated the seeds [22:05] bfiller, rsalveti I can not top approve [22:06] I'd just like the second one to be approved first [22:06] ChickenCutlass: no worries, I can take care of it [22:06] rsalveti, ack thanks [22:06] rsalveti, I bottom approved both [22:07] bfiller: actually, it seems you want the seeds approved first === salem_ is now known as _salem [22:07] rsalveti: yup, stuff needs to be in image before adding launchers to shell [22:07] cyphermox: can you re-enabled daily-release for address-book-app? we're ready to land it now [22:07] let me push the seeds changes [22:08] or robru ^^^^^ [22:08] bfiller: can we then remove phone-app from the ppa? [22:08] rsalveti: yup [22:16] bfiller: seeds and meta-package pushed [22:16] robru, cyphermox: here is MR to turn address-book-app daily-release back on https://code.launchpad.net/~bfiller/cupstream2distro-config/enable-address-book-app/+merge/182770 [22:16] rsalveti: thanks [22:17] rsalveti: you can happrove this one can you? ^^^^^^ [22:20] bfiller: yes [22:20] awesome [22:20] done [22:21] just need https://code.launchpad.net/~bfiller/unity8/add-new-telephony/+merge/182184 [22:21] will build & test [22:21] rsalveti: https://launchpad.net/ubuntu/+source/android/20130828-0ubuntu1/+build/4915218 [22:21] actually, just install the package from the zip [22:21] xnox: awesome [22:21] rsalveti: i'm afk until tomorrow =) [22:21] xnox: no worries [22:21] good luck [22:22] xnox: do we have packaging branch or such? or just plain tarballs? [22:22] just in case I need to fix something [22:22] rsalveti: lp:ubuntu/android =) [22:22] good [22:23] rsalveti: well, it's out of date. I just did, old-school debian by hand src package. [22:23] sure, no worries [22:23] rsalveti: but e.g. uscan & uupdate will fetch updated tarball et.al. [22:23] great [22:25] xnox: there's an interesting foo.txt in debian/patches :-) [22:26] hm? [22:27] check https://launchpad.net/ubuntu/+archive/primary/+files/android_20130828-0ubuntu1.debian.tar.gz [22:27] rsalveti: indeed, i'll let you work it out what it is =) [22:27] haha, right :-) [22:27] xnox: thanks for the upload [22:27] rsalveti: well I know, AllanBell knows and ubottu knows. [22:30] who me? [22:30] rsalveti: hm... not sure if the proprietary blobs package needs updating. [22:30] AlanBell: yes, you do. ;-) [22:30] xnox: probably, I can update them [22:31] AlanBell: remember we drafted that list together. [22:31] rsalveti: well, the package pulls from branches. But it's been a while, might need refresh, cause there have been 4.3.x releases.... [22:31] we need to update them with the 4.2.2 binaries [22:32] at least the same ones we're using internally === Deihmos1 is now known as Deihmos [22:33] * AlanBell lacks context [22:35] AlanBell: android package has foo.txt with http://paste.ubuntu.com/6038279/ and I said " I know, you know and ubottu knows that list" =) [22:35] AlanBell: i'm not sure there is anything else to say =) [22:39] hi there [22:39] Is there anyone online ? [22:48] is there anyone online ? [22:48] I need little help [22:52] bfiller: new phone app is awesome, happroving [22:53] bfiller: happroved https://code.launchpad.net/~bfiller/unity8/add-new-telephony/+merge/182184 === kentb is now known as kentb-out [22:55] is there anyone online ? [22:55] rsalveti, mind doing the image build yourself ? seems the package is still building and i dont want to stay online for another hour .... the cron build is enabled again so it will kick in at 8 UTC [22:55] ogra_: no worries, was planning to do it anyway [22:56] great, thanks [22:56] rsalveti: can you help me whit flashing ? [22:57] eXee: what is the issue? [22:58] i have nexus 7 3g and i`m flashing it like it`s described in https://wiki.ubuntu.com/Touch/Install#Step_2_-_Device_unlock but it stuck at the Google sign [22:58] all was fine whit x-term [22:59] no errors appear [22:59] mhall119,jdstrand: I'm in a thread with some of the online-accounts people to discuss this. [23:02] bfiller: can you send a mail to paul lars and and doan about the new test names? [23:02] so they can enable the right ones? [23:02] rsalveti: any ideas ? [23:03] bfiller: plars and doanac` :) ... sorry [23:03] thax [23:03] eXee: hm, I'm afraid the 3g version will not work out of the box [23:03] might still need some changes in the kernel/initrd used by it [23:04] rsalveti: how can i make it work (the noobs way) ? [23:05] would need to better understand what are the differences between the 3g version and the wifi one, it could be the partition table layout, it could as well be the kernel [23:05] so it's not necessarily trivial [23:06] starting to understand if the partition table is at least similar would already help [23:06] then checking if the kernel is actually different from the wifi one [23:06] sergiusens: http://bazaar.launchpad.net/~sergiusens/+junk/network/view/head:/network_gprs_provision_test.sh would not necessarily work if wifi is already enabled, right? [23:09] cjwatson: worry, I missed your ping. what is 'this'? [23:09] s/worry/sorry/ [23:09] cjwatson: please don't worry :) [23:11] jdstrand: This is following up to "22:09 kenvandine: hey, a question came up during a session today about whether online-accounts providers can run under app isolation and be installable as click packages" [23:11] ah [23:12] cjwatson: I didn't need a thread. I said 'no' :P [23:12] of course, I tempered that somewhat and said, not for 13.10 [23:12] jdstrand: Apparently some such plugins are just XML config files, which (I think) wouldn't need careful confinement [23:13] jdstrand: For those that involve arbitrary code, I agree, of course [23:13] cjwatson: well, I can create a template for them [23:13] But if it's just plain old data with API URLs and keys and the like, that should be handleable with the existing click hook code [23:13] jdstrand: I gave them some pointers [23:13] they would only need the install directory from apparmor's POV [23:14] since nothing would execute there. that would allow automatic reviews [23:14] If it's just linking an XML config file into some directory, that doesn't need apparmor, just a click hook provided by online-accounts. [23:15] it would be silly to have a profile for nothing though, but if they want to cc me in the thread, that would be fine [23:15] Even if it requires a bit of sanitisation/mangling, that can still be a hook, just with a script. [23:15] right. the point is, what happens if the next upload ships code [23:15] Well of course you don't write your hook to permit that :-) [23:15] as long as the review can be automated, it doesn't matter [23:16] (to me) [23:16] automated because it is safe that is :) [23:19] we could check for the presence of that hook, and then check what the package ships and verify they are xml (or whatever). yes, seems doable [23:22] The hook should do the checks itself, either by its pattern only creating links that are by definition safe, or by the Exec script doing whatever sanitisation is needed. [23:23] Either of those should be possible, rather than having to push things up to the review code. [23:40] bfiller: so, I'm ready to build a new image, but I'd like to wait for your unity8 to be in [23:40] bfiller: otherwise it'll fail to load the phone-app [23:40] I approved it 46 mins ago, should be merged soon [23:40] then we need to ping cyphermox to trigger the CI for it [23:40] so we can build a new image in ~2 hours [23:45] I don't trigger any kind of CI, only daily release [23:46] rsalveti: ack, I'll try it out when done [23:46] cyphermox: yeah, the daily release [23:46] that's CI as well [23:46] :-) [23:50] rsalveti: it's not clear when you just say CI, because there are other CI jobs in another jenkins that I also look at [23:51] right [23:51] but yeah, except atm there is lots of stuff running [23:51] we'll wait until they're done [23:51] yeah, the mr is still not yet merged [23:51] just phone? [23:51] hrm [23:51] not good though [23:52] yeah, it's merged [23:52] ouf :) [23:56] https://code.launchpad.net/~bfiller/unity8/add-new-telephony/+merge/182184 [23:56] approved one hour ago [23:57] will probably take a while to merge and be released I bet