=== SkavenXXI-[OFF] is now known as SkavenXXI === Ursinha-afk is now known as Ursinha === Ursinha-afk is now known as Ursinha === SkavenXXI is now known as SkavenXXI-[OFF] === Ursinha is now known as Ursinha-afk === Ursinha-afk is now known as Ursinha === chihchun_afk is now known as chihchun [02:31] Dummies_freelanc: sorry for the delay, yes you can use qtquick for desktop apps too === Alys is now known as The_Letter_M === juicyjones|away is now known as juicyjones === juicyjones is now known as juicyjones|away === mmrazik is now known as mmrazik|afk === mmrazik|afk is now known as mmrazik [07:16] good morning === dholbach_ is now known as dholbach === chihchun is now known as chihchun_afk [07:43] Hi, I inserted a SIM in Galaxy Nexus with latest (147) Ubuntu touch image, but phoning possible [07:53] phoning not possible, any ideas? === chihchun_afk is now known as chihchun [07:53] it is not asking for SIM PIN [08:15] didrocks: before I depart to the tax-enforcers - I see that otto failed the Apps AP tests because of "packages: ERROR" [08:15] didrocks: what does that mean? [08:16] sil2100: did you try to look at the artefacts? I'm interested to see if the messages are clear enough for figuring it out :) [08:17] sil2100: btw, once you are back, mind looking at bug #1186225? (see the rationale, explanation and what's need to be reverted in both branches) [08:17] bug 1186225 in Canonical Upstream To Distro "Needless package build in case both control and changelog changed" [Undecided,Invalid] https://launchpad.net/bugs/1186225 [08:22] didrocks: will check and will look further - for now the logs just make fun of me, as the main console output said: "E: A run step failed. Check summary.log for details", while summary.log says only: "packages: ERROR" ;p [08:22] sil2100: right, tell me once you start looking at the artefacts :) [08:22] (I think we should change packages: ERROR as "packages setup: ERROR") [08:23] so if I were you, I would check the "logs" directory in the artefacts ;) [08:25] didrocks: yes yes, I see it - it's actually all nicely 'put in place', and it's rather intuitive that when there is a packages ERROR, I would look at the otto-setup.log [08:25] Nice ;) [08:25] sil2100: got it! :-) [08:26] sil2100: so it seems the ppa was in a middle of publishing [08:26] and apt-get update failed to catch it [08:26] sil2100: I just added a retry 3 times, separated by 10s [08:26] that should do it ;) [08:26] (in case of apt-get update fails) [08:26] sil2100: I'll rerun it, but nice to have your feedback! [08:26] Good idea, since there are a few cases where it could fail for no real reason [08:26] right [08:27] packages setup: makes more sense to you? [08:29] Yes, more informative indeed! [08:33] sil2100: why did you relaunch unity head? [08:33] sil2100: it was just in manual publishing, isn't it? [08:33] or is it my chrome with cache which is tricking me? [08:33] * didrocks ctrl + R [08:33] oh right, thanks jenkins cache :/ [08:33] sil2100: forget about it ;) [08:38] sil2100: oh btw, if you look at the job, you will see that you have labels with the stack/series/ppa === 64MAAD496 is now known as aquarius === chihchun is now known as chihchun_afk === chriadam is now known as chriadam|away === chihchun_afk is now known as chihchun [10:05] ogra_: I cheated I part chopped the crespo against galaxy nexus :) no more build worries :) [10:13] cool ! [10:17] ogra_: I got feed up of it not building and when I finally forced it to it not working :( so now I be a happy bunny :) === mmrazik is now known as mmrazik|afk [11:11] popey, ogra_: is it me or does the twitter feed not move on the galaxy nexus? I'm scrolling up nothing scrolling down nothing [11:12] * ogra_ has no idea === patricia is now known as Guest54184 === MacSlow is now known as MacSlow|lunch [11:42] mhall119: ping [11:43] mhall119: When you have time, can you test my MP https://code.launchpad.net/~nik90/ubuntu-clock-app/change-timer-logic/+merge/166982 [11:44] mhall119: We can then get that merged and I can then focus on the visual UI layout change. === pete-woods is now known as pete-woods-lunch [12:08] sergiusens, rsalveti: hi! Are we ok for the QA (autopilot) and unity stacks switch to saucy? Were you able to release the final raring image last week? === dholbach_ is now known as dholbacj === dholbacj is now known as dholbach [12:14] sil2100: btw, it seems that unity head depends on indicator raring stack, I think it should be now indicator head stack, isn't it? [12:14] sil2100: also, the hud stack (basically all rdepends stacks) should be built on saucy as a consequence [12:15] (btw, we're going to plug hud with otto, don't relaunch a stack without telling us ;)) [12:16] Ok ;) [12:22] Hi! === mmrazik|afk is now known as mmrazik [12:37] Mirv: ping [12:38] Mirv: Do you have some time ? I need some help on a Qt5 cmake module ^^' [12:40] Oranger: pong.. I haven't used the cmake much myself (mostly qmake all the way), although I know it's used and preferred by many of the developers [12:40] maybe fire away with questions and let's see if anyone can answer them? [12:41] Mirv: I asked to didrocks ( mhall119 said he is a pro in cmake ) and he told me to see with you :p [12:41] But ok :) [12:43] ;) [12:44] hmm, maybe from some of the packages we'd see who has used cmake a lot together with Qt5, not only one or another [12:44] Mirv: Thank you anyway :) I can't ask it here because it is a very specific problem so I'll try figure it out alone ;) [12:45] Mirv: Yes, I did some researchs, a lot of packages use cmake with Qt5, all with the same file (FindQt5.cmake) but it seem to don't work for me === MacSlow|lunch is now known as MacSlow [12:49] Oranger: I could continue with the list of potential persons to ping with Saviq, from looking at the dee-qt code history. I bookmarked that code branch at one point since it's an example of using cmake to build both against Qt4 and Qt5 [12:49] Mirv: Ok, thank you :) [12:49] Oranger, I'll keep an eye on the channel [12:50] Saviq: Wow, you are fast :) === pete-woods-lunch is now known as pete-woods [13:07] rsalveti, hi, are you familiar with how carriers communicate timezone + time + date to a phone when it's on automatic time setting? [13:11] folks. Have just phablet-flashed a nexus 10, but failing to get past demo user login screen as anything but guest [13:11] suggested lower case names as passwords don't seem to work [13:12] john-mcaleely: I think it changed to "password" now [13:14] mzanetti: yup - that works thanks :-) [13:19] <[SLB]> would installing ubuntu touch also wipe a possibly custom recovery from the phone? [13:20] [SLB], we're using a tweaked ClockworkMod recovery AFAIK, but it shouldn't be a requirement [13:20] sergiusens, can you comment ^? [13:20] <[SLB]> thanks Saviq [13:24] [SLB]: only if you use phablet-flash -b [13:24] <[SLB]> ah okay [13:24] <[SLB]> and is the current image actually usable or still the demo one? [13:28] sil2100: define unity stack if you mean something different that http://bazaar.launchpad.net/~cupstream2distro-maintainers/cupstream2distro-config/trunk/view/head:/stacks/head/unity.cfg [13:29] [SLB], some of us are using it as main phones, so it's possible [13:29] sergiusens: yes, I mean this + qa, hud and all other unity dependencies [13:29] jounih: The ubuntuShape and app theme capabilities sdk support are still being worked on. This essentially means that the visual design implementation can only be started after a few weeks. [13:29] [SLB], there's of course a lot of features still missing [13:30] sil2100: oh.. ok, so we shouldn't call all that unity stack or rename the stack in cupstream2distro ;-) [13:30] sil2100: anyhow... we have a discussion regarding that today... so I'll get back to you soon [13:31] it's not called unity stack, it's the unity stack + all dependent stacks (which has other names: qa, hud, indicators…) [13:32] Yes, I meant, the idea is to switch the unity stack to saucy, but to do this also qa, hud and other stacks need to be switched as well [13:32] I didn't say they're all called 'unity stack' [13:33] <[SLB]> thanks Saviq, nice [13:33] rsalveti, are you able to boot with the latest grouper kernel === ara_ is now known as ara [13:34] * ogra_ goes mad, i cant get the flipped container boot on grouper and there is no reason why it wouldnt [13:34] (root is mounted fine, it seems to fail after run-init) [13:37] ogra_: the latest kernel without or with the flip? [13:37] with [13:37] ogra_: I can give it a quick go either way [13:37] recovery with the same kernel works fine [13:38] i can debug /init down to it executing run-init [13:38] i asssume i even properly end up in the rootfs [13:38] but upstart doesnt move on [13:41] maguro works flawless btw [13:42] (apart from https://code.launchpad.net/~ogra/session-manager/fix-oom_adj/+merge/166975) [13:42] nik90: ok let me speak to kaleo about the ETA for those. Is there anything we can start with now? [13:43] jounih: I can try to match the placement of elements like the text to match the designs but otherwise not much. [13:43] jounih: The background gradient, ubuntuShape etc all require the sdk to provide support [13:47] sergiusens: where is your test suite? [13:47] sergiusens: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdGR3c1NUM2RnYkNBYjVMTkMxVjFqb2c#gid=0 can you fill out the phonedations row there: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdGR3c1NUM2RnYkNBYjVMTkMxVjFqb2c#gid=0 [13:49] asac: is this for tomorrow? [13:49] sergiusens: no [13:49] :) [13:49] not sure what tomorrow is [13:50] we have wednesday a call about saucy upgrade [13:50] Tuesday [13:50] afaik [13:50] mpt: :-P [13:50] asac: I'm not sure either... asked and got no reply [13:50] * mpt ducks [13:50] * asac hunts down the duck :) [13:50] sergiusens: this spreadsheet is about saying where your test suite is [13:50] and how to run it [13:50] and where it shoujld be run [13:50] daily [13:50] asac: well... I have no test suite yet... at least no formal one [13:51] asac: I'll create one [13:51] sergiusens: mfrey told me phonedatils has a test suite [13:51] that you run regularly [13:51] locally [13:51] thats what i am looking for [13:52] asac: he's probably talking about the manual ones? Or are you talking about the autopilot ones? [13:53] sergiusens: he said u do manual tests, but also run scripts etc. that are automatically testing stuff [13:53] asac: yeah, the autopilot ones then [13:53] sergiusens: do you also run those locally ? [13:53] if so, then yes... those are the ones :) [13:54] sergiusens: guess i need a list of autopilot packages that you run [13:54] sergiusens: check out the spreadsheet [13:54] asac: I am [13:54] something similar: one url, one set of commmands that will run everything, report true or false and produce logs [13:55] ok let me know if thats impossible :) [13:55] asac: anyways, with this new org, I'd like to get rid of those and do more lower level ones [13:55] asac: impossible today... but I can do something that aggregates all [13:56] sergiusens: I woul dlike to get what exists for now, but delegate responsibility of what is rewally in ther eto the phonedations team [13:56] asac: ack [13:56] i guess you can just provide a branch that has two simeple scripts: [13:56] ./list-test-suites [13:57] ./run-test-suite TESTSUITE OUTFILE.log && echo success [13:57] then you can firest run your autopilot stuff [13:57] and later just run whatever you find is better [14:04] mhall119: sforshee is working on powerd, but ChickenCutlass's team will eventually take over [14:06] thanks mfisch [14:06] sforshee: ChickenCutlass: https://bugs.launchpad.net/powerd/+bug/1186469 [14:06] Launchpad bug 1186469 in powerd "segfault on nexus 7" [Undecided,New] [14:07] because of this, you can't suspend the screen on a Nexus 7 in build >=145 [14:07] * ogra_ sighs === chihchun is now known as chihchun_afk [14:10] mhall119: hi! is the touch application entry still using .desktop file? [14:13] The following packages have unmet dependencies: [14:13] ubuntu-touch : Depends: demo-assets-video but it is not installable [14:13] GRR ! [14:14] ogra_: how did that make it there? [14:14] i synced the seeds [14:15] as well as all scripts etc [14:15] so that wer are on par between raring and saucy [14:15] ogra_: I was creating a bp to sync all this stuff :-) [14:15] i had to binary copy the new demo-assets in the PPA but they should be there [14:15] ogra_: as in do it in the right order [14:16] well, your only seed change regarding the demos wha the one that drops demo-assets and adds the -video|-music|-.... ones [14:17] penk: yes [14:17] python-gi is in the archive ... so there is no reason why it wouldnt be installable [14:17] ogra_: regarding the demos yes [14:17] right [14:17] i didnt merge anything that looked like it could break us [14:18] ogra_: ack... let me fix demo assets for you [14:18] and i dont get why that package isnt found .. [14:18] ogra_: python-gi? [14:19] sergiusens, you added a dep on that [14:23] sergiusens: don't you have the ppa sync script already? [14:23] we might want to sync the packages again [14:23] ogra_: only difference with nexus 7 is that we're now using the packaged kernel [14:24] but if should work, as it works fine with the current raring image [14:24] sil2100: as sergiusens said, guess we'll talk a bit more about the transition plan at our daily (in 35 mins) [14:24] awe_, hi, how much do you know about timezone + time + date communication between phones and carriers? [14:24] mpt: hey [14:24] rsalveti, well, i can see it getting to run-init and then it hard reboots [14:24] ah, hi rsalveti :-) [14:25] mpt: the carrier can send time + date + timezone offset to the phone [14:25] mpt: but it's not the timezone itself, it just says utc - [14:25] rsalveti: add your ideas here https://blueprints.launchpad.net/touch-preview-images/+spec/foundations-1305-saucy-migration [14:25] interesting === chihchun_afk is now known as chihchun [14:26] rsalveti, does it ever send timestamp + timezone, as opposed to just timestamp? [14:26] mpt: I can get the date [14:26] rsalveti, sorry, I meant the other way around [14:26] mpt: * "data" is const char * pointing to NITZ time string [14:26] * in the form "yy/mm/dd,hh:mm:ss(+/-)tz,dt" [14:26] ARGH [14:26] rsalveti, does the carrier ever send *just* the timestamp and not the time zone [14:26] demo-assets-video has an "s" in the end ... [14:26] silly me [14:26] guess sergiusens had the same issue [14:27] ogra_: yes I did! [14:27] but the branch was up to date [14:27] mpt: well, it sends both (but sometimes it can indeed just send the date/time) [14:28] let me check what I'm getting here, one sec [14:28] rsalveti, sergiusens: ok, so give me a sign once you guys have all talked out, until that I'll be waiting with the transition for a green light from you guy [14:28] *guys [14:28] rsalveti, the reason I'm asking was that in I assumed that it never sent the time zone, so you had to set that yourself first [14:28] sil2100: I'm mostly sure it's ok, but let's do this in sync [14:29] rsalveti, but if you often but not always get the time zone, then the time zone setting should be positioned after the automatic/manual setting [14:29] * ogra_ would appreciate if someone with a nexus4 could test the flipped container images [14:29] ogra_, I can. URL? [14:30] ChickenCutlass, http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ [14:30] install with adb, first the armel zip, then the armhf one ... [14:31] to get a shell you need https://code.launchpad.net/~ogra/session-manager/fix-oom_adj/+merge/166975 [14:31] (adb should work all the time though, at least it does on maguro) [14:32] ogra_, so which files do I need exactly? [14:32] well, what arch is n4 ... mako ? [14:32] yes [14:32] you need the armel zip for mako and the armhf zip with the rootfs [14:33] http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/saucy-preinstalled-touch-armel+mako.zip [14:33] http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/saucy-preinstalled-touch-armhf.zip [14:33] these two [14:33] (identical with raring installs) [14:33] ok [14:33] ogra_, how about bootimage? [14:33] not needed [14:33] inside the armel zip [14:33] ok [14:34] (like the tarball is inside the rootfs zip) [14:34] that will all get cleaned up [14:34] mpt: right, let me dig a big more [14:34] ok [14:45] ogra_: do you see text in the ubuntu-terminal-app? [14:46] i have never seen the terminal app in action [14:46] only on screenshots [14:46] ogra_: ah okay :) [14:46] mhall119: do you see text in the ubuntu-terminal-app? [14:47] davmor2: I can't see what I'm typing [14:48] rsalveti: ah thanks I see it flash but then it goes black again :) but I'm glad it's not just me :) [14:48] alias for 3gup and 3gdown created though woohoo [14:49] right [14:50] davmor2: increase the font size to 14 in the settings tab [14:50] mhall119: oh okay [14:51] mhall119: Yay thanks [14:51] mhall119: and how do you trigger esc for vim? [14:52] davmor2: in the toolbar there's a "Panels" button [14:52] one of the panels (ctrl I think) has an esc button === dandrader is now known as dandrader|afk [14:52] mhall119: ah nice thanks, I had the control button just not the ecs thanks [14:53] rsalveti, any luck? [14:53] mpt: comparing now with the logic in android [14:53] I got 13/06/03,14:29:11-12,00 here [14:54] "-12,00"? [14:54] -12 is -3 [14:54] divide by 4 [14:54] because of weird timezones, such as the one in venezuela [14:54] Time zones are in 15-minute increments? [14:55] and the time I got was local time, so with the timezone offset applied already [14:55] mpt: yup [14:55] ricmm: would be nice to check yours, you'd need to add -d to your /etc/init/ofono.conf and reboot [14:56] then look for UNSOL_NITZ_TIME_RECEIVED in your /var/log/syslog [14:56] rsalveti, but you say that string sometimes doesn't include the timezone. [14:56] rsalveti: apparently upping the font size does display the text which is useful to have :) [14:57] rsalveti: gimme 5 [14:57] mpt: at least when I was checking that wasn't always necessarily true, that's why I'm now checking how android deals with this [14:57] davmor2: https://bugs.launchpad.net/ubuntu-terminal-app/+bug/1187015 [14:57] Launchpad bug 1187015 in Ubuntu Terminal App "Set default font size to 14" [Undecided,New] [14:57] cool [14:58] mpt: so in android there are 2 options, automatic time and automatic time zone [14:58] they split it in 2 different options [14:59] hmm [14:59] You might want both on if they're both right [14:59] right [14:59] You might want neither on if they're messed up [14:59] When would you want automatic time and manual time zone? [14:59] ok, it seems they try to grab the country code from the sim [15:00] and then create the right timezone entry based on country code + tz offset [15:00] International roaming? [15:00] right, the country code the sim is registered [15:00] Several countries have multiple time zones... === derp is now known as Durbinator [15:05] afternoon [15:08] mhall119: do you have any issues with the twiiter web app? currently I can't scroll down my time line, however it works fine for facebook. [15:08] twitter even === dandrader|afk is now known as dandrader [15:09] davmor2: I don't really use twitter [15:10] mhall119: I was just testing how dogfoodable it was for me before I took my sim out of my S3 and dropped it into the galaxy nexus :) [15:11] sergiusens, where is the build number on the phone now? [15:12] zsombi: ping [15:12] mhall119: pong [15:12] zsombi: do you have time for a hangout to discuss responsive layouts? [15:13] pmcgowan: same location [15:13] mhall119: oh, completely forgot about it!!!, sorry, will join soon!!! [15:13] sergiusens, remind me! [15:15] pmcgowan: /system/ubuntu_stamp [15:16] ty === ckpringle_ is now known as ckpringle [15:36] mhall119, ping [15:36] ZDmitry: pong [15:37] mhall119, which font size do you want to set as default? [15:37] 14 seems to be the magic number, that or higher seems to fix it [15:40] mhall119, but is it comfortable size? Not small? [15:40] ZDmitry: it's larger than the default I get on the Nexus 7, it's a good size for me [15:41] Hi all. Just installed ubuntu touch on my galaxy nexus. I'm sure theres a page somewhere that will inform me how to delete the place-holder data and install the develepment apps, but thought I'd ask here for more update information === chihchun is now known as chihchun_afk [15:48] mhall119, sorry. That was mistake. It would better to set bigger font size from beginning. Then let 14 be default. [15:49] Hmm, case in point: https://wiki.ubuntu.com/Touch/ReleaseNotes says that Ubuntu touch is based on 12.10 ? [15:51] folf: updated it to say 13.04, thanks for spotting that [15:51] folf: as far as I am aware, the process for removing place-holder data hasn't changed [15:52] mhall119, anyway, this will be quick fix for the bug. Problem is still in konsole qml plugin. But I'll fix it only on the next week. [15:53] thanks ZDmitry [15:53] np [15:53] ZDmitry: oh, I spoke with the upstream KDE developer managing konsole, he's looking at your plugin branch and will try and help us get your changes integrated into their brances [15:54] mhall119, good [15:56] mhall119: then I just need to figure out where that information is :-) [15:57] folf: https://wiki.ubuntu.com/Touch/ReleaseNotes#Sample_content ? [16:00] mhall119: yeah, I just tried to look there. The first part doesn't seem to work for me, but then again there are no contacts. So maybe that doesn't apply anymore? [16:00] folf: could me [16:03] folf, we took all the sample apps and contacts and messages out [16:03] only videos remain === msvb-lab is now known as msvb [16:11] pmcgowan: OK, thanks. So I guess that could be taken away from the release notes page mhall119 === olli_ is now known as olli [16:15] NIGGER [16:15] BITCH [16:16] !ops [16:16] Help! Channel emergency! (ONLY use this trigger in emergencies) - oubiwann, corp186, popey, k1l, Corey, wgrant, idleone, and ogra. [16:16] nIGGER [16:16] CUNT [16:16] !OPS [16:16] Help! Channel emergency! (ONLY use this trigger in emergencies) - oubiwann, corp186, popey, k1l, Corey, wgrant, idleone, and ogra. [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:16] !OPS [16:17] !OPS [16:17] !OPS [16:17] gee, why am i an op ... [16:17] ogra_: Nice command [16:17] heh [16:18] * ogra_ doesnt even remotely know the runes too kick or ban people ... once i have looked them up the troll is gone :P [16:18] yes, but you can ban the troll... forever \o/ [16:18] heh [16:18] and the peace will be back in the Ubuntu Touch world === ogra_ is now known as ogra [16:20] sad to see some people just don't have a life :( [16:21] mhall119: The instruction "Run manage-address-books.py delete " from this page https://wiki.ubuntu.com/Touch/ReleaseNotes#Sample_content does not make sense to me. Perhaps I'm missing something obvious [16:27] folf, whats the confusing bit ? [16:28] you execute the command "manage-address-books.py delete" which clears the address book === dandrader is now known as dandrader|lunch [16:31] mhall119: is there a way to switch out apps in launcher? ie is there a config file for it somewhere? [16:33] mhall119, fix done: https://code.launchpad.net/~hiroshidi/ubuntu-terminal-app/font-size-fix/+merge/167087 [16:35] hey guys now I have an up-to-date ish build of ubuntu-touch I have to say well done it rocks :) [16:35] apart from twitter [16:36] davmor2, it's in applications.js file from qml-phone-shell package atm [16:36] Saviq: ah cool thanks [16:36] davmor2, we're working towards having it proper in conf atm [16:36] Saviq: I will hold off then and try it when it lands [16:36] ogra: the confusing bit is that I can't seem to locate "manage-address-books.py" [16:37] davmor2, well, depends on how pressing it is for you, it's still at least two weeks away ;) [16:37] folf, oh, it might be poutdated, the demo users were removed in the most recent images [16:37] -p [16:38] Saviq: not pressing at all I only wanted to drop a couple and add the terminal app, neither of which were ergent [16:38] urgent even === mmrazik is now known as mmrazik|afk [16:44] mpt: so, it's kind of ugly, it gets the time + timezone offset (if it's available), and together with the carrier ID + country code, it tries to match and find the correct timezone the user is in [16:44] that's why it works [16:45] mpt: and that's why you have both options, to set automatic time & date and also to automatically set the timezone [16:46] and it always uses ntp if it's available (if there's a data connection), as that's probably more correct [16:48] mpt: so don't know if you want to allow auto timezone as well in the ui [16:48] from the link you gave it gives the impression that it'll only set the time and date, and not necessarily the timezone [16:48] as I might want to manually select a timezone [16:49] sorry, you do have timezone specifics below [16:50] rsalveti: is this (https://wiki.ubuntu.com/Touch/ReleaseNotes#Timezone) what you are looking for?? [16:50] folf: no, looking for the new ui design to cover such features [16:50] but thanks :-) [16:52] mpt: so, for time we don't necessarily need an active data connection, as it can get that value from the carrier (same for timezone) [16:53] so it'd only need a data connection in case there's no modem at all (e.g tablet) [17:01] So to whoever cares: in the https://wiki.ubuntu.com/Touch/ReleaseNotes#Adding_Content pages it says to look at /usr/share/demo-assets/contacts-data/data.csv but that's not present on my phone. [17:01] folf: apt-get install demo-asstest-contacts [17:01] err demo-assets-contacts [17:02] folf: feel free to update the wiki [17:02] sergiusens: thanks [17:03] sergiusens: (and I wasn't aware that I could do that; the "login" button had eluded me) === dandrader|lunch is now known as dandrader === greyback is now known as greyback|away [17:39] end of day :) [17:39] almost [17:40] whats news on knocking out the saucy switch over? [17:40] lucky you [17:40] is container flip landing, yes or no? [17:40] asac, we will only be able to tell ionce the apps work or dont work [17:40] that will need some time [17:41] ok so we dont do this before switch [17:41] sounds good [17:41] we will switch the jenkins builds to saucy and roughly stablize them [17:41] nice [17:41] in parallel cdimage will build the flipped images [17:41] what do we define as "stablized enough"? [17:41] ack++ [17:42] and we will research whats working and what not within the next two weeks on these flipped images [17:42] swounds good. we don't block the switch over, but keep ability to flip over open\ [17:42] comes with some additional overhead to allow us to do that, but seems we figured out how [17:42] dunno what we define as "stablilized enough" ... i guess thats still open for definition :) [17:43] * ogra cries .. grouper is treating me mean today :( [17:43] right. i think the right angle is to find an answer is to think about "what is the inner ring of image stakeholders and what would block them" [17:43] ogra: grouper == N4? [17:43] 7 [17:43] the flipped container image works absolutely flawless on maguro (gnex) [17:44] ok. that one i won't kill without dual install ... it has my preferred pinball game on it :) [17:44] but not at all on grouper [17:44] it commits suicide when it should run run-init and switch to the rootfs [17:44] and i dont know why [17:45] what kind of suicide? silence death? or crying out loud? [17:45] reboot [17:46] seems you found the reboot feature :_) [17:46] hehe [17:46] thats one of these crappy android "features" [17:46] yeah [17:46] thats the CIA feature i am sure [17:47] ogra: how do you plan to debug? [17:47] just curious [17:48] well, currently i add a lot of stuff to /usr/share/initramfs-tools/init ... that writes to a file on disk [17:48] to get me all info about the environment [17:48] what about /proc/last_kmsg (assuming ubuntu kernel has ram console enabled)? [17:48] asac: reboot is easy first it lands on your left foot you chip it over to your right and then boot it again right? [17:48] but it doesnt reveal enough ... so tomorrow i'll build a new kernel with USB-serial enabled (which means killing adb) [17:49] davmor2: yeah. actually with good applciation lifecycle model and snappy boottime, a reboot shouldnt be that bad [17:49] user just sees slight screen black disruption and then bang... the device is back exactly where it was before :) [17:49] [ 6.597432] The USB cable is connected [17:49] [ 6.597510] elan-ktf3k 1-0010: Update power source to 1 [17:49] [ 7.006120] Kernel panic - not syncing: Attempted to kill init! [17:50] not really revealing much more than i know [17:50] i know that GPU driver folks have a full OS and they also optimize the "reboot GPU in case there are troubles we don't undersatnd" case :) ... so maybe its valid in the future on full devices too :) [17:50] well, usually you just use a serial console in such cases [17:50] ogra: hm, yeah, didn't really expect much more, it has to reboot if init crashes :/ [17:51] tassadar_: it could restart init :) [17:51] tassadar_, pfft it doesnt on my desktop [17:51] the point is that it isnt clear which init crashes :) [17:51] ogra: rename our init? [17:51] we run one in the initrd ... and one on disk [17:51] to real-init? [17:51] and then ? [17:52] then hope that the message tells you which init crashed? [17:52] do you think the kernel will tell me "hey your freaky named init crashed" [17:52] it will print exactly the same :) [17:52] its a hardcoded msg [17:52] ogra: what does it do on desktop? fall back to some "shell"? Oo [17:52] just telling you that the init currently in use died [17:52] you dont know which one is currently in use though [17:53] tassadar_, depends when exactly the crash happens, but yeah [17:55] that seems a bit odd, I didn't think kernel has something like that built-in [17:56] there is no ubuntu without initrd ;) [17:56] no need for that in the kernel [17:56] well kernel does this if process with PID 1 crashes, there is nowhere else to go [17:56] right [17:56] (or, can't be found) [17:56] it would do that for /init .inside the initrd [17:57] but doesnt do that for /sbin/init of the rootfs [18:00] tassadar_, what massively bothers me that the same thing just works on a galaxy nexus but doesnt on the nexus7 [18:01] same rootfs and initrd [18:02] maybe something wrong with kernel config? you'll need to get the log out anyway thouhg [18:03] I saw some info echos in ubuntu's init script, where does that go? [18:04] hi guys, can ubuntu touch apps be made using vala? [18:04] tassadar_, /dev/console ... [18:05] ogra, sergiusens: hey, so is there any work yet to make phablet-tools work with the saucy images on cdimage.u.c? It seems http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ doesn't have the full set of files that phablet-flash looks for [18:05] slangasek: I have a bluepring for that one... will land tomorrow if everything gets approved [18:05] slangasek, no, these saucy images arent really enduser friendly yet [18:06] slangasek, there are still to many issues with the flipped container images (which these are) ... tomorrow we will start to build jenkins saucy images in parallel so endusers have something while we all fix the flipped images [18:07] ogra: so the move to cdimage is still blocking on the container flip? [18:08] slangasek, i dont have the time to re-implement non flipped images in parallell and at the same time work on the flipped ones [18:08] ogra: ok, so what are the remaining issues with the flipped images? [18:09] jenkins seemed like a good compromise for a two week thing [18:09] two weeks is a really long time [18:09] how can we help you get the container flip done sooner? [18:09] slangasek, well, currently i can get the nexus7 to run run-init on the bindmount [18:10] generally the platform api likely needs to learn about how to even run apps in the new model [18:10] i can start apps but end with a white screen [18:10] then we need to make sure device access across the containers works [18:10] so that rild and friends are still able to work [18:10] (rild is the bit caring for calls and completely closed) [18:11] slangasek, if you want you can install the cdimage image on a gnex ... manually [18:12] just not with phablet-flash [18:14] ogra: so, what would the manual process be? [18:15] ogra: first use phablet-flash to install the current raring image, then manually flash the bits from cdimage? [18:15] slangasek, well, you only need the raring bootstrapping ... if you have the partitions and recovery you are good [18:16] download the armhf.zip file ... download the armel+$yourdevice.zip file [18:16] ogra: ok - I guess my question was, is there anything else I need from jenkins for saucy support, or can I just use whatever phablet-flash has given me so far, then overlay the two images [18:16] boot the device into recovery [18:17] adb push /path/to/armel.zip /sdcard/autodeploy.zip && adb reboot recovery [18:17] once it rebooted, remove the battery (it triesd to boot into something nonexistent here ... ) [18:17] then repeat the adb dance with the armhf [18:18] on nexy boot you end up with a black screen and adb access [18:18] ok [18:18] apply that change in /usr/bin/ubuntu-session https://code.launchpad.net/~ogra/session-manager/fix-oom_adj/+merge/166975 [18:18] and on next boot you get the phone shell [18:19] (i would have uploaded that to saucy to fix it ... but i think we should first discuss the name ... i didtn really want to call it ubuntu-session) [18:20] slangasek, oh, well, the jenkins sucy images will just come out under ubuntu-touch-preview and phablet-flash will re-grow the -l (latest) argument if i understood sergiusens right [18:23] ogra: ubuntu-session wasn't supposed to go into saucy [18:23] well, we will have to start our session bits somehow [18:24] and nobody works on the upstart porting yet [18:24] so why not have it in saucy [18:24] (but probably as ubuntu-touch-session) [18:25] ogra: ack... do we have an item to move most of this to upstart? [18:26] ogra: I'm thinking we will have to have the PX env var picked up from somewhere sitll [18:26] not really, i have it on my whiteboard on the wall and wanted to attack it once the containers work on all devices [18:26] right [18:26] we will just use the user session support from upstart [18:27] and have the PX stuff in the phone shell job i guess [18:31] im asking when will be relased the day usable beta [18:33] rsalveti, FYI, the old kernel shows the same issue [18:34] i wonder if we miss anything that makes upstart work [18:34] ogra: well, good and bad then =\ [18:35] ogra, what about calls/sms on porting ? [18:35] yeah [18:36] ogra: what's broken with kernel/upstart? [18:36] stgraber, well, my init dies on the nexus7 [18:36] with the flipped container [18:36] works fine on the gnex [18:37] i'm slowly running out of ideas ... i see the mounts are fine but right after or at run-init i get the "kernel panic... attempting to kill init" and a reboot [18:38] hmm [18:38] i could try the quantal nexus7 desktop kernel ... havent tried that one yet [18:39] ogra: just a thought, from the initramfs, try running: init --user [18:39] i cant usi it interactively [18:39] hopefully that'll let you see where upstart fails to start without actually running it as PID1 [18:39] i dont think it even comes up [18:40] ogra: tries the desktop kernel, maybe we're missing a config or such [18:40] yeah, thats what i suspect [18:40] i havent heard from other devices yet though [18:40] i dont know if it isnt probably just working because it is a maguro, who knows [18:41] I can try on mako [18:41] that would be helpfu;l [18:42] http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ [18:42] grab the armel+mako.zip and the armhf.zip at the bottom [18:42] ogra: can't we get some sort of usb-serial with grouper? [18:42] flash the armel one first, then the armhf one [18:43] so we can try grabbing the trace log [18:43] mako should probably just work [18:43] rsalveti, yes, thats for tomorrow if i still didnt get anywhere [18:43] rsalveti, thats what i thought about grouper as well :P [18:43] well, but we got it to work with mako already [18:43] nobody tried grouper before :-) [18:47] err [18:47] Linux version 3.1.10-g06b7e9c (ubuntu@ps-android-precise-server-amd64-smp-2) (gcc version 4.6.x-google 20120106 (prerelease) (GCC) ) #1 SMP PREEMPT Thu Apr 18 08:36:34 UTC 2013 [18:47] * ogra scratches head [18:48] thats from the desktop bootimg ... [18:48] oh, hehe [18:49] extrachted the wrong one [18:52] rsalveti, hmm, ok, the raring desktop kernel seems to go further, but not much [18:53] hmm, same issue [18:54] it just blanks the screen during initramfs [18:55] ogra: flashed the new rootfs + armel+mako zip and then fastboot flash the bootimg-mako. My phone boots properly and the Android container is running. [18:55] phew [18:55] so its grouper [18:55] ogra: only problem I see is that I can't start the shell from the phablet user but that's another issue [18:56] http://paste.ubuntu.com/5730295/ if someone cares [18:56] stgraber, https://code.launchpad.net/~ogra/session-manager/fix-oom_adj/+merge/166975 [18:56] do that change and you get the shell on boot [18:58] (at least you should, it might be that i'm missing a device for mako in the lxc upstart job) [18:58] ogra: root@android:/ # initctl list | grep ubuntu-ses [18:58] ubuntu-session start/running, process 1280 [18:58] but no shell [18:59] did you apply the fix above ? [18:59] yep [18:59] oh [18:59] is surfaceflinger running ? [19:00] yep [19:00] hmm, then it is likely device permissions [19:01] ogra: yeah, it's. chmod 666 on /dev/fb0 and /dev/kgsl-3d0 did the trick [19:01] fb0 shouldnt be necessary [19:01] i'll add it to the lxc job though === dandrader is now known as dandrader|afk [19:02] ogra: yeah, not sure about fb0, I just did that one too as I remember seeing it on some strace a few days ago. I'm pretty sure we need kgsl-3d0 though and it was 600 root:root [19:02] * ogra would love to just ignore grouper ... but then i know that most of our devs have it [19:03] stgraber, could you try without fb0 ? [19:03] sure, 1sec [19:04] ogra: confirmed, only kgsl is needed [19:05] aweesome, will add it to the upstart job then [19:05] i guess in the end we will ship a gigantic udev rule for all these possible device :P [19:06] are there any other kind of udev rules? :) [19:07] ogra: so the android lxc container is already alive in the saucy images? [19:07] for some definition of alive, yes ;) [19:07] :) [19:07] but still no udev on ubuntu site? [19:09] there should be udev [19:09] ok [19:10] feel free to grab the image if you have a mako or maguro [19:10] so you create a symlink between the /dev on android site and the one on ubuntu site to share for example /dev/sockets? [19:10] ogra: will do [19:10] ogra: nope, it's still disabled here (/etc/init/udev.override) [19:10] maguro only here [19:10] stgraber, ah, we need to drop that [19:10] morphis: nope, you just let both ueventd and udev create the same devices each in its own namespace [19:10] right [19:11] morphis: as long as the type/major/minor match between the two /dev, you'll have access to the same device [19:11] you need to know how either side calls them to get the permissions in sync though [19:11] so no need to do symlinks/bind-mounts or other tricks for those we have uevents for, we only need to manually create those that do not emit uevents (if any) [19:11] stgraber: yes, I mean for example /dev/socket/rild which is created not by ueventd but by android init [19:11] i guess that and getting the platform-api fixed will be our big challenge with the flip [19:12] morphis: that will need to be done separately I'd guess [19:12] as this is needed for example for ofono to talk with rild [19:12] but yeah, we're not linking the paths there yet [19:12] rsalveti: ok [19:12] morphis: right, any manual mknod done in android will need to be replicated in Ubuntu, either through our own init scripts or through bind-mount of a whole directory if easier [19:13] atm I can only think of /dev/sockets which is needed [19:13] yeah, I believe so [19:13] maybe access to the propertyservice would be nice [19:14] in my tests last week I noticed some stuff trying to access /dev/alog, so I had that directory bind-mounted too. Not sure we actually need it though [19:15] (might be handy to have adbd fetch the logs I guess) [19:15] the lxc paxckage ships a udev rule already [19:15] for /dev/alog [19:15] ah cool, so we just need to make udev work then ;) [19:16] i wanted to move the other chmods from the upstart job too [19:16] yeah [19:16] that comes from the ubuntu-session package as well [19:16] anyway, I'm out for a while now, very late lunch break here, ttyl [19:16] i guess getting that into the archive should be our first step [19:16] yeah, i'll end my day soon === dandrader|afk is now known as dandrader [19:18] * ogra re-flashes his grouper from scratch [19:18] morphis: you said there's a socket for property as well [19:18] yes [19:18] morphis: we might want to hook that into hybris in some way [19:18] /dev/socket/propertyservice afaik [19:18] cool [19:18] rsalveti: yes, Stskeeps and me already had this idea too [19:19] but it is ro [19:19] oh, right [19:19] for rw we need th share memory area too [19:19] would help at least to probe the variables there [19:19] which we only get afaik if we're started by the android init daemon [19:19] yes === dpm is now known as dpm-afk [19:19] but for example for setting the usb mode it would be nice to have a setprop utility even in the glibc environment [19:20] indeed [19:20] root@ubuntu-phablet:/# lxc-info -n android [19:20] state: RUNNING [19:20] pid: 464 [19:20] root@ubuntu-phablet:/# ls /proc/464/root/dev/socket/ [19:20] adbd installd property_service rild rild-debug [19:20] ... [19:20] that should be easy to script [19:20] right, we'd only need to bind mount the socket dir there [19:20] yes [19:21] morphis, 90% of the getprop bits can be done with echo in sysfs [19:21] morphis: we need to think more about the property system as well, might make sense to hook that up in upstart itself somehow [19:21] ogra: yes, but it would be easier to have it once scripted on the android site [19:21] or provide it via scrips or such for others to use [19:22] ogra: as every device has it's own set of id's to set for the various gadget modes [19:22] for now we cant even enter the container [19:22] ok, thats more important :) [19:22] i mean you can chroot into the /proc node ... but that doesnt really get you a full set [19:23] rsalveti: for the first we need the ro only in libhybris so be sure we provide all needed properties to the libs we're accessing [19:23] and lxc-console kind of expects a getty [19:23] as atm it's only using /system/default.prop [19:23] ogra: yeah, getting getty into android :) [19:24] busybox would do [19:24] yes [19:24] i guess thats the better way ... since it can also provide all other missing commands in case we find any [19:25] morphis: right [19:25] rsalveti: I already have some code for this lurking around somewhere [19:25] hi [19:26] ubuntu touch in p3100 function 3g is call phone? [19:26] morphis: cool, can give it a shot later as well [19:26] ogra: I do have terminal with nexus 4 (using uart) [19:26] how can i control the icon size of a ListItem.Subtitled? [19:26] ogra: but didn't investigate how that's actually called [19:27] rsalveti, wrt getty you mean ? [19:27] ogra: yeah [19:27] well, i want lxc-console to work [19:28] so you can actually chroot [19:28] Any QML guys in? [19:28] ubuntu touch functionin the samsung tab2 p3100? [19:29] * snwh is away: Away [19:29] * snwh is back (gone 00:00:01) [19:29] !devices |cri [19:29] cri: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices [19:29] cri, see the wikipage [19:30] Cellular Radio [19:30] [19:30] Being worked on by Ubuntu devs. [19:30] -.- [19:42] is the keyboard a know problem on the galaxy nexus [19:42] mhall119, I want nice syntax highlighting for qtcreator for python.. I added the syntax file, but it doesn't seem to give me highlighting :-( [19:43] I get some basic highlighting already, wat file did you add? [19:45] autopilot [19:45] "I added the syntax file [19:45] " [19:45] rsalveti, [ 5.485081] Warning: unable to open an initial console. .... hmm, i wonder if thats my issue with grouper [19:45] was the file I was asking about [19:45] ahh, http://qihome.org/qiliang/2012/01/09/more-syntax-highlight-files-for-qt-creator/ [19:46] it doesnt fail there though [19:46] only later in run-init [19:46] ogra: what grouper issue are you referring to? [19:46] ogra: hm, wonder how it works with the desktop version [19:46] mhall119, grouper being the only arch not working in the new world order [19:46] mhall119, open a *.py file.. it should prompt you and ask if you want to get a syntax highlighter for it [19:46] mhall119, i.e. after the container flip [19:47] rsalveti, it uses fbcon there .... but the desktop kernel didnt work either here [19:48] balloons: doesn't prompt me, and I don't recall it ever doing that, but maybe it did a long time ago [19:48] mhall119, you have highlighting? [19:48] ogra: that's weird [19:48] yup [19:48] it's under options;text editor;generic highlighter [19:49] balloons: yup [19:49] rsalveti, and i tried all i could ... i'll try a kernel with usb serial tomorrow [19:49] but even that will only tell me the same [19:50] mhall119, kk.. it just hate sm [19:50] balloons: ah, highlighting is provided by Kate, maybe I had installed a python syntax on that a long, long time ago [19:50] for some reason upstart doesnt get along with that arch [19:51] ogra: well, you might be able to get more from your kernel there [19:51] ogra: would be nice to grab the kernel trace at least [19:51] well, i have the oops here [19:51] but that wont help much i guess [19:53] http://paste.ubuntu.com/5730485/ is dmesg if anyone is intrested [19:53] (most of it at least) [19:56] mhall119: do you know if its possible to control the icon size on a ListItem.Subtitled? [19:56] Dubstar_04: yes, it is, let me find you an example [19:57] great. I have made my own delegate that replicates the ubuntu.components but its a waste really if i can use the standard components [19:58] Dubstar_04: http://mhall119.com/2013/04/building-an-ubuntu-sdk-app-rev-3/ scroll down to the "Visual Tweaks" header [20:04] sergiusens, http://cdimage.ubuntu.com/ubuntu-touch-preview/raring/monthly-05/ for you with flowers :) [20:04] pmcgowan_, ^^^ [20:05] w00t [20:05] ogra: could care less, I'm on saucy ;-) [20:05] lol [20:05] thanks :-d [20:05] thats so last week [20:05] haha [20:05] mhall: thanks for the link, it doesnt seem to work for me. I am trying to display non-square images from movie posters if i use __iconHeight: units.gu(10) [20:05] __iconWidth: units.gu(5), it still shows a square image or scales the images based on the amount of text in the subtitle [20:06] yeah, saucy is today, but you can even live in the future .... http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/has the flipped container images :) [20:06] ogra rock on [20:06] Dubstar_04: ah, it might be because the image is being put into an UbuntuShape, which might always be squared [20:06] kgunn, not working on nexus7 though, i hope you guys dont rely on exactly that arch [20:07] mhall119: so in this case its best to use a custom delegate? [20:07] ogra no worries...we got a variety of devices on team [20:07] good [20:08] nexus4 is missing a fix i just uploaded ... galaxy nexus works but you need https://code.launchpad.net/~ogra/session-manager/fix-oom_adj/+merge/166975 to make the phone shlle come up [20:08] *shell [20:11] Dubstar_04: might be, yeah [20:11] mhall119: this is the result: https://lh6.googleusercontent.com/-6Xlm-UCCJBw/Uaz4M4RVurI/AAAAAAAAHbw/aalMH2Qg_QE/w681-h551-no/Screenshot+from+2013-06-03+21%253A09%253A17.png [20:12] heres the code: http://paste.ubuntu.com/5730527/ === dandrader_ is now known as dandrader [20:18] Dubstar_04: that first one looks like it has different height and width [20:19] mhall119: the icon dimensions seem to be based on the amount of text in the subtitle [20:21] narf, rsalveti: I was wrong, /dev/socket/property_service is w only and not r [20:23] morphis: oh [20:24] morphis: let me look at the init code [20:25] rsalveti: https://code.google.com/p/android-source-browsing/source/browse/libc/bionic/system_properties.c?spec=svn.platform--bionic.0d787c1fa18c6a1f29ef9840e28a68cf077be1de&repo=platform--bionic&r=0d787c1fa18c6a1f29ef9840e28a68cf077be1de#257 [20:26] part of bionic? wtf [20:26] yes [20:26] thats the client site [20:27] rsalveti: service site is here: https://android.googlesource.com/platform/system/core/+/master/init/property_service.c [20:28] cool [20:31] ogra: ok, so I followed your instructions and on mako, I have ubuntu-session running but I still have a black screen; any hints? [20:31] mhall119, thank you sooooo much for getting canonical-auth in the ppa... it's working great on my n7 :) [20:32] kenvandine: thanks aquarius too, there was very little I had to do [20:33] ogra: as for /proc/$$/oom_adj not being writable, there's a rather suspicious disconnect between the pids shown in 'ps' output, and the directories I find under /proc [20:33] mhall119, i had started a QML UI for that and was going to just snag the .js files aquarius had for it [20:33] but never finished it up [20:33] now i never will... since we have something that works :) [20:33] kenvandine: it was a good use-case for shipping a native-looking HTML5 aqpp [20:33] indeed [20:34] and it does look native [20:34] with the css applied, yeah, and we'll see more of that in the future [20:35] mhall119: This is the custom delegate: https://lh6.googleusercontent.com/-vM5HRKv0ips/Uaz9gCACHAI/AAAAAAAAHcE/anmYpwNFef8/w441-h551-no/Screenshot+from+2013-06-03+21%253A32%253A38.png [20:40] Dubstar_04: looks much better that way [20:40] mhall119: its a shame the ubuntu.components aren't more flexible [20:41] Dubstar_04: we're working on that, feel free to file bugs against ubuntu-ui-toolkit on Launchpad when you run into walls like this [20:45] rsalveti: https://github.com/morphis/libhybris/commit/d52cb4ea040b8ba4b735fcb164958b52d7edadb0 [20:47] mhall119: bug filed. [20:47] rsalveti: that is what I wrote some time ago [20:48] let me check [20:49] kenvandine: the new powerd makes the power button responsive, but it still doesn't look like it completely turns of the LCD's backlight [20:49] it has a very faint, but noticable amount of glow [20:49] morphis: that's nice [20:49] just cleaned it up but did not tested it yet [20:49] that isn't related to my fix :) [20:49] kenvandine: is it a known issue? [20:49] we'd just need something to get the properties [20:49] my fix just kept it from crashing if it didn't have a proximity sensor [20:50] no idea [20:50] ok, thanks just the same [20:50] mhall119, hmmm... mine seems to be cycling [20:50] like the display is off [20:50] morphis: looks good, can give that a try later today [20:50] and every few seconds it gets a little brighter [20:50] then darker again [20:50] not sure if that is new behavior or not [20:51] it's new for me [20:51] yet another bug :) [20:51] though I recall it happening when I used an older version of powerd from bzr [20:51] I assumed it was related [20:51] but now I'm thinking otherwise [20:52] I wonder if it's the new grouper Android bits [20:52] because it didn't happen on image build <=140 [20:52] perhaps [20:52] i guess file a bug for that [20:52] now that i've seen it... it's rather annoying :) [20:53] kenvandine: any idea how I can get some useful information to add to the bug? [20:54] mhall119, nope... sorry [20:54] i don't know much about powerd [20:54] ok, I'll file it and then maybe someone will tell me what extra info to gather [20:54] just the fix for the crash was pretty obvious once i started bisecting it [20:55] kenvandine: can you attach your branch to https://bugs.launchpad.net/powerd/+bug/1186469 [20:55] Launchpad bug 1186469 in powerd "segfault on nexus 7" [Undecided,Fix committed] [20:55] I've updated it to Fix Committed already [20:56] mhall119, oh... there was another bug filed [20:58] mhall119, bug 1186706 [20:58] bug 1186706 in touch-preview-images "Cannot suspend by pressed power button" [Critical,Fix committed] https://launchpad.net/bugs/1186706 [20:58] i duped your's [20:58] oh? Launchapd is usually good about telling me when I'm filing a duplicate [20:58] different project [20:58] ChickenCutlass: https://bugs.launchpad.net/powerd/+bug/1187135 is my current problem [20:58] Launchpad bug 1187135 in powerd "Screen doesn't fully turn off on Nexus 7" [Undecided,New] [20:58] and your's was first [20:58] kenvandine: ^^ if you want to mark yourself as affected [20:58] but i had referrenced that one in my branch [20:58] sure [21:23] stgraber: which device have you been doing testing on wrt the container flip? [21:23] slangasek: nexus4 (mako) [21:23] stgraber: ok. and it's working for you? You get video out? [21:23] slangasek: ogra is having kernel/initrd problems with the nexus7 [21:24] cat /home/phablet/.ubuntu-session/logs/maliit-server.log [21:24] WARNING: Cant find EGLConfig, returning null config [21:24] CRITICAL: loading "libmaliit-keyboard-plugin.so" [21:24] slangasek: no, there's one small problem with it at the moment. [21:24] slangasek: chmod 666 /dev/kgsl-3d0 [21:24] slangasek: restart ubuntu-session [21:24] ah [21:24] right, that's the thing you mentioned earlier [21:24] ogra fixed that in the archive an hour or so ago, so that's going to be in the next image [21:24] ok, cool [21:25] sweet, there it is [21:25] thank s:) [21:25] does it seem like it is taking ever longer to enable "developer mode"? [21:25] slangasek: note that at the moment it's not possible to start any app, but at least getting the shell running should make it easier for the right people to debug that [21:26] rickspencer3: I think it's installing more developer tools than it used to [21:26] used to be it justed installed openssh, I think [21:41] stgraber: I also don't see any wireless networks in the UI... makes it more difficult than anticipated to apt-get debugging tools on ;) [21:42] slangasek: ah yeah, that part is broken too, just use ifupdown [21:42] I don't want to use ifupdown to configure my wpa wifi ;) [21:42] * slangasek implements apt-adb instead, because of course that's simpler [21:42] ;) [21:42] slangasek: actually, maybe it's just a matter of starting network-manager [21:43] hmm, any idea why that didn't auto-start? [21:43] still no network list, though [21:43] I believe it might be started [21:43] it wasn't here [21:43] but I know android has some specifics for wlan drivers and such [21:43] slangasek: well, udev is currently blacklisted, so I expect some other bits to be broken as a result of that [21:43] weird [21:43] oh, right [21:43] oh [21:43] we need to remove that [21:44] what problems can I expect if I unblacklist udev? [21:44] because I would expect that to be a fairly important prereq for container flip [21:44] but that would break the traditional image (without container flip) [21:44] right [21:44] starting it manually kills my adb connection. Haven't tried booting the phone with it enabled though, maybe that works [21:44] * stgraber tries [21:44] heh, didn't just kill my connection here, it rebooted my device [21:45] stuck in a reboot loop here... [21:45] yep [21:46] right, so time to re-create the file from recovery :) [21:46] maybe I shouldn't have removed the .override for testing :D [21:47] what happened in builds 148 and 149? [21:47] slangasek: just do "echo manual > /data/ubuntu/etc/init/udev.override" over adb with the phone booted in recovery [21:47] stgraber: sure [21:48] sergiusens: are the terminal and file manager going to make it into raring images before we switch to saucy? [21:48] mhall119: I don't think we started such builds [21:48] ....if I manage to get it out of the reboot loop [21:48] sergiusens: should we include that? [21:48] mhall119: they could ... they shouldn't from what we are supposed to be focusing on [21:49] rsalveti: I guess so [21:49] slangasek: right, so spawning udevd itself isn't the problem, I suspect it's the following udevadm trigger that breaks the world [21:49] would be useful for 3g data [21:49] slangasek: trying now with udevtrigger disabled [21:51] sergiusens: did we land the fix nexus 7 related powerd fix? [21:51] actually, did we find what was the issue there? [21:51] slangasek: right, booting with udev enabled, udevadm disabled, then running "udevadm trigger --action=add --subsystem-match=net" and finally starting network-manager gives me a working list of network [21:51] rsalveti: the segfaulting bug is fixed in lp:powerd [21:51] don't know if mfisch knows as well [21:52] mhall119: awesome [21:52] sergiusens: let's do a release of that as well [21:52] guess you're already taking care of that [21:52] rsalveti: it's released (powerd) [21:52] I didnt know we had a segfault [21:52] build 151 will have it [21:52] sergiusens: great [21:52] rsalveti: it was sensors related [21:52] mfisch: I pinged you about it over the weekend [21:52] stgraber: and, any idea what bit is causing the poweroff? [21:53] slangasek: nope, I'm assuming one of our udev rule is somehow causing a system crash, would have to go subsystem by subsystem until we find which [21:53] I was wondering if tailing the udev log over adb while running the trigger would tell us anything useful [21:54] sergiusens: rsalveti: 2 our of 3 reboots on my nexus 7, apps won't display and the OSK won't come up [21:54] what can I do to trace the problem down to file a bug? [21:54] mhall119: the osk one is quite old actually [21:54] the apps might be new [21:54] rsalveti: well they seem to go together [21:54] if I don't have OSK, I won't see the apps either [21:54] ricmm might be able to help you on that [21:55] ricmm: you got a 7 as well, right? [21:55] I get a white screen, and if I start to swipe-from-left to switch it'll suddenly show me the app's content for a few seconds, then it goes black [21:55] mhall119: sorry about that, I've been spending a lot of time outdoors on weekends [21:55] it might be platform-api/ubuntu-session related [21:55] mhall119: officially I'm working on something else, but I'll help when I can [21:55] when I switch back to the apps lens, the running app thumbnail for it disappears [21:55] mfisch: kenvandine fixed the segfaulting [21:56] cool [21:56] rsalveti: reading [21:56] ricmm: let me know what I can do to track down the cause [21:56] sforshee: ^^^^ - there was a segfault and kenvandine fixed it, FYI [21:56] stgraber: last events before reboot (from udevadm monitor -e): http://paste.ubuntu.com/5730840/ [21:57] ricmm: I'm currently in a boot that is experiencing this, so if there's something I can try I will [21:57] mhall119: lol, that wasn't my code ;) [21:57] * mfisch eyeballs the sensor guys suspiciously [21:57] I'm the sensors guy [21:57] slangasek: odd, subsystem-match=firmware doesn't cause the reboot here [21:57] ricmm: ;) [21:57] stgraber: right, second attempt comes up with something completely different [21:57] ken hid the segfault and now im pushing something into the platform api to guard non existant sensors better [21:58] slangasek: I'm going through /sys/class trying to figure which one [21:58] stgraber: ok [21:58] ricmm: thats probably a better fix. was the fault platform specific? [21:58] stgraber: seems like a better approach [22:01] mfisch: the problem is the untested return of the proximity sensor getter [22:01] but the other helper functions would benefit from a guard in the platform API [22:01] if the client forgets to check the pointer [22:02] ricmm: got it. sforshee is out today and part of tomorrow IIRC, but I'll bring him up to speed when he's back [22:02] ricmm: I can do a review for you if needed [22:03] slangasek: confirmed that if I go through all of them I get the crash, now running a for loop which tries each and wait 5s, let's see where it reboots [22:03] mhall119: do you know of other people experiencing this issue? [22:07] slangasek: I'm affraid we've got some kind of race... [22:07] bfiller, hey [22:07] are you familiar with a bug where recieving a call has no audio? [22:07] jono: wassup [22:07] the call connects but I can't hear the other person [22:08] slangasek: "for dir in *; do echo $dir; udevadm trigger --subsystem-match=$dir; sleep 5; done" worked fine here [22:08] jono: haven't seen that one. have seen the one where you don't hear the ringtone but see the notification and can't answer a call [22:08] slangasek: "for dir in *; do echo $dir; udevadm trigger --subsystem-match=$dir; done" makes my phone reboot (when done after boot, doing it after the delayed one is fine apparently) [22:08] bfiller, I get the tone but when I answer, nothing [22:08] jono: which hardware? [22:08] bfiller, happened three times today for me [22:08] bfiller, Galaxy Nexus [22:09] jono: can the person on the other end hear you? [22:09] bfiller, not sure [22:09] will test with skype [22:10] one sec [22:10] bfiller, I used to get that but not for some time, actually may be a nexus specific one [22:10] pmcgowan_: haven't had the one happen to me yet [22:10] bfiller, back when I had a nexus the wrong numbers used to go that way [22:10] cant recall it on the nexus 4 [22:10] bfiller, just tested now via skype and it worked [22:10] odd [22:11] jono, was the phone sleeping when the calls came in [22:11] pmcgowan_, yep, screen was off [22:11] jono: mind changing the parameters for ofono in your /etc/init/ofono.conf? [22:11] exec ofonod --noplugin=atmodem, to exec ofonod -d --noplugin=atmodem [22:11] so we can have debug [22:11] bfiller, evolution not automatically picking up my backup, probably need to kick it to know that addressbook exists [22:11] rsalveti, one sec, grabbing my USB cord [22:12] then use it a bit more, and once you reach the problem, please save your /var/log/syslog somewhere [22:12] pmcgowan_: you copied the stuff back into .local/share/evolution? [22:14] sergiusens: happroved https://code.launchpad.net/~saviq/unity/phablet.raring-backport-726/+merge/167085 [22:15] rsalveti, I can't seem to edit text in my shell [22:16] in either nano or vi [22:16] jono: weird, using it via adb? [22:16] rsalveti, [22:16] vi should work [22:16] yeah [22:16] trying in regular terminal and not terminator [22:16] mind trying via ssh? [22:16] might a weird VT mode [22:16] yeah [22:16] adb sets up some console stuff [22:17] rsalveti, yeah doesnt work [22:17] rsalveti, neither vi nor nano let me save the file [22:17] but no perm denied [22:18] jono: weird, works fine here [22:18] ssh should work better [22:18] rsalveti, just so I am clear, I am doing this: [22:18] adb root [22:18] adb shell [22:18] ubuntu_chroot shell [22:19] nano /etc/init/ofono.conf [22:19] right, nano doesn't work as expected [22:19] but vi does [22:21] rsalveti, how do I enter INSERT mode in vi? [22:21] pressing 'i' doesnt work [22:21] jono: try insert [22:22] rsalveti, no luck [22:22] jono: weird, try using vi via ssh (or just install vim, it should behave better) [22:23] jono: sed -i 's/ofonod/ofonod -d/' /etc/init/ofono.conf [22:23] that as well [22:23] stgraber, wise :-) [22:23] but check with vim as well [22:23] rsalveti, ok changed with sed [22:23] wonder if that would work [22:23] it works fine here [22:23] do I restart the service? [22:24] jono: better restarting the phone [22:24] you won't be able to restart the service from adb as you're not in the right pid namespace, you could from ssh or just reboot the phone [22:24] and the app might not necessarily work after restarting ofono [22:24] if you've got ssh access, you should be able to mount a remote directory (in this case, on the phone) on your laptop/desktop via Nautilus over ssh protocol, targeting the phone. Then just edit it with gedit/something else. Right? Just an idea. Maybe it wouldn't work. [22:25] rsalveti, stgraber ok rebooting the device [22:25] rsalveti, now what do I do when the bug occurs? [22:25] is there a log somewhere? [22:25] jono: save your /var/log/syslog somewhere, and email it to me [22:26] rsalveti, ok [22:26] hopefully I will see the bug occur again soon [22:26] cool, that should tell what ofono is doing, and if it's indeed setting up the audio route correctly [22:26] rsalveti: I dont get why grouper would fail to register started apps [22:26] mm [22:27] ricmm: no idea either [22:27] ricmm: but it happens quite often actually === SkavenXXI-[OFF] is now known as SkavenXXI [22:27] happens every time for me [22:28] ricmm: I haven't heard of anybody else with this problem, no [22:28] ricmm: even better :-) [22:28] maybe only mhall119 has flashed grouper with 147 :) [22:29] :-) [22:29] what was the issue with powerd that was nexus 7 specific? [22:29] * rsalveti branches trunk [22:29] nonexistant proximity sensor [22:29] no guards for it [22:29] oh, ok [22:29] so not related at all [22:29] no [22:30] stgraber: gah :/ [22:30] stgraber: how about a bisect, running the for loop for only half the directories? [22:34] ricmm: it's been doing it to me for a while [22:35] slangasek: testing [22:35] since before 140 at least [22:40] slangasek: making some progress, down to just a few possible subsystems [22:41] mhall119: thats a long while [22:43] slangasek: down to 2 [22:43] stgraber: which two? [22:43] slangasek: udc and video4linux [22:43] hmm, what's udc? [22:44] yeah, I would always tell myself "I need to file a bug about this, but first let me reboot until my tablet is usable again", then 5 minutes later "Sweet, it's working again, time to go on Reddit" and bug filing never happens :( [22:44] slangasek: hmm, actually, the issue is video4linux [22:44] slangasek: it's just that apparently depending on what you trigger before it, it doesn't make the phone reboot [22:45] slangasek: if you just do a clean boot, start udev and do "udevadm trigger --subsystem-match=video4linux" you'll get the reboot [22:45] speaking of N7 stuff not working, is there any work being done to get the camera and audio to work? [22:45] mhall119: camera not sure, audio it seems to be missing a default mixer config [22:45] diwic was looking into it [22:46] oh is that all? I thought it was going to be some driver thing [22:47] Hi, did anyone manage to get the software in a HTC Desire? [22:48] !devices | abonilla [22:48] abonilla: You can find the full list of devices, official images, community images, and works in progress at https://wiki.ubuntu.com/Touch/Devices [22:48] if it's on that list, somebody is going something with it, if it's not, then you get to be first :) [22:49] slangasek: removing 60-persistent-v4l.rules appears to prevent the reboot [22:49] mhall119: the instructions are there but not clear. ie. does not say which image/files. [22:50] slangasek: which is weird considering that all it does is create a bunch of symlinks [22:50] stgraber: that's not all it does, it also calls v4l_id [22:51] stgraber: /lib/udev/v4l_id /dev/video39 > triggered an instant reboot here [22:52] slangasek: ok, so it's some kind of kernel/hardware bug then which triggers the reboot when accessing some properties [22:53] stgraber: seems so. can you file a bug on the kernel for this? [22:53] hmm, and it is indeed not 100% reproducible... this time, v4l_id succeeds on all the nodes [22:54] slangasek: yep, I'll file a bug, sadly can't provide a whole bunch of debug info [22:57] stgraber: slangasek: is that mako specific? [22:57] or can you also reproduce with maguro/grouper/manta? [22:58] rsalveti: can't boot grouper at this point, so hard to tell [22:58] and I don't have grouper or manta [22:59] right, ok [22:59] but it's probably mako specific [23:03] it probably is, yes [23:03] so that's odd, basically opening the file after starting udev + udevtrigger causes the reboot. Doing so without udev is fine [23:03] I could check grouper later, but that requires getting it past the non-booting point where others have already gotten stuck [23:04] so, how should we work around this? the rootfs is device-independent, right? [23:04] I got a grouper on which I'm testing some stuff right now [23:04] rsalveti: what do you need to try? [23:05] if there's a generic ubuntu-touch package, we could put a dpkg-divert in place on the .rules to keep them from being run; that will affect all devices, but that's probably an ok stopgap if it lets us get to the point of turning udev on [23:05] ricmm: step 1) run udev [23:05] abonilla: did you go to https://wiki.ubuntu.com/Touch/Devices/bravo ? [23:05] it has a link to an XDA forum thread that has the images [23:08] slangasek: bug 1187189 [23:08] bug 1187189 in linux-mako (Ubuntu) "Kernel crash and reboot when accessing video device" [Undecided,New] https://launchpad.net/bugs/1187189 [23:08] stgraber: great, thanks [23:09] slangasek: no step2? :) [23:09] ricmm: step 2) run 'udevadm trigger' [23:10] ricmm: (with the full set of Ubuntu udev rules - and in particular, /lib/udev/rules.d/60-persistent-v4l.rules) [23:13] mhall119: I did but mentions to download an Desire image and the site has many different images... I'll see.. [23:14] abonilla: oh, then you'll need to get clarification from the XDA developer working on it [23:20] mhall119: is Ubuntu Touch aka "Ubuntu Phone"? [23:31] stgraber: so, I've enabled udevtrigger now after diverting the v4l rules; it no longer crashes, but I also no longer get any GUI [23:32] slangasek: I wonder if the udevadmtrigger is maybe resetting some permissions that ogra hardcoded in his scripts [23:32] stgraber: possibly. Any idea which scripts? [23:33] slangasek: try: chmod 666 /dev/pvrsrvkm /dev/tiler /dev/binder /dev/ion /dev/kgsl-3d0 [23:33] slangasek: from /etc/init/lxc-android-config.conf [23:34] ok... as soon as the device comes back up again (what's taking it so long?) I will [23:35] abonilla: it's the software for phone and tablet, yes [23:38] stgraber: well, heh; trying to stop and restart the lxc container also caused a phone reboot, this time with the v4l rules diverted [23:39] slangasek: yeah, I noticed that before, I assumed that some device didn't like seeing its firmware loaded twice or something along those lines [23:47] OK, HTC Desire. FAIL. === chriadam|away is now known as chriadam