[00:00] in which ppa is that done? [00:00] ogra_: asac the ppas we use are all real [00:00] ok [00:00] so now i wonder where the jenkins uploads go to [00:00] or how that things tests builds [00:00] (if not in a ppa) [00:00] asac: jenkins has pandas hooked up to it [00:01] ok. how can i find the right job [00:01] for a build attempt? [00:01] * popey stabs poxy proxies [00:01] Saviq: lol, seems to be almost impossible to land a mr against unity8 [00:01] /usr/include/c++/4.8/type_traits:98: confused by earlier errors, bailing out [00:01] The bug is not reproducible, so it is likely a hardware or OS problem. [00:01] asac, ^^^ [00:01] ogra_: sure, but where is it? [00:01] where can i see this in the infra [00:01] dunno [00:01] or was that a local result? [00:02] nope [00:02] asac: if it's a merge, jenkins votes upon the merge with links to the logs [00:02] it usually sends you a mail with the log link if a build fails [00:02] ogra_: so seems there are manual uploads [00:02] asac: if it's daily release, it's a manual check [00:02] https://code.launchpad.net/~unity-team/unity/trunk [00:02] the last commit [00:03] merges such an upload [00:03] i dont see any merge request though [00:04] https://code.launchpad.net/~phablet-team/unity/unity8-mir [00:04] guess that must be it? [00:04] asac: the daily release ones don't have an easy way to check upon but manually looking at https://jenkins.qa.ubuntu.com/view/cu2d/view/Head/ [00:05] https://code.launchpad.net/~saviq/unity8/fix-network-page/+merge/174828 [00:05] thats the related merge [00:05] ok so they crewated a new project calle unity8 [00:05] tricky [00:05] would have thought it would be in unity :) [00:06] https://jenkins.qa.ubuntu.com/job/unity8-saucy-armhf-autolanding/87/console [00:06] thats whats linked from there [00:06] asac, it was in lp:unity/8.0, but it's unmanageable in LP to keep two different codebases under the same project [00:06] Caused by: java.io.IOException [00:06] yeah, as usual, when stuff's important, everything fails on jenkins [00:07] Saviq: can one just request a new review by the bot to get him rekick a job? [00:07] :) [00:07] asac, http://s-jenkins:8080/job/unity8-autolanding/88/console [00:07] asac, it's already running [00:07] ah [00:07] but is that the approach? [00:07] i dont have s-jenkins mapped [00:07] asac, you just re-approve [00:07] what ip is that? [00:07] 97? [00:07] 10.97.2.10 [00:07] or 98? [00:07] asac: 10.97.2.10 [00:08] oh [00:08] a new :) [00:08] i had two other [00:08] you probably had the ones without java failures :) [00:08] this one has them builtin it seems [00:09] * ogra_ heads bedwards [00:09] Saviq: ok so re-request and then approve triggesr jenkins to come back [00:09] sounds good [00:09] ogra_: n8 [00:09] ogra_: 15.2 still doesn't do wifi [00:10] check the dashboard tomorrow :) [00:10] lol [00:10] you of course knew this and I am late to the party [00:10] popey, did you read what i wrote above ? :) [00:10] no ☻ [00:10] go to bed ☻ [00:10] * popey hugs ogra_ [00:10] popey, there is one in pending, but it has the wlan issue still [00:10] :) [00:11] \o/ [00:11] * popey gets the bed warm for ogra_ [00:11] lol [00:11] susie wont like that [00:11] hmm, or perhaps she would ... [00:11] haha [00:11] who knows [00:11] n [00:11] n [00:11] * ogra_ & [00:15] yeah so we have different jenkins versions it seems... the one we hit hear seems to be 1.424.6 while 10.97.0.1:8080 is already rocking the 1.480.3 [00:15] popey: so what was the issue with phablet-flash? [00:16] asac: well I know of no reason why the former wasn't upgraded [00:16] let me guess [00:17] a) we have custom developed plugins that need to be ported [00:17] I am confused about the desktop 'image' that was once pitched as Ubuntu for android and is now being offered on the OEM and Operator page of the Ubuntu phone. . Is it ever going to be part of a 'regular' phone distro? is it to be concidered proprietary? [00:17] b) we rely on plugins that are not avialable for the new one [00:17] c) we found other issues [00:17] d) we veoted and upgrade and never looked back :) [00:18] sergiusens: ? [00:18] e)? === chriadam|away is now known as chriadam [00:19] asac: might be the denial to shutdown at the time the .1 server was upgraded [00:19] asac: I know there's planned downtime at the end of July [00:19] so d? [00:20] yes [00:21] anyone ? [00:22] matv1: which image is that? [00:24] matv1: Ubuntu for Android is not really the topic of discussion here, but in essence it is operator driven, I don't think you will find it on cdimage if that's what you were looking for [00:25] robert_ancell, how do I enable mir/unity mode in lightdm? [00:25] I know. But isnt the docked-device that runs as full desktop that is now offered to OEMS the same thing as the old Ub for Android [00:26] I mean halfway down this page [00:26] http://www.ubuntu.com/phone/operators-and-oems [00:26] matv1: well, the elements would still look look roughly the same if you refer to http://assets.ubuntu.com/sites/ubuntu/503/u/img/devices/converged-device-440x267.jpg [00:26] matv1: there is no Android as you know it in Ubuntu Touch [00:27] the target is pure Ubuntu convergence [00:27] ogra_: do you have a way to check if the builders can reach http://people.canonical.com/~sergiusens/click_packages/ ? [00:27] the image builders that is [00:27] if they are different [00:27] yea but I am guessing the way to lauch the desktop is some kind of virtualization [00:28] not with ubuntu touch [00:28] dont know exactly how ubuntu android offering does that [00:28] but the experience looks pretty much the same when docked as both solutions offer a full ubuntu desktop goodness [00:28] for users [00:29] I see. So in touch, you are actually using the phone OS itself? [00:30] matv1: not sure what that tries to say [00:30] as oposed to the android solution where I am sure virtualization was involved [00:31] i have no clue about that as i said :) [00:31] ubuntu phone that we work on here doesnt use virtualization afaics :) [00:31] okay [00:32] not saying that we will never use some or all the convepts of virtualization in some form to achieve something of course :), but the kernel boots and we start stuff pretty much "normally" here [00:32] sergiusens: not sure, got it flashed in the end [00:34] Saviq: doesnt look so good [00:34] #88 that is [00:34] armhf buld fell over [00:35] maybe panda-1 is down? [00:35] can you force on a different executioner? [00:35] do you have any rights? [00:37] asac, no, I don't [00:37] fginther, help [00:37] i really think that device might be busted [00:37] failed abunch of stuff in a row [00:38] yeah [00:38] everything fails a bunch of stuff... [00:38] http://s-jenkins:8080/computer/ps-panda-4/builds [00:38] do you guys ever succeed any build? [00:39] asac, not tonight it seems... [00:39] ps-panda-5 seems to be an option [00:39] it succeeded builds during the last 3 hours [00:39] ps-panda-7 [00:39] is also strong [00:40] very strong :) [00:40] Successunity8-saucy-armhf-autolanding #85 [00:40] even succeeded to do that 4h 5min [00:42] Saviq: is fginther the rigth guy? [00:42] asac, yeah, he's on dinner, though [00:43] only option? [00:43] asac: this MR has no code https://code.launchpad.net/~asac/phablet-tools/check-device-ip-state-after-sleep/+merge/174889 ... going to check if it's just a launchpad issue [00:43] asac, should be back after, and no, we can harass someone else in #ps-qa on i.c.c [00:43] even though they've bounced me towards fginther for now [00:43] asac: Saviq I have access [00:43] thomi: veebers: do you know about the s-jenkins? [00:43] can you help us direct a job to a certain panda? [00:44] sergiusens, ps-panda-1 needs to go away http://s-jenkins:8080/computer/ps-panda-1/builds [00:44] asac: yes, Saviq was just asking us that [00:44] cool [00:44] so i am only in #qa [00:44] and #ubuntu-quality [00:44] can we shut down #ps-qa :) [00:44] i really have enough lists [00:44] err channels [00:44] the group doesn't exist anymore [00:45] sergiusens: ummm... excuse me? [00:45] sergiusens: is there a way you can reboot panda's ? [00:45] sergiusens: we still exist :) [00:45] thomi: Product Strategy doesn't exist anymore [00:45] well, the name might be under discussion [00:45] robert_ancell, poke [00:45] we still call outselves PSQA, I guess the name stuck :) [00:45] the folks are certainly there and just operate in a meaningful vacuum :) [00:46] thomi: so what can we do about all those pandas? [00:46] can we do a complete PDU pipeclean reboot of everything and reflash :)? or just turn off and have someone tomorrow check? [00:46] asac: sergiusens can help you now, or fginter can help you when he gets back after dinner [00:46] thomi: asac Saviq I marked panda-1 as offline, no builds should go there [00:46] sergiusens: http://pastebin.com/y9XWCZ7C [00:47] sergiusens: panda-4 is also bogus [00:47] at least risky :) [00:47] asac: if it's something that requires physical intervention you can contact rfowler or retoaded, or file an RT [00:47] sergiusens: can we somehow route a job to an explict board? [00:47] sergiusens: i assume that's device specific stuff that should be removed? [00:47] i would really like to run this exactly on panda-7 [00:47] Is there a changelog available for the new flipped daily builds? [00:47] asac: if I shut that one down we will have only one panda [00:48] sergiusens: but you cant bring them back then, do you? [00:48] asac: do you know what's up with the calxeda servers? [00:48] asac: yes I can [00:48] coming, but not avaialble :) [00:48] so not an option today [00:48] sergiusens: can you really pipeclean some [00:48] ? [00:48] like reflashing from scratch, checking whats up with them? [00:48] (just wondering) [00:49] how sure? if you are sure you can bring panda-4 back you could disable it [00:49] cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ no longer contains saucy-preinstalled.changelog & ubuntu_stamp like cdimage.ubuntu.com/ubuntu-touch-preview did [00:49] start the unity8 [00:49] so it goes to 7 and then bring it back [00:49] asac: I already disabled panda-4 [00:49] :) [00:50] panda-5 is still there, no? [00:50] 5 and 7 [00:50] Q: is that because apt-get dist-upgrade is now able to update all system components, to include kernel and lxc android adaptations, etc. ? [00:51] asac: but if you want it to run on panda-7 I have an easier solution [00:51] because https://wiki.ubuntu.com/Touch/ReleaseNotes#Upgrades still states that some parts cannto be updated [00:51] asac: I just restricted unity8 to panda-7 [00:51] cool [00:51] when will it start? [00:51] what is your guess> [00:52] asac: well bamf is currently building on it now, I can retrigger the land [00:52] asac: do you have the MR handy? [00:52] this one? https://code.launchpad.net/~saviq/unity8/fix-network-page/+merge/174828 [00:53] i think it was about http://s-jenkins:8080/job/unity8-autolanding/88/console [00:53] i think thats the one [00:53] https://code.launchpad.net/~saviq/unity8/fix-network-page/+merge/174828 [00:53] that one refers to 88 [00:53] so i guess thats correct [00:56] asac: so now you have it waiting on panda-7 http://10.97.2.10:8080/job/unity8-saucy-armhf-autolanding/ [00:57] thanks sergiusens [00:57] asac: I would just fasttrack this to be honest unless you want to continue with this experiment [00:57] hmm it says "all are offline" [00:57] hope its normal [00:57] asac: there's only one [00:57] let's have those restarted or something [00:58] asac: :-/ I got the label wrong, fixing.. [00:58] sergiusens: whats fasttracking? [00:58] upload straight to archive? [00:59] asac: sergiusens: unity8 fix-network-page should get landing soonish [00:59] asac: this is my gripe with upstream merger, ci builds and runs tests and votes. When you Happrove, you have to run the same thing again. I would argue to only do that if there were merges in between [01:00] asac: fasttracking would be to skip the full package build and tests (which already ran) and have the rest of the process complete [01:00] sergiusens: after approve it goes to ppa first [01:00] ? [01:00] e.g. after build approving? [01:00] asac: nope [01:00] asac: that's daily release... want to set somethig up for tomorrow? I'll walk you through everything [01:00] so fasttracking means: upload to the ppa the image picks up without testing build? [01:01] sergiusens: well. do whatever it needs us to do to get a new image :) [01:02] sergiusens: i think all my issue went away by running make multiple times. now it's complaining about missing blobs (but i cant get adb to work on virtual box and windows) [01:02] what i tried to understand is what you plan to do ... just upload to archive (e.g. through an untested jenkisns job for instance) [01:02] asac command acknoledged [01:03] sergiusens: so thats what we usually do then? [01:03] bypass the build tests? [01:03] (if it doesnt go ahead in infrastructure) [01:05] asac: it's not the tests that fail, it's the package build [01:06] cyphermox_: it's landed [01:07] asac: I'm still completing my thoughts from over the weekend for that process document [01:07] no thats fine... just try to understand the fasttracking operation :) [01:07] asac: skip builds and tests [01:07] ight [01:07] and then commit to archive [01:07] etc. [01:07] sergiusens: is a changelog available for daily builds? [01:07] asac: fairly confident of it since it already ran once and worked [01:07] right [01:08] sergiusens: cdimage.ubuntu.com/ubuntu-touch-preview/ is no longer being updated [01:08] mterry, yo [01:08] sergiusens: do you know if it goes through proposed testing this way? [01:08] bjv: there isn't really much work going in there, what do you need? [01:09] asac: after this, the daily build process comes up and picks up on all the trunks that have updates and runs tests again pushes to ppa:ubuntu-unity/next the build in that ppa is what goes into proposed [01:10] robert_ancell, hihi. Just playing with lightdm and my greeter branch on my nexus4. How do I get lightdm to not want to spawn X? [01:10] right so now it goes to ppa [01:10] gest build [01:10] gets tested [01:10] and then pushed to archive through poropsed [01:10] gotcha [01:10] thanks [01:10] annerajb: I'm not a Windows person, but you could install adb on Windows [01:10] mterry, not spawn X at all? [01:10] asac: in a nutshell, yes [01:10] well not sure if extract-files.sh will run but ill give it a shot thanks [01:11] robert_ancell, well... work with mir on my nexus4 at least. If it's xmir, that's fine. But the lightdm.log file said it couldn't find the X executable [01:11] mterry, i.e. run a native Mir greeter and not through XMir? [01:12] sergiusens: do you know if most packages we do in daily-release us cmake or automake ... or somewhat balanced? [01:12] use [01:13] asac: cmake for almost everything we did for ubuntu touch [01:13] asac: it's our prefered tech [01:13] sergiusens: I noticed that. I was wondering what the difference was between cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/current/ and pending/ [01:14] bjv: that's not ubuntu-touch-preview though, this is the _flipped_image [01:14] bjv: pending may break bad, current may break but not bad [01:15] sergiusens: so http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/20130714/ is the same as http://cdimage.ubuntu.com/ubuntu-touch-preview/daily-preinstalled/20130712/ ? [01:16] only flipped? [01:16] bjv: no [01:16] because there is no ubuntu_stamp in 20130714/ [01:16] oh [01:16] bjv: ubuntu-touch-preview is unflipped, ubuntu-touch is flipped [01:17] sergiusens: i cannot find a changelog on the web for the flipped, ubuntu-touch though [01:18] asac: I don't recall where it was documented, but tvoss was one of the authors of it [01:18] or list of packages/versions [01:18] bjv: there isn't any [01:19] bjv: ubuntu-touch uses pure cdimage build scripts whilst ubuntu-touch-preview was an artifact copy of what we built on jenkins [01:19] sergiusens, aborted? did you push manually? [01:19] Saviq: not manually [01:19] sergiusens, yeah, fast-tracked, k [01:19] Saviq: yup [01:19] sergiusens, thanks [01:20] Saviq: as I said, no reason to run build+tests all over again if the CI one worked... IMO at least [01:20] sergiusens, yeah, and especially when we can see what was really failing [01:20] sergiusens, q, now, will daily release pick it up still? curfew passed over an hour ago? [01:21] sergiusens: ah, I see [01:22] anyway... I'm gonna go sleep now... [01:22] sergiusens, thanks again [01:22] sergiusens: can .manifest files not be produced with the cdimage scripts for Touch? [01:25] is there a way we can easily produce a kernel/build/whatever that we can start with 256m, 512, 1024m etc.? [01:26] like a "maguro-265m edition" build :) [01:26] guess 512m would be more interesting for now [01:28] wonder if extract-files.sh can be made to work thru adb over TCP/IP [01:46] mhall119, ping [01:47] iBelieve: pong [01:48] mhall119, concerning merge requests for the File Manager app, will I have the other 2 devs still review my merge requests? That's what the Touch Developer docs seem to say, but I haven't seen either of the other 2 doing that, so I'm a little confused. [01:48] Saviq: regarding the question, cyphermox_ can still manually trigger [01:51] iBelieve: it's best practice to get somebody else to review every MP [01:52] mhall119, that's what I thought, but I was just checking. Thanks! [01:52] no problem [01:54] robert_ancell, heyo! I timed out earlier, if you responded to me [02:02] mterry, I said: "i.e. run a native Mir greeter and not through XMir?" [02:13] robert_ancell, yeah, ideally. I have the mir branch of unity and it can work. I just want lightdm to spawn my greeter, but it complains about X not being available [02:14] mterry, right, lightdm doesn't support native mir greeters/sessions yet. I am currently working on that [02:14] robert_ancell, oh OK. But it supports Xmir, eh? [02:14] mterry, yes [02:18] Hello [02:19] hmm did somebody changed anything? 3 hours ago the envsetup.sh worked fine now i get this dpkg-source: warning: failed to verify signature on ./libhybris_0.1.0+git20130606+c5d897a-0ubuntu10.dsc [02:23] Saviq, ping [02:32] i think i got it fixed brunch has to be run once before make to setup some files like the kconfig [02:34] Saviq, ps-panda-1 has been taken offline [02:35] fginther: I forced unity8 to build only with ps-panda-7 [02:36] fginther: although it is a hot fix [02:36] sergiusens, ack [02:37] sergiusens, does it build under qemu? [02:37] fginther: I don't think so, the unit tests would fail [02:38] sergiusens, :-( [02:38] fginther: it was one of the first reasons we needed real hardware ;-) [02:38] sergiusens, I fear we may be short on arm builders starting tomorrow morning [02:39] what hardware you guys are using to compile? [02:39] panda boards [02:40] http://www8.hp.com/us/en/products/proliant-servers/product-detail.html?oid=5375897#!tab=features [02:40] Hey all. [02:40] annerajb, yeah, I [02:41] I'll take a few of those :-) [02:41] fginther: i work for hp they are awesome.. * i r biased lol* [02:41] I just jumped in the IRC for a quick question that i can't seem to find an answer to... [02:42] annerajb, HP demoed one at UDS in copenhagen. it was a nice looking box, despite the jet engine sound [02:42] Will Ubuntu touch only support the ARM architecture? Or will i be able to load it on just about any tablet or slate? [02:43] Cruellyricist: we are only focusing on arm these days, but only because we are basing out of cyanogenmod. So if it runs on CM, it will run easily [02:43] fginther: i havent seen a fully loaded one yet but i bet is loud lol [02:43] a port to a different device is not out of the question, but you just need to do more work [02:44] http://www.gigabyte.com/products/product-page.aspx?pid=4456#kf this is specifically why i asked... I would LOVE to load Ubuntu touch on this baby when it sees its first full release. [02:45] It seems so right for it... Then again I'd also load it on the Microsoft Surface too... [02:50] Anyone, feedback? === chihchun_afk is now known as chihchun [03:03] yay Running: mkyaffs2image -f -c 4096 -s 128 /home/annerabj/android/out/target/product/epicmtd/data /home/annerabj/android/out/target/product/epicmtd/userdata.img annerabj@annerabj-VirtualBox:~/android$ [03:33] asac: that's simple, you just need to limit the memory via kernel argument [03:36] amazing, unity8 just got published === Namidairo`bnc is now known as Namidairo [06:09] anyone seen Saviq? === bzoltan1 is now known as bzoltan [07:12] how i can install ubuntu phone to alcatel onetouch idol ultra [07:12] anasp, port it [07:15] noskcaj can you gave a little introduction [07:22] good morning [07:22] morning [07:25] anasp, the info is at https://wiki.ubuntu.com/Touch/Porting === alan_g is now known as alan_g|afk === MrDHat|afk is now known as MrDHat [07:44] rsalveti: would it work? e.g. without graphcs driver gonig crazy? [07:44] ogra_: hi [07:44] did folks get the stuff through yesterday still? [07:45] i dropped out at 4am (like 5 hours ago :)) [07:45] dunno, i went to bed before [07:46] * ogra_ gets some coffee to wake up [07:47] ogra_: hehe... ok thought you aklready checked [07:49] didrocks: do you know if unity8 made it through this morning? [07:49] asac: I think it did, as he was luckily not building on ppc [07:49] https://launchpad.net/~ubuntu-unity/+archive/next/+sourcepub/3365714/+listing-archive-extra [07:49] i think thats it [07:49] (the rest is blocked on the ppc build) [07:50] asac: right, that's the one [07:50] that one has powerpc [07:50] pending [07:50] asac: yeah, it will build-dep forever [07:50] we don't have qtdeclarative building on ppc because of v8 [07:50] can you check if it made it into the archive? [07:50] we need a new image [07:50] still [07:50] asac: which archive? [07:50] saucy [07:50] real thing [07:51] asac: hum, I see no decision on the ML for the archive yet [07:51] s/ML/emails/ [07:51] err [07:51] whatever place [07:51] we currently use to get it on image [07:51] asac: "next", it's where you set it [07:51] ic [07:51] so you have it [07:51] ogra_: kick off image builder if the last one failed then [07:51] didrocks: do you know how to kick off image build? [07:52] asac: we have daily-build ppa as a staging env (where it's testing) -> daily-build-next [07:52] ooopsss [07:52] sorry: [07:52] daily-build ppa -> distro [07:52] daily-build-next ppa -> next ppa [07:52] right ... yeah i remember [07:52] didnt sleep much :) [07:52] so have to wake up too [07:52] I read that! :) [07:52] asac: on image build -> unfortunately, not [07:52] ic [07:52] kk [07:53] asac: you just get close to the first coffee machine you cross :) [07:54] lol [07:54] yeah... i have coffee now [07:54] goes up from now on i am sure :) [07:55] bah ... it tastes pretty bad :) [07:59] asac: well… it's your coffee, so are you to blame? :p [08:01] bah, I remember the time where asac was claiming that sleep is for the weaks ;-) [08:01] seb128: he's getting old… [08:02] yeah [08:04] asac, the scheduled build starts in 20 min ... relax :) [08:04] oh, and good morning german friends ;-) [08:04] and didrocks [08:04] moin moin :) [08:05] hey again seb128! [08:05] so reading the backlog, it's "stay away from the current image"? [08:07] seb128, wait 1.5h for the new one, or make sure to immediately upgrade unity8 after install [08:07] ok [08:07] (or just use phablet-network-setup to configure your wlan [08:07] ) [08:07] the network indicator is broken ... [08:08] s/is/was/ [08:08] ok [08:08] ogra_, thanks [08:08] :) [08:09] heh ... [08:09] https://code.launchpad.net/~saviq/unity8/fix-network-page/+merge/174828 [08:09] so it took only 7 jenkins retries and 14h to get that 2 line fix in :P [08:09] ogra_, yeah, and it was fast-tracked in the end [08:09] seb128: i am awake :) ... so good morning [08:10] Saviq, LOL ! [08:10] ogra_, it was a HW failure [08:10] asac, good morning ;-) [08:10] goota love our infrastructure [08:10] ogra_, some of the pandas were acting up [08:11] we shouldnt use them anymore [08:11] we have someting != pandas ? [08:11] :) [08:11] asac, we have calxeda machines since over a year [08:11] was told they dont work :) [08:11] actually the work since a few days [08:11] dunno how many nodes, but i think a magnitude more than we have pandas in the DC [08:11] but still need to be brought up [08:12] yeah [08:12] and wont be enough to also have them for daily bot i think [08:12] (but i only think) [08:12] i coulld copy/paste you that very same sentence from several IRC logs over the last year [08:12] that they worked? [08:12] * ogra_ heard that to often to belive in it until he sees the first package build coming out of one [08:13] asac, that they dont work, but just have been brougt up and will be available soon [08:14] we were now told that they worked, survived stress tests etc. [08:14] if i hear it i always have to think of vinyl ... and scratches .... === didrocks1 is now known as didrocks [08:14] lool: i think i am in now... feel free to pop in when you are ready [08:14] .oO( ah ... the needle jumped back 3 months again) [08:15] slangasek: how does the calxeda box look like? [08:15] :) [08:15] ogra_: image build kicked? [08:16] asac, 10:32 our time is when the scheduled build runs anyway === alan_g|afk is now known as alan_g [08:16] (the builder runs other builds before ) [08:17] (and we only have one thats half way reliable) [08:18] is a 20130716 image building now? [08:18] I like to test something that probably went in saucy last night, so if its building I'll wait with my new phablet-flash [08:22] it should be burped up by cdimage in around 1h [08:25] ok, thanks === chriadam is now known as chriadam|away [08:45] Good morning all, happy Corn Fritter Day! :-D [08:46] yummy, popcorn [08:51] ogra_, seems like a different thing [08:51] * seb128 googling for corn fritter [08:51] oh, yeah [08:51] nothing to do with frying :) [08:59] hum, I wonder if the image changed that much since the previous download I had [08:59] didrocks, hey, when jenkins merges latestsnapshot, is that before or after the ppa build? [08:59] or if rsync is just not working great on those images [08:59] Saviq: this is on publication, meaning, when publishing to "dest" [09:00] (if manual approval for publication, this is after this approval as well) [09:00] didrocks, so it's already available in dest at that point? [09:00] didrocks, as in, for people to download? [09:00] Saviq: when dest is a ppa, I think launchpad takes ~10 minutes to publish it [09:00] seb128, the armhf zip is always completely downloaded [09:00] the device specific ones work fine [09:00] ogra_, :-( [09:01] Saviq: if it's -proposed -> release, it's taking a little bit more time [09:01] didrocks, k thanks [09:01] seems that putting a tar.gz into a zip doesnt work so well [09:01] even though the tarball is gzipped with --rsyncable [09:01] ogra_, the armhf zip is the biggest one [09:01] yeah [09:01] well, image based updates will solve that [09:02] since they already are diffs [09:02] (before downloading) [09:02] didrocks, so... https://launchpadlibrarian.net/145055955/buildlog_ubuntu-saucy-armhf.unity8_7.81.3%2B13.10.20130716ubuntu.unity.next-0ubuntu1_UPLOADING.txt.gz says "Finished at 20130716-0311" [09:02] didrocks, but latestsnapshot got merged at -0332 [09:02] Saviq: no, don't look at the build log not the changelog [09:02] nor* [09:03] Saviq: look at when the MP was submitted [09:03] didrocks, -0327 then [09:03] and -0332 LP marked it merged === anasp_ is now known as anasp [09:03] Saviq: it's even a little bit before, it's when the MP is submitted [09:03] not merged :) [09:04] didrocks, yeah, -0327 is submitted [09:04] didrocks, so about 10 mins later it was available? === oreneeshy_ is now known as oreneeshy [09:05] Saviq: approximately, yeah (it seems publishing a ppa doesn't follow a deadline rule like for distro publishing, but it's in practice ~10 mins) [09:05] why? [09:05] hmm.. I get the feeling that after I install a new phone image and the phone reboots, my laptop always gets disconnected from my wireless network [09:05] didrocks, I'm doing a crisis report about yesterday's wifi fail [09:06] ogra_: I understand that we actually have them physically in London now, which is a major step :) [09:06] Saviq: ah ok [09:06] cjwatson, yay ! [09:06] cjwatson, yeah, that gives me some confidence it will actually happen :) [09:06] infinity has been mocking one of them being shipped with cardboard baffles [09:07] didrocks: how do you mean doesn't follow a deadline rule? [09:08] didrocks: the PPA publisher tries to run every five minutes, just like the primary archive publisher; IIRC it generally only manages to run once every ten (similarly, the primary archive publisher doesn't actually manage to run every five-minute slot) [09:09] cjwatson: hum, sometimes, I'm under the impression it's taking 15 minutes to publish, sometimes 5, is that because one publication is taking more time? (and just one is publishing at a time) [09:09] cjwatson: what is the discrepancy influenced by, number of total new packages in all ppas? [09:11] didrocks: since it takes about ten minutes, that means that there's a range depending on when you upload relative to the publisher starting ... [09:11] didrocks: just as for the primary archive :) [09:12] cjwatson: I was under the impression that distro publishing was taking ~30 minutes for primary archive [09:12] back then ... in the old days [09:12] didrocks: outdated :) [09:12] :) [09:12] didrocks: we sped it up a fair bit recently [09:12] it is breezingly fast nowadays [09:12] cjwatson: outdated from how long? I remember staring last month or even closer to it :) [09:12] oh nice! [09:13] a couple of weeks [09:13] not that outdated then! :) [09:13] great news anyway ;) [09:13] didrocks: the PPA publisher walks over all PPAs, so it has some constant cost with its current design [09:14] didrocks: it's now about five minutes if it only has to operate on saucy-proposed, and around 20 if it has to do things involving saucy [09:14] excellent news :) [09:14] * ogra_ hasnt had a build that took more than 1h from dput to binary-in-archive since a while [09:15] ogra_: you're working on too small packages ;) [09:15] heh [09:23] ogra_: new image arrived? [09:24] still building [09:24] should be ready any minute [09:25] hmm, build is done, nothing published and i didnt get any failure mail [09:26] ah [09:26] i'm just to impatient ... it is publishing [09:27] asac, there now ... phablet-flash --pending to get it [09:27] * ogra_ rsyncs [09:28] * ogra_ goes afk while the sync runs [09:29] asac, bare time report after yesterday's fail is in your email [09:30] * popey flashes [09:30] mhr3: Hey [09:31] mhr3: can't physically join the meeting location you've proposed :-) [09:31] lool, hehe, no worries, you'll join virtually [09:32] just give me a sec to get the wired network and stuff [09:32] gema: gimme a few [09:32] mhr3: grabbing a coffee in the mean time [09:32] asac: only because you ask nicely ;) === vrruiz_ is now known as rvr [09:32] ogra_: we have pending? nice!! [09:32] :) [09:33] lool, https://plus.google.com/hangouts/_/cd522c45ba33d085c4186ba86183e50fcd3b98fa [09:34] ogra_: can you confirm that i dont loose my content in /home/phablet? [09:36] asac: correct, don't use -b [09:36] asac: if you use -b it will wipe _everything_ [09:45] \o/ [09:49] asac, right, /home content is preserved [09:53] ogra_: my mako has a dead battery so can't flash that for a bit. grouper flashed though.. [09:54] ok [09:54] Morning all how's the images today? [09:54] * ogra_ still has a few mins until sync is done [09:55] hey rsalveti - do you think it'd make sense to have an editing session for the porting guide update - where we'd update it in a etherpad together? [09:56] I'm sure people like sergiusens, ogra and others would be able to help [09:56] (I could probably fix a few small things and proof-read) [09:57] dholbach, prob is that i'm trying to do a port since two weeks but there is always comeing some package or image issue in my way, so i dont really have much first hand info i could add [09:58] i'm trying to invest a day during this week to get my SGS2 ported with flipped [09:59] * ogra_ flashes maguro [10:04] 3G Data connection assistant on Ubuntu Touch | http://askubuntu.com/q/320596 [10:04] dholbach: when you do update it, I'd like to test the process [10:05] popey, cool [10:05] dholbach: off to oscon next week so prefer if it could be updated this week ☻ [10:06] popey, that's not in my hands [10:06] popey, I'm not an expert when it comes to all of this [10:06] woah, the boot got a *lot* slower [10:06] (at least the first boot) [10:07] hmm, my mako flat out wont boot [10:07] it took like a minute with black screen here [10:08] i guess the new ubuntu-touch-session runs slower ... [10:08] which is weird, when i tested for ricmm it was actually a lot faster [10:08] i wonder what changed vs the stuff i tested for him [10:09] * ogra_ removes wifi config to test if that works now [10:09] ah, second boot is faster [10:10] yay, wifi works fine [10:10] apps run, i can browse, looks fine [10:12] ah, and the dashboard tests are done too [10:13] hmm, mako still missing though [10:25] popey: how did the mobile thing go yesterday? [10:26] hmm, do we have others with mako to test ? grouper and maguro look fine [10:26] * ogra_ would like to mark the image good [10:26] this has been on charge for 2 hours and still wont boot [10:27] not even when holding the power button down for a while ? [10:27] popey: is it the phone? if so try taking the battery out if you can [10:28] n4 makes that hard :) [10:28] you need a glass cutter i suppose :) [10:29] i think the n4 has the power button wired in a way that it does a hard reset when holding it though [10:29] ogra_: I thought there was a rule that europe passed that said all batteries should be accessible with x amount of screws [10:29] well, it has screws :) [10:30] i was just joking with the glass cutter [10:31] ogra_: mind you the same people also said that there should only be one power supplier and settled on micro usb oh apart from a certain fruit based product that shall remain nameless [10:32] hehe, well, they always went their own non-std way in the past [10:32] popey: I had the same thing the other day [10:32] popey: you have to plug it [10:32] and press the power button for *60s* [10:32] i have held it for 10 seconds [10:32] blimey [10:33] * popey holds and types [10:33] it will get out hibernate === sil2100_ is now known as sil2100 [10:34] nope [10:35] ogra_: so talked to gema... we will use --pending now for producing the dashboard [10:35] asac, yeah, looks like thats already happening [10:35] i think it might be implicit [10:35] because it still pulls the latest by date? [10:35] or you say that the team was faster than us? [10:35] and just fixed it [10:35] ? [10:35] popey: To bring the Nexus 4 out of hibernation: [10:35] 1. Plug Nexus 4 to a wall charger. [10:35] 2. Press and hold the Power button for 60 seconds [10:35] popey: this is what worked for me ^ [10:36] asac, oh, indeed, it could just pull latest [10:36] ogra_: so please ensure we get results on dashboard [10:36] yeah, just tried that [10:36] you have jcollado to poke jobs [10:36] popey: there is an alternative: http://productforums.google.com/forum/#!msg/mobile/eREo__Lhuwg/Kuly9n_cmscJ [10:36] in case they return nothing [10:36] or weird stuff [10:36] asac, dashboard looks fine except that mako is sill missing [10:36] ogra_: once dashboard is not regressing, we propagate if manual test looks fine too [10:36] davmor2: ogra_: do be honest, micro-usb is crap - it's still not upside-down reversible & it's limited to low wattage, hence iphone's charge quicker since it mandates higher W [10:36] an given popey's issues we dont have a mako test at all yet ... [10:36] ogra_: ? [10:36] please compare with results from yesterday [10:37] since he is usually the only manual mako tester [10:37] there are bunch of tests not yet finished [10:37] asac, green today .... [10:37] we want all the autopilot tests that succeeded at least to be there and those that are orange [10:37] 100% [10:37] all fine [10:37] ogra_: look at the test count [10:37] ogra_: it didnt run everything yet [10:37] is jenkins down? [10:37] ogra_: just the quyick tests [10:37] ogra_: so wait :) [10:37] 11 tests passed [10:37] ogra_: compare: http://reports.qa.ubuntu.com/smokeng/saucy/image/3010/ [10:37] looks all good ... [10:37] mako isnt even listed [10:37] xnox: I didn't say it wasn't justifiable it is more that fact that everyone else is happy to go along with the standard except apple [10:37] ogra_: its not the truth yet [10:38] ogra_: with http://reports.qa.ubuntu.com/smokeng/saucy/image/2998/ [10:38] ogra_: the green doesnt say anything until we have results for all of those [10:38] ogra_: yeah, you need to integrate the number of tests running :p [10:38] asac, it says 11tests, 11passes per arch [10:38] ogra_: so work with jcollado to find out of those jobs are still running etc. [10:38] davmor2: micro usb is the single worst mobile phone connector [FACT] ☻ [10:38] ogra_: look at the other [10:38] dont tell me thats not true ... [10:38] that's why for daily release, we fail if we have less tests results than the previous run :p [10:38] ogra_: its not all tests... it just is one test suite ... the rest is still running [10:38] what good is a dashboard if it doesnt show the right numbers :) [10:38] we want all results to be there [10:39] ogra_: it shows the right number, it doesnt show what hasnt finished yet. [10:39] gema and folks are working on making that better [10:39] so you see what to wait for before approving [10:39] ogra_: so check with jcollado if those autopilot jobs are still running ... otherwise he might want to run them etc. [10:39] jcollado: can you check? [10:40] e.g. are all the autopilots running? [10:40] or are we waiting for dashboard publishing? [10:40] asac: Let me see [10:43] asac, ogra_: Umm, I see that the smoke test case run, but not the downstream projects. I'll trigger those jobs to run manually now. [10:43] jcollado: all the autopilots, security etc. [10:43] yeah [10:45] asac: Yes, I've scheduled eventstat-saucy-touch-armhf-install-idle-maguro, saucy-touch-maguro-smoke-sdk, saucy-touch-maguro-smoke-security and smoke-saucy-touch-apps-maguro. [10:45] jcollado: do we know why that happens? [10:45] popey: works for me :P [10:46] popey, what woudld you chose ? thunderbolt ? [10:46] thunderbolt is better than any usb connector [10:46] lol [10:46] all usb, A, B, full size, mini, micro, they all suck giant donkey balls [10:47] asac: Today jibel was trying to figure out why the jenkins instance was restarted. I don't know why that happened, but that probably affected the job execution. [10:48] ogra_: lisettte has a device in the same state as mine [10:48] oh, maybe not, hers at least gives google logo, mine doesn't [10:48] jcollado, there were jenkins HW issues last night, not sure that was the same machine though (it failed CI stuff) [10:49] jcollado, several machines rebooted, I suspect a power failure, retoaded is on the case [10:50] ricmm: ^^ [10:50] ogra_, jibel: Ok, thanks. [10:50] bah, still no buildd stamp in the image ... hurry up IS ! [10:52] ogra_: hey, did you see my last question last night? [10:52] sergiusens: thanks [10:52] sergiusens, hmm, nope [10:52] ogra_: if there is a way to check if the image builders can see http://people.canonical.com/~sergiusens/click_packages/ [10:53] hmm, not beyond having livecd-rootfs try it and print results in the build log i guess [10:54] probably cjwatson has a better way [10:54] shouldnt be hardd to add a temporary script that tries a wget or so [10:54] ogra_: ok, I'll ask him when I get back, need to drive my partner now [10:54] thanks [10:54] yeah, i can do that too unless he comes up with something better [10:55] great, I'll get to that in a bit [10:56] ogra@chromebook:~/branches/livecd-rootfs$ grep -r people.canonical [10:56] live-build/auto/config:SEEDMIRROR=http://people.canonical.com/~ubuntu-archive/seeds/ [10:56] ah, seems to already be used anyway [10:57] (like we were told before as i now remember) [10:57] popey: do you happen to know if there is a bug against the browser/webapp for twitter in that you can't scroll your timeline === greyback is now known as greyback|food [11:04] davmor2: yes, i filed one [11:05] bug 1190212 [11:05] bug 1190212 in webbrowser-app "Unable to scroll in twitter web page" [High,Confirmed] https://launchpad.net/bugs/1190212 [11:05] popey: I was about to ask that :) === chihchun is now known as chihchun_afk === MacSlow is now known as MacSlow|lunch === schwuk is now known as schwuk_away === schwuk_away is now known as schwuk [11:30] Is Ubuntu Touch compatible with Oppo Find 5 | http://askubuntu.com/q/320624 [11:50] finally got mako to boot! [11:50] yay [11:51] popey, well, asac wants us to wait for the dashboard ... and since there seem to be jenkins issues i think there is no hurry [11:51] had to repeatedly plug/unplug the charger to get it to flash red, then press power button and it woke [11:51] well I'll flash anyway [11:51] yeah [11:51] once it charges a bit [11:51] 3% [11:51] to me the image looks good enough to mark === alan_g is now known as alan_g|lunch === muhammad is now known as Guest2012 === greyback|food is now known as greyback === MacSlow|lunch is now known as MacSlow === _salem is now known as salem_ [12:29] ogra_: are vendor/cm bits required for ubuntu-touch build? [12:29] ogra_: we don't really use their ringtones do we? [12:32] /msg popey hey, no mako build today? any idea what is up? [12:33] sergiusens: hello! [12:33] *sigh* [12:33] /me notes the negative impact of unexpected spaces? [12:33] /msg rickspencer3 yeah! [12:34] xnox: no, you can delete that, the fonts, the layout xmls and what not [12:34] sil2100: hey [12:34] rickspencer3: ogra_ was planning on marking it good. i had problems with my mako today not working [12:34] /kick popey [12:34] just flashing it now [12:34] ☻ [12:34] sergiusens: ack. but it looks like some prebuilds are still copied across.... sysinit et al. [12:34] thanks popey [12:34] xnox: are you still picking them up from the private branch? [12:35] popey, is that why it is not showing up on reports.qa.ubuntu.com ? [12:35] not sure about that. [12:35] xnox: I can clean it up now if you want [12:35] sergiusens: I'm doing a packaging review of demo-assets right now, and I noticed that all the graphical assets inside are not copyrighted in debian/copyright - is there some common license you wish to use for all those? [12:35] sergiusens: vendor/cm is from git. No, it's ok. I'll just add it my list of excludes when generating the tarball for the package. [12:35] i dont actually know how reports.qa.u.c is built... gema ? [12:35] sergiusens: since, for instance, there are some .png's in there [12:35] sil2100: I thought we were dropping demo-assets from the image, pmcgowan ? [12:35] sergiusens: i'd rather keep "brunch" command working, without modifying our fork of cynogenmod too much. [12:35] rickspencer3, asac asked me to wait for the dashboard to finish, but sadly jenkins was acting up [12:36] sil2100, right but you probably didnt know that [12:36] thanks ogra_ [12:36] rickspencer3, from my perspective the image is fine [12:36] /me tries phablat-flash [12:36] sil2100, but we will still have packages for samples in the ppa, so good to get it straight [12:36] (i would mark it good after hearing that popey can make calls) [12:36] rickspencer3, phablet-flash --pending ;) [12:36] sil2100, we use a creative commons license, let me get a reference [12:36] ogra_: just made a call [12:36] xnox: ok, I'll take a look at the priv bzr branches, brunch should keep on working for our case [12:36] rickspencer3, that will give you the one waiting for approval [12:37] pmcgowan: ah, ok, since didrocks mentioned to me yesterday to conside demo-assets for daily-release, so hm, I thought it will still be valid [12:37] popey, yay [12:37] pmcgowan: demo-assets has images from imdb (or similar) [12:37] sergiusens: well, vendor/cm is in git =) not in private bzr. [12:37] ogra_: http://popey.com/~alan/device-2013-07-16-133714.png [12:37] sergiusens, those we dont own for sure [12:37] sergiusens: unless there is a private bzr of vendor/cm..... [12:37] pmcgowan, sergiusens: for some of our graphics we used CC BY-SA 3.0 usually [12:37] xnox: yup, I know about vendor/cm, it's only in git [12:37] sil2100, exactly [12:37] for ones we made [12:38] sergiusens: ok. =) so i'm using the one and only then. [12:38] popey, you are sideways [12:38] I am! [12:38] xnox: yes, one vendor/cm to rule them all ;-) [12:38] sil2100, we are going to restructure the "demo" packages and leave them in the PPA and not in build [12:38] xnox: we can probably rename it to vendor/ubuntu :-P [12:38] geez, why do all maguro tests show failed ? [12:39] sergiusens, are you doing a weekly update email? [12:39] pmcgowan: ok, so there is no plans for daily-releasing any "demo" packages, yes? [12:39] sil2100, thats the proposal yes [12:39] pmcgowan: for the meeting? let me think really hard... everything is in progress from last week [12:40] sil2100, just forwarded the email to you [12:40] sergiusens, yes thanks [12:40] dont think too hard [12:41] pmcgowan: thanks! [12:48] asac, so obviously all the dashboard tests are failing due to not being able to connect to dbus by whatever reason (dbus is just fine on the image here and does it's duty), should i still wait until dashboard finished (likely with 99% failure) or shouldnt we rather finally give something to the waiting users [12:49] * ogra_ really thinks we shouldnt depend on dashboard tests until they actually work reliable [12:50] ogra_: where was the dashboard [12:50] ? [12:50] gema, is it a known issue that utah cant connect tp dbus on the phones ? [12:50] sergiusens, http://reports.qa.ubuntu.com/smokeng/saucy/ [12:50] ty [12:50] i suspect they dont run as the phablet user [12:51] most likely, so no DBUS_SESSION_BUS_ADDRESS [12:51] or use a wrong way to become that user [12:51] right [12:52] Jul 16 11:38:45 ubuntu-phablet utah: 11:38:45.192 ERROR importer:51 - Could not find any typelib for Gdk === alan_g|lunch is now known as alan_g [12:52] mfisch, Hi! [12:52] yeah, that too [12:53] though the dbus errors dont seem realted to gdk [12:53] oh org.freedesktop.DBus.Error.NoServer [12:53] right [12:53] yeah, the gdk thing is probably utah specific [12:53] which is nonsense [12:54] dbus runs else we wouldnt be able to do the manual tests indeed [12:54] so i suspect it simply doesnt run as phablet user [12:54] ogra_: well if my memory works well, from the logs they are running as a utah user, going to be very complicated if they don't run as phablet [12:54] is there a way to turn on the screen from command line through powerd ? [12:54] ogra_: doanac is the right person to talk about that [12:54] * ogra_ just doesnt want to hold back the image the whole day for that [12:55] or any other way for that matter [12:55] om26er: powerd-cli [12:55] sergiusens, ack [12:55] sergiusens: we can run as phablet, that's not a problem [12:55] sergiusens: I'd rather you check with doanac how they are actually run now [12:55] gema, that will surely help to get the dbus address [12:55] ogra_: which job are you looking at? [12:56] currently maguro, the others seem to still not run [12:56] gema: https://jenkins.qa.ubuntu.com/job/saucy-touch-maguro-smoke-gallery-app-autopilot/7/consoleFull [12:56] and mako is completely missing [12:56] ogra_: yes, doanac was trying to get those running yesterday, I don't know where that ended [12:56] the job tha provisions mako was failing [12:57] ogra_: I agree we need to get them running stably [12:57] doanac: will be online soon [12:57] right, my point is more towards asac, i dont want to block the publishing of the images based on that [12:57] ogra_: asac and I have been discussing that this morning, we are not going to do that [12:57] until we know they run at least half way reliable [12:58] ogra_: we are going to hook you and maybe sergiusens in the process and get all the issues fixed, the PS QA guys are helping the teams get their testing to a reliable place [12:58] ogra_: a lot of work is going into making these reliable [12:58] great [12:59] (i wasnt doubting that the plan is to make them more reliable :) ) [12:59] nobody is planning to hook the image publishing to a bunch of failing tests [12:59] well, currently we do [12:59] ogra_: right now the dashboard is running tests after the images are published [12:59] since i was asked to hold back publishing until all tests ran :) [12:59] sergiusens, i tried: 'sudo powerd-cli display on dc' screen is still off :/ [12:59] gema, unlikely [13:00] om26er: why sudo? [13:00] gema, it should run the images from /pending [13:00] ogra_: I don't believe the tests are running until theres a new image in current [13:00] gema, while phablet-flash by default pulls from /current .... [13:00] ogra_: I know, I need to talk to doanac about that, this is a requirement from this morning [13:00] and i'm the person moving them from one to the other manually [13:00] ogra_: ok [13:00] so currently the 0716 image isnt released [13:01] waiting for me to pulll the trigger [13:01] which i think i will just do now ... manual testing was fine across the board [13:01] sergiusens, You must be root to run powerd-cli [13:01] * ogra_ does [13:01] ogra_: I suggest we implement this later in the week [13:01] ogra_: we need some time to create the jobs that'll run on pending images [13:01] yeah, np [13:02] gema, phablet-flash has a switch for that, do you use it for flashing ? [13:02] hmm, ogra_ when the battery goes to complete zero, is it expected that the date/time reverts to 12:00 1970? [13:02] om26er: was the screen always black? what does logcat say? [13:02] ogra_: yes, but right now we are not using it [13:02] ogra_: we need to configure some jobs to do that in the lab [13:03] rsalveti, sergiusens, ogra_ (and anyone else), I put up http://pad.ubuntu.com/rw9y4Sk0D7 where we can update the porting guide together [13:03] maybe that'll make it a bit easier to get the work started [13:04] dholbach, doesnt make sense before we have code :) [13:04] what do you mean? [13:04] dholbach, the current porting process hasnt changed [13:04] sergiusens, i don't see any change in logcat when i run powerd-cli [13:05] only if we moved ports to flipped it will [13:05] ogra_, and that's not what we want? [13:05] dholbach, ratrher end of the week i'd say [13:05] dholbach: ogra_: and even then, we will try to make $ . ./build/envsetup.sh && breakfast $foo && make otapackage [13:05] generate everything that's needed. [13:05] people already asked for updates of the guide, that's why I thought it'd make sense to start the work on the update [13:05] xnox, yeah, but the modifications you will have to do will change === oreneeshy_ is now known as oreneeshy [13:06] and none of us has yet run through that process with a new tree [13:06] (well sans failing at various bits that need "fixing-up" aka "porting" until it finishes building that is =) [13:06] exactly [13:07] * om26er what does telepathy do when a call/sms notification appears to lighten the screen [13:07] just writing a guide out of the blue wont work [13:07] ogra_, no, not "out of the blue" [13:07] but rather update the current guide where necessary [13:07] i'm running a phablet-dev-bootstrap since a few hours here ... and hope to have a flipped SGS2 port by end of the week [13:08] i will take notes along the process [13:08] cool [13:08] dholbach, right, but we need to identify the changes first :) [13:09] .... 20130716 marked current btw .... [13:09] morning [13:09] * rsalveti reading backlog [13:10] rsalveti, summary: all better than yesterday [13:10] :) [13:16] ogra_: yeah, great [13:16] nobody going crazy before I start my day [13:16] :) [13:16] lately nobody == asac :P [13:16] haha [13:19] ogra_: is there an easy "on device" way to tell what you're running on? [13:19] like something in /proc or /sys? [13:20] getprop|grep cm.device [13:20] ogra_, xnox: I could imagine that even if some bits needs fixing up, the information you have would already help and if it's just in a "tips" bullet points section [13:20] nice one thanks [13:20] oh, even without grep [13:20] getprop ro.cm.device [13:20] thats better [13:20] much! [13:20] dholbach: sounds good [13:21] dholbach: will try to get to it during the next few days :-) [13:21] meanwhile I'm trying to finalize some of the work in phablet.u.c [13:21] I'll send a reminder mail to all of you :) [13:22] great [13:22] i'll happily get to it once i have assembled some notes by doing it myself [13:22] as i said, by enmd of the week i should have a port [13:23] dholbach: i see. [13:23] thanks guys [13:25] thanks dholbach (for making sure the porting guide gets updated at some point in time :) ) [13:37] hello [13:38] so I got a nexus 4 today and bought/cut a sim card for it; that works (tested under android) [13:38] is there any magic to make this actually work under current u phone builds? it neither asks me for a PIN, nor use it in the phone app [13:38] pitti: any logs? [13:38] rsalveti: lol [13:39] popey: which one [13:39] pitti, there is a command line method to give the PIN [13:39] is that what you need? [13:39] pitti: under ~/ubuntu-touch-session/ [13:39] rsalveti: i try to keep the stuff in /msg so you can't push back anymore :-P [13:39] j.k. [13:39] lol [13:39] pmcgowan: could be -- this is pretty much a virgin install, and my first time that I try this [13:39] pmcgowan: is that documented on the wiki? [13:39] rsalveti: also was stitting here yesterday/today till 4:30am [13:39] trying to understand how the pain goes when stuff doesnt go through daily-release :) [13:39] but i was up and runnig at 9:30 [13:40] popey, just in an email as I recall [13:40] not sure where mterry is on the UI [13:40] asac: hahah, did you feel the jenkins/panda pain? [13:41] meh, it forgot my wifi connection again, hang on [13:41] rsalveti: yeah i observed it [13:41] and took notes [13:41] pmcgowan, on PIN UI? That's a good thing for me to work on today actually [13:41] asac: good [13:41] Saviq's notes are great too though [13:42] ogra_: +1 [13:42] yeah ... have to read those still :) [13:42] well Saviq +1 [13:42] :) [13:42] mterry, indeed [13:42] * ogra_ hugs mterry [13:43] pmcgowan, (mzanetti was going to look into it in a bit, but I have some cycles today; mzanetti: I'm going to look at the backend part of it) [13:49] ogra_: did our contacat disappear? [13:49] jcollado: ah :) ... sry forgot your nick [13:49] heh [13:50] jcollado: so mako seems to be missing [13:50] he is there, the tests fail badly missing a dbus connection though [13:50] and maguro has a bunch of jobs panding [13:50] yeah there are 5 still outstanding [13:50] ogra_: do you see the same locally? [13:50] e.g. when running the autopilot with autopilot run ? [13:50] not when i manually use the device, no [13:51] its simply that the tests dont run as the phablet user [13:51] so they dont get access to the session bus [13:51] see backlog :) [13:52] ogra_: they ran the days before [13:52] asac, ogra_: In which jenkins job are you interested? From what I see, raring-touch-mako-smoke-default has been disabled after two failures. Is that the one you're looking for? [13:52] jcollado: guess thats it, yes [13:52] jcollado, saucy indeed [13:52] asac: hum, daily-release != upstream merger for once more (backlogging :p) [13:53] * jcollado looks for the saucy job [13:53] didrocks: ok i will distinguis between those two bots :) [13:53] jcollado, todays saucy smoketests of maguro on the dashboard [13:53] that was the one i specifically looked at [13:53] popey: which log is interesting? ofono-setup.log? [13:53] asac: thanks! [13:54] popey: that just says "ofono/ofono/account0 initialized" [13:54] pitti, grep ofono /var/log/syslog might be more intresting [13:54] asac, ogra_: Maguro or mako? [13:54] jcollado, maguro [13:54] pitti: do you need pin to unblock the sim card? [13:54] mako is missing completely [13:54] rsalveti: yes [13:54] jcollado: mako is completely down [13:54] jcollado: maguro has a few jobs missing still on dash [13:54] so hard to look at :) [13:54] but also failed hard [13:55] rsalveti, ogra_: so would it help to install android again, disable the pin, reinstall ubuntu? [13:55] pitti: so we just have support to unlock in ofono, not in the app yet [13:55] pitti: you can unlock using the ofono scripts [13:55] pitti: install the ofono-scripts package [13:55] rsalveti: I can't put it in my normal phone, I just cut the sim card to fit into the n4 [13:55] rsalveti, should be seeded, shouldnt it ? [13:55] yep, I have it [13:55] ogra_: it is [13:55] yeah, i thought so [13:56] pitti: then /usr/share/ofono/scripts/unlock-pin "pin" [13:57] funny, what other options are there for "pin" ? [13:57] (teh script name kind of implies i want to unlock the pin, doesnt it _ [13:57] a few, guess could be pin2 puk puk2 [13:57] ah [13:57] yeah [13:58] ogra_: Regarding maguro, both saucy-touch-maguro-smoke-security and smoke-saucy-touch-apps-maguro failed [13:58] jcollado, yes, because they dont run in the pahblet session [13:58] asac: ogra_: the dbus issue might be related with the new session manager [13:59] they should be run as the phablet user, so you get access to the session dbus [13:59] but we should always run the tests as phablet user [13:59] rsalveti: nice, works now; thanks! [13:59] yes [13:59] pitti: great [13:59] pitti, sweet [13:59] rsalveti: right. so it might be a genuine real issue [13:59] ogra_: What do you mean? [13:59] asac: the issue is running it as root instead of phablet user [13:59] rsalveti, it might expose it ... but as you said, testes need to run as phablet user [13:59] rsalveti: i thoguth folks used phablet-test-run [13:59] which should do the right thing (TM) [13:59] jcollado, the errors are obviously caused by not being able to attach to dbus [13:59] sergiusens: ^ [14:00] e.g. i couldnt run any autopilot test locally without using that [14:00] jcollado, dbus is provate to the phablet user [14:00] *private [14:00] e.g. just root -> autopilot run [14:00] you need to sudo -u phablet when running the tests [14:00] will fail miserably and the results we seeen the last days match what i saw as phablet user [14:00] rsalveti: yeah, we discussed with ogra_ earlier ;-) [14:00] else you wont get access to the session dbus [14:00] rsalveti: RTFBacklog ... LOL [14:00] haha [14:01] sergiusens: I did, but too long to notice all the details in there [14:01] asac, ogra_: Regarding mako, saucy-touch-mako-smoke-default didn't run I guess for the same reasons didn't run some of the maguro jobs. Running them now. [14:01] sergiusens: what is the outcome? [14:01] jcollado, great [14:01] rsalveti: wait for doanac to connect [14:01] cool [14:01] rsalveti: but most likely they are running as a utah user, so missing the DBUS_SESSION_BUS_ADDRESS and stuff like that [14:02] right [14:02] right [14:02] rsalveti: coming? [14:02] or did i kill your invite? [14:02] asac: yup, 1 se [14:02] sec [14:03] asac: calxeda box> sorry, what do you mean? [14:03] rsalveti, ogra_: fyi, the security tests are broken into two parts: privileged/ and unprivileged/. the former need to run as root [14:03] jdstrand: right, that's fine [14:03] jdstrand, well, as long as they dont need session bus access thats fine [14:03] no, they do not [14:04] yeah, all fine then [14:04] pmcgowan: summary sent [14:04] the app tests actually need to interact with the session bus [14:04] I didn't read all scrollback, just say the mentioning of the security tests and 'must run as the phablet user'. please ignore me then :) [14:04] s/say/saw/ [14:05] slangasek, we want more infinitys to bring it up faster :) [14:05] sergiusens, awesome thanks [14:05] jdstrand: it's the autopilot tests that fail they need a dbus session :-) [14:05] * jdstrand nods [14:05] I just wanted to make sure we didn't swing too far the other way :) [14:06] jdstrand: oh, I added really basic security manifests to the click packages and they are here now: http://people.canonical.com/~sergiusens/click_packages/ [14:06] Saviq, i just did a update and after reboot can no longer connect to the session bus over ssh or adb [14:06] sergiusens: nice! [14:06] sbeattie: ^ [14:06] Saviq, do i maybe need to reflash? [14:07] kenvandine, update as in re-flash ? [14:07] apt-get :) [14:07] i flashed on friday [14:07] well, the whole session mgmt was completely reworked [14:07] most likely that gets in your way [14:07] i would re-flash with 0716 [14:08] sbeattie: note that the ubuntu-filemanager-app and ubuntu-terminal-app should use '"template": "unconfined"' as per https://wiki.ubuntu.com/SecurityTeam/Specifications/ApplicationConfinement/Manifest#Click [14:08] ogra_, where can i get that? not on cdimage [14:08] rsalveti, you packaged ofono-qt for saucy. Is it maintained upstream still? [14:08] sbeattie: I mention that cause as soon as the click apparmor hook is implemented, those two may blow up :) [14:08] kenvandine, definitely on cdimage ... and phablet-flash will get it by default [14:09] oh... i was looking at the wrong url :) [14:09] jdstrand: initially I just set them all to be like the calculator app, will make these discriminations as we start to have hooks [14:09] kenvandine, heh, i hope we can wipe that old cruft soon [14:09] mterry: yes, still alive but without a lot happening there [14:10] we're mostly the ones doing changes in there [14:10] sergiusens: yeah, that's what I figured. that is probably fine, but note the terminal and file manager apps will either need to be unconfined or need a permission set not supported by the sdk (cause they are expected to have wide access) [14:10] ogra_, thanks! [14:10] jdstrand: understood, if it's only those two, I'll make the autochange now [14:11] pwd [14:11] ls [14:11] oops [14:11] sergiusens: that would be great. have policy groups as empty (or omit it), and then add that template line [14:11] ogra_, is that why I can't "restart ubuntu-touch-session" anymore? [14:12] Saviq: you should be able to restart unity8 as the phablet user (let me check the name and recall upstart user sessions) [14:13] sergiusens, ah, so we're under upstart now, sweet [14:14] Saviq: yup, no more shell scripts :-) [14:14] Saviq: yeah, unity8 is unity8 [14:14] unknown job: unity8, though :/ [14:14] oh, that is good to hear [14:14] -bash: /home/phablet/.dbus-session: No such file or directory [14:15] ricmm: ^ [14:15] * sergiusens flashes to see [14:17] probably need to get rid of that from .bashrc [14:17] rsalveti: ogra_ jcollado the autopilot tests may be failing to to that [14:17] in the hooks [14:17] that being upstart [14:17] sergiusens, no [14:18] they fail due to running under the wrong user ... the switch to upstart just exposes it more widely :) [14:18] that they worked before was a matter of luck (or even insecurity) [14:18] ogra_: sure, but it may also fail as the shell needs to be stopped to test the apps [14:18] ah, then they need to be reworked for this [14:19] fun, since we will change the whole thing again once lightdm enters the picture [14:19] sbeattie: if you didn't see in backscroll, sergiusens is updating the security manifest to unconfined for those two [14:20] ricmm, ++ .... we should hopefully end up with a default .bashrc by release :) [14:22] ogra_: yes, the .bashrc we have right now used to source the old hack [14:22] yep [14:22] right now we are using the same dbus session job that the desktop uses [14:22] also /etc/environment is quite a mess [14:35] om26er: hey === Namidairo is now known as Namidairo`bnc [14:44] mfisch, hi [14:45] mfisch, I wanted to ask a way to turn on the screen from commandline, powerd-cli does not seem to work for me [14:45] ogra_, ok, so reading the scrollback... sounds like right now we just can run apps from a console? [14:46] kenvandine, ? you always could [14:46] i still fail to connect to the session bus [14:47] kenvandine, https://wiki.ubuntu.com/Touch/ReleaseNotes#Ubuntu_SDK_Alpha [14:47] scroll down a bit [14:47] oh... my .dbus-session file is old [14:48] om26er: did you turn the screen off with the power button or via timeout? [14:48] mfisch, yeah, I turned it off with the power button [14:48] om26er: the power button overrides all states [14:48] om26er: if you let it timeout instead the display on calls will work [14:49] om26er: this is somewhat covered in the testing section of the wikipage [14:49] mfisch, hmm, it would actually work for my cases, since the purpose is to do that in the lab, where we don't have physical access to the phone [14:49] om26er: then how did you press the power button? === alan_g is now known as alan_g|tea [14:50] mfisch, i did that locally on my phone for testing, but the actual tests will run in the lab [14:50] om26er: sorry, I thought you said "wont work", yes, it should work fine for you [14:51] mfisch, is there a way to simulate the power button? [14:51] mfisch, as I am not sure if powerd-cli will work as it does not return prompt unless ctrl+c is pressed [14:52] example: [14:52] phablet@ubuntu-phablet:~$ sudo powerd-cli display on dim [14:52] Requesting Display On [14:52] Requesting Brightness Dim [14:52] ** Message: Got cookie: 3f26f743-9f20-4e3c-b0ea-5dce8f8bad58 [14:52] Display State requested, cookie is 3f26f743-9f20-4e3c-b0ea-5dce8f8bad58. [14:52] Press ctrl-c to exit. [14:52] om26er: thats right, because it kills requests when the caller dies on dbus [14:52] om26er: so we just wait [14:53] pmcgowan. ogra_ : I'm not sure if this is filed in the right place but you'll get the idea https://bugs.launchpad.net/platform-api/+bug/1201830 [14:53] Launchpad bug 1201830 in platform-api "Clocks need to sync better and use the same backend" [Undecided,New] [14:53] davmor2, yeah, thinking about timed derivative [14:54] davmor2, i'm not sure either, but thanks for that one ... definitely annoying [14:54] haha, yeah [14:54] om26er: what are you trying to do exactly? [14:55] mfisch, i want to turn on the screen each time autopilot tests are run in the lab. and then unlock the screen with a swipe [14:56] om26er: once the screen is on, and you swipe the screen it will stay on for 60 seconds [14:56] due to the activity timer [14:57] mfisch, let me try that here === alan_g|tea is now known as alan_g [14:57] om26er: so you could have a script kill powerd-cli [14:57] om26er: we can always investigate adding some extra debug interfaces as well if they're needed [14:57] ricmm, is there a hack i can use to access the session bus again? [14:58] it has been sooooo handy to be able to test stuff over ssh :) [14:58] kenvandine, just sudo -u phablet -i isnt enough ? [14:58] mfisch, what would be most useful is if we send a signal through powerd-cli to turn on the screen, similar to pressing the power button [14:59] ogra_, nope [14:59] unable to connect to session bus [15:00] om26er: and powerd-cli display on isn't enough because it holds the request and waits? [15:00] hmm that should actually work, you could probably dump the address somewhere on session startup [15:00] it used to get dumped to .dbus-session [15:00] but it isn't anymore [15:00] mfisch, yep, exactly [15:00] that's what ricmm was pointing out a bit ago [15:02] mfisch, something like: powerd-cli display on 60 (which won't hold command prompt) [15:02] om26er: can't you simulate that by wrapping it in a script? === pete-woods2 is now known as pete-woods [15:03] om26er: I'd say we could add an interface to poke the activity timer, but that code is moving into Unity [15:04] mfisch: I'd rather that the cli have a timeout after which it releases the request than an interface to poke the activity timer [15:04] mfisch: remember, the inactivity timeout will be moving to unity [15:05] mfisch, i tried: powerd-cli display on dc & /home/phablet/unlock_screen.py (which worked) I should just kill powerd-cli after that [15:07] om26er: how about this: powerd-cli display on cd & ; sleep 60; killall -INT powerd-cli [15:08] kenvandine, yeah, thats a bug [15:09] sforshee, we don't need to keep powerd-cli running for 60 seconds as we generate a touch event with unlock_screen.py so the screen would be on for 60 seconds anyways. [15:09] yep [15:09] om26er: you can also change the activity timeout if you want it longer, although that setting will likely change when the code moves to unity [15:10] mfisch, we won't need to do that since during autopilot test run a lot of touch events are generated so the screen won't turn off [15:11] om26er: perfect [15:16] mzanetti, btw, I've started lp:~unity-team/unity8/ofono (currently has a simple qml plugin that proxies libofono-qt for us) [15:17] jhodapp: hi! [15:17] hey sil2100 [15:17] jhodapp: I have a question related to qtubuntu-sensors [15:17] jhodapp: https://bugs.launchpad.net/qtubuntu-sensors/+bug/1190698 [15:17] Launchpad bug 1190698 in qtubuntu-sensors "should libqtubuntu_sensors.so.1.0 be a private lib?" [Undecided,New] [15:17] sil2100, ok [15:18] jhodapp: what do you think? Is that a private lib or a standard public lib? [15:18] How is it used? [15:18] jhodapp: since I want to clear this out [15:18] (and fix if needed) [15:19] sil2100, sure, let me check (it's been months since I've looked at this code) [15:19] Thanks :) [15:20] sil2100, well it's not intended to be used by any app directly, so I guess in that sense it's meant to be private [15:21] sil2100, but I don't know if it should/could be moved for use with how Qt5 expects things to be set up [15:22] jhodapp: it's a Qt5 plugin, yes? [15:22] sil2100, yeah [15:23] sil2100, it implements some parent classes from QtSensors [15:24] sil2100, so here's what I would say...if you can make it private and it still works with various apps, then that's how it should be [15:25] jhodapp: what current apps use this library? [15:25] sil2100, to my knowledge, only qtubuntu utilizes it directly [15:25] sil2100, though it's been somewhat changed since I last touched it [15:26] sil2100, just open an app that has orientation change working and make sure it still rotates === tvoss is now known as tvoss|dinner [15:41] Hi all [15:42] Is anyone else seeing weird time zone type issues with the "clock" app? [15:42] can you be more specific? [15:42] Sure - so I set the OS level timezone already, per the doc, and the purple initial screen is correct tz [15:42] the clock shows UTC [15:42] If I browse down to world and add my local area (AZ), I get something REALLY random for time [15:43] So the center of the clock shows 00:43 (UTC) [15:43] and the current location shows 12:43AM [15:43] It's 8:43 AM here [15:43] what does "date -u" on the device say? [15:43] ooh, it's showing utc for me too [15:44] Tue Jul 16 15:43:53 UTC 2013 [15:44] * popey files a bug [15:44] thanks for spotting mbybee [15:44] Thanks popey :) [15:45] Loving Touch so far - can't wait to be able to distribute my own apps for it [15:45] I made a quickie one for G+, btw, in the QT creator [15:45] Not sure if you guys noticed, but if you go into Gmail (via app or web), the Drive/Calendar/Mail links work - but G+ just messes everything up [15:45] So I made a dedicated app for G+ - just based off the generic web template [15:46] nice [15:46] I think it's Google fault, actually ;) [15:46] How would I package this to submit it to you guys? [15:46] or would it be a feature request? [15:49] mbybee: https://bugs.launchpad.net/ubuntu-clock-app/+bug/1201858 [15:49] Launchpad bug 1201858 in Ubuntu Clock App "Clock reports time in UTC by default." [Undecided,New] [15:49] could you confirm that please [15:50] I added myself as 'it affects me too' - did you want any other info in there? [15:50] mbybee: qtcreator has an option to create a package. you could put it on launchpad and let us know the link and we could test it [15:50] no, thats fine, thanks [15:51] Ok, I'll poke around in there and find that option. I'm new to the tool, so it'll take me a bit [15:57] stgraber, so i'm looking at your dbus upstart job for X sessions ... why dont we use dbus-launch anymore ? seems doing a dbus-daemon --session makes us miss some stuff [15:58] ogra_: I'd have to dig through what I did a while ago to give you specifics, but the main reason was that dbus-launch wouldn't let us grab the PID and have upstart handle the respawn and re-attach properly [15:59] stgraber, hmm, k ... we are not able to attach to it anymore in the new model [15:59] since ~/.dbus/ doesnt get populated (which i think dbus-launch does automatically) [16:00] (attaching from another session but being the same use doesnt work to i.e. launch apps via ssh) [16:00] s/use/user/ [16:00] ogra_: hmm, DBUS_SESSION_BUS_ADDRESS should be in the environment [16:00] it seemingly isnt for ssh users [16:01] or if it is, it might be wrong [16:01] ah, yeah, I'd indeed expect ssh not to have it set [16:01] seems we used to create ~/.dbus-session in the former iteration of the touch session setup ... [16:02] let me see probably just dumping the vars in there helps [16:03] export DBUS_SESSION_BUS_ADDRESS=`sed 's/.*DBUS_SESSION_BUS_ADDRESS=//' /proc/$(pidof unity8)/environ | sed 's/UPSTART.*//'` [16:03] or we could generate the .dbus/... stuff, it's just a file that's - (with the latter not existing in Mir so not too sure what libdbus will try to use) and contains DBUS_SESSION_BUS_ADDRESS [16:03] does wonders :) [16:03] haha [16:03] kenvandine: initctl list-env | grep DBUS :) [16:03] yeah [16:04] hmm [16:04] root@ubuntu-phablet:/# sudo -u phablet -i [16:04] phablet@ubuntu-phablet:~$ env|grep DBUS [16:04] DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-NYQGkgcmzr,guid=04a8863d22e6ea4a8624ee5d51e4059c [16:04] phablet@ubuntu-phablet:~$ [16:04] works via adb [16:07] hmm [16:07] phablet@ubuntu-phablet:~$ env|grep DBUS [16:07] DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-NYQGkgcmzr,guid=04a8863d22e6ea4a8624ee5d51e4059c [16:07] phablet@ubuntu-phablet:~$ [16:07] i also have it via ssh [16:08] i dont get why kenvandine needs to export it separately [16:08] initctl list-env fails too [16:08] initctl: Rejected send message [16:09] as the phablet user [16:09] yeah, i have the same here [16:09] that rather points to upstart then ... [16:10] but the sed trick in my .bashrc works fine :) [16:10] yeah, but thats not a fix werer can ship ... .bashrc needs to be the default again :) [16:11] without all our hackery [16:11] -rer (i didnt type that, wheer did that come from) [16:19] aha [16:19] phablet@ubuntu-phablet:~$ env| grep XDG_RUNTIME_DIR [16:19] phablet@ubuntu-phablet:~$ [16:19] thats the issue i guess [16:20] hmm, even exporting it doesnt make initctl work [16:22] To restore Android the command ./flash-all.sh does not work | http://askubuntu.com/q/320704 [16:22] HAHAHAHA ! [16:23] i think i found it ... [16:23] a simple typo [16:24] hmm, no, thats not enough [16:24] though that typo must go [16:25] Ok - does anyone have a couple minutes to help a QT noob figure out how to deploy a QT package to launchpad? The publish is grayed out, build is grayed out, test/debug/run on ubuntu touch all work fine [16:25] Google isn't being helpful for me lol [16:35] popey: is your phone up and running now? [16:35] yes [16:36] popey: if you take a photo you can scrap please, then goto into the gallery select the photo open the hud and select auto enhance what happens to the photo [16:36] known bug [16:36] popey: ah cool [16:36] guess who filed it ㋛ [16:36] https://bugs.launchpad.net/gallery-app/+bug/1180345 [16:36] popey: your kids [16:36] Launchpad bug 1180345 in gallery-app "Auto enhance crashes gallery app and deletes photo" [Critical,Confirmed] [16:37] hehe [16:37] rsalveti, one for you https://code.launchpad.net/~ogra/session-manager-touch/upstart-session-typo/+merge/175066 [16:37] (still doesnt fix anything, i think the whole setup is a bit weird) [16:39] popey: I know they do your testing for you really ;) [16:39] this is true [16:39] popey: when does the minecraft app land ;) that's your kids happen then right well as long as they can get you off the phone and tablet :D [16:39] katie, poke about demo/discoverable first boot bits. Was there any further thoughts on "what happens if user swipes into their session during demo?" [16:39] s/happen/happy [16:41] ogra_: happroved [16:41] thx [16:41] ogra_: will you merge && push? [16:43] rsalveti, there are more issues [16:44] rsalveti, but yeah, will do [16:44] katie, oh, nm. I see that we expect the user to swipe into session now [16:44] mterry, hi, yes [16:44] ogra_: more? :-) [16:45] mterry, so its a contained journey - they can't exit without clicking on the 'exit' or skip intro text [16:45] mterry, did you see the video? its a link from the spec [16:45] rsalveti, XDG_SESSION_DIR doesnt end up in the env [16:45] katie, yup, makes sense. And launcher/indicator knowledge will extend to greeter [16:45] rsalveti, i fear thats the cause of all the test failures ... [16:46] katie, watching [16:46] mterry, great [16:46] rsalveti, also /var/log/upstart/ubuntu-touch-session-setup.log has: "mkdir: cannot create directory '': No such file or directory" [16:46] ogra_: hm, right [16:47] that whole two fold upstart job thing looks pretty weird to be honest [16:47] I think that was just following what was supported on upstart, not sure [16:47] ricmm: ^ [16:48] ogra_: http://upstart.ubuntu.com/cookbook/#non-graphical-sessions-ubuntu-specific [16:49] * ogra_ reads [16:49] ah, yeah, k [16:49] yay, so the error from the log is gone at lest [16:49] *least [16:50] but i still dont get XDG_SESSION_DIR in my env === edulix2 is now known as edulix [16:58] wowm so starting clock-app from cdmline while some other app is running in teh shell makes the terminal get stuck , i cant even ctrl-c ... until i close the running app in the shell [16:58] (and clock-app doesnt come up) === dholbach_ is now known as dholbach [16:59] * ogra_ wonders if thats related to what sergiusens told today [17:00] hmm, if i dont have any app running in the shell i cant get my terminal back in adb [17:00] ogra_: are you launching from an adb shell? [17:00] yes [17:01] i can get my prompt back if i start another app on the shell and stop it [17:01] ogra_: the not showing up is related to what I mentioned... you also need to add a -desktop_file_hint=[path to desktop file] [17:01] yeah, i'm used to that last bit [17:01] * sergiusens doesn't recall the extact switch, it's too long [17:02] its dashes, not underscroes i think [17:02] ogra_: it should still be in your Running Apps [17:02] ah, no, actually underscores are right [17:02] ogra_: give me 5 as im in the car [17:02] * ogra_ gives ricmm 5 [17:02] :P [17:03] if not, app lens where you can focus it, if it's not in focus the appmanager sigstops the app [17:03] and its --destop_file_hint= [17:03] stop IRCing while driving ! [17:06] yeah [17:06] popey: ping [17:06] doesnt change a thing in behavior [17:06] and i dont see it in the running apps list [17:06] in a taxi :) [17:06] nik90: pong [17:06] popey: I saw your bug https://bugs.launchpad.net/ubuntu-clock-app/+bug/1201858 [17:06] Did you set the correct location in the clock app? [17:07] nik90: actually no, there seems to be an issue there [17:07] nik90: there's no way to add a location [17:07] nik90: and there's something wrong with the tabs along the top, i can't switch to timer/stopwatch [17:07] ogra_: do you have the clock installed, can you test that? [17:08] popey: Really? Since when did switching tabs break? [17:08] I did not commit anything recently which could affect that [17:08] jodh, are you still around ? i have some issue with http://upstart.ubuntu.com/cookbook/#non-graphical-sessions-ubuntu-specific ... (XDG_SESSION_DIR doe not end up in the users env) [17:08] only just noticed nik90 [17:09] popey, works fine here [17:09] popey: Is this happening on your phone or the desktop? [17:09] phone [17:09] because I cannot reproduce it [17:09] hmm [17:09] * popey reboots [17:09] it is extremely slow though [17:10] oh, because ueventd acts up [17:10] * ogra_ reboots too [17:10] ogra_: which device are you trying it? [17:10] pfft, works now [17:10] maguro [17:10] the galaxy nexus? [17:10] yeah [17:10] the tabs are all fine [17:10] ogra_: yeah we need to do lot of performance optimisation for the clock app. colin king already created a bug report about it [17:11] apart from alarm telling me it isnt implemented, everything seems to work just right [17:11] tackling it now [17:11] odd i cant add a city now nik90 [17:11] i searched for london, found it, tapped it, it goes back to the clock [17:11] and doesn't update [17:11] nik90: lemme do a video for you [17:12] popey: you need to look at the world cities and then click it again to set it as your current location [17:12] tapping it in the search results will only add it to the list [17:12] it doesn't show up [17:12] i.e. it doesn't get added to the list === alan_g is now known as alan_g|EOD [17:12] oh, thats supposed to work ? [17:13] hasnt worked for me ever [17:13] damnit, kids just came home [17:13] going to be a noisy video [17:13] hmmm, this phone is whacky [17:13] it's been sat at the google screen for ages [17:13] ogra_: that is supposed to work since 2 weeks ago actually. But will take a loot [17:14] i'm stuck with only UTC in the list [17:14] also if i scroll up to see the list it is hard to scroll down again [17:14] Hi, what is the official Contact API if I want to access Contacts from my Ubuntu Phone application? [17:14] i have UTC and Er in mine [17:15] ogra_: ok im back [17:15] the clock app is showing UTC by default for everyone since it cannot detect the user's location. [17:15] so whats sthe issue? [17:15] nik90: right, but i cant add a city from the list of cities to the main screen [17:15] but now, my phone won't boot [17:16] popey: is there a way you can try to remove the local storage database and try later? [17:16] i will when i get this thing booted [17:16] okay [17:16] Because on my side I am perfectly able to add a world city and then set it as my current location [17:16] ricmm, well, dbus not being connectable if you log in via adb or ssh ... but it seems the dbus address is in the env ... what isnt in the env is XDG_SESSION_DIR ... not sure you can help with that, i suspect thats an upstart issue [17:18] jodh: ping [17:19] ogra_: well we manually set up XDG_RUNTIME_DIR, not SESSION_DIR [17:19] this is usually all handled by the pam login [17:19] ricmm, yeah, well, still doesnt help, initctl doesnt work [17:19] ok, this is odd. the device wont boot and it's gona back to giving me a red light [17:19] ogra_: initctl works for me [17:20] like it did this morning, but it's been charging all day, can't imagine it's dead battery [17:20] what exactly are you trying to do? [17:20] if I ssh in, not adb [17:20] ricmm, http://paste.ubuntu.com/5881553/ [17:20] ogra_, ricmm: [17:20] you're trying to connect to the upstart user session with initctl? [17:20] wrks with neither for me === dandrader is now known as dandrader|afk [17:20] ah list-env [17:20] slangasek, thats what stgraber suggested to get the dbus pid [17:20] Hey guys, can anyine give me the full bug list for the Galaxy Nexus (maguro) [17:20] slangasek, for the session bus [17:20] ah, ueventd killing cpu [17:20] ogra_: well, that only helps if you're already inside the session ;) [17:21] youd need to export UPSTART_SESSION [17:21] from list-sessions [17:21] popey, yeah, seems it got worse ... i never had it on maguro, now its there every second boot [17:21] [color=red] anyone? [/color] [17:21] slangasek, well, i'm the owning user of the session (the actual issue that got me here is that autopilot cant connect to dbus) [17:22] since we switched to upstart [17:22] you own it but that doesnt mean you are in it [17:22] Strike_Riki: the bugs are spread over a number of apps, what specificially you looking for? [17:22] right - UPSTART_SESSION=$(initctl list-sessions | awk '{ print $NF; quit }') initctl $stuff [17:22] you need to join it if you want to list the env and export it, I guess [17:22] Can i get basic functionality? e.g. GSM, WiFi, Sound, and many of the other core functions [17:23] however, it becomes questionable what the right behavior here is if you ever have more than one upstart session on the system [17:23] the session does have dbus and all the stuff you want, you just arent in it if ssh'ing in [17:23] yes Strike_Riki [17:23] not a problem on the phone, will be a problem down the line [17:23] slangasek: is there a nice way to make all ssh sessions spawn inside of the upstart session by default? [17:23] but "many of the other core functions" is very vague [17:23] ricmm: no [17:24] Strike_Riki: https://wiki.ubuntu.com/Touch/Devices/Nexus gives current status [17:24] ok [17:24] There is a video posted on youtube by Jono Bacon showing the latest updates. However, i cannot find a place to download it. [17:24] Strike_Riki: https://wiki.ubuntu.com/Touch/Install [17:24] well I guess we could hack it into .bashrc [17:24] ricmm: you can do not-nice things with /etc/profile, I guess? [17:24] or that [17:24] Strike_Riki: that video is already outdated [17:24] slangasek, there is only one user session by user usually though no? [17:24] /etc/profile.d/phablet-be-crazy.sh [17:24] It is? [17:25] How close it to being bugless? [17:25] Strike_Riki: no software is bugless [17:25] slangasek: right now we already have some ugly stuff in bashrc, maybe we can keep it there ;) [17:25] seb128: that's not a safe assumption, particularly if we ever switch to having upstart sessions for console logins [17:25] ricmm: fair enough :) [17:25] and yes, we will go multi user later via lightdm to handle user logins [17:26] Well, i mean no noticeable bugs. I have looked at ther spread sheet and a lot of things are working. But aside from the core functions, i have 2 more questions. [17:27] Can i install android apps on there? [17:27] slangasek, ogra_, ricmm: we should just make the dbus script dump the env in e.g ~/.cache/upstart/dbus-env or something and then we can source that file from .bashrc [17:27] Strike_Riki: no [17:27] And which apps have the most bugs? [17:27] seb128, ++ [17:28] How to Install UBUNTU on a SHV-E250L? | http://askubuntu.com/q/320736 [17:28] that involves overriding the normal session dbus script [17:28] seb128, thats what i was planning ... but i *see* the dbus address in my env [17:28] Strike_Riki: i dont know, you'd have to go through each app and see [17:28] Strike_Riki: things change daily [17:28] ricmm, nah, just dumping the vars in place [17:28] even if we have pam login and lightdm we will still need ssh access to join the session [17:28] Alright. Do i need a different kernel to run it? === Zic_ is now known as Guest44134 [17:28] ricmm, worse is that autopilot actually seems to run via adb [17:29] which doesnt process pam at all [17:29] Strike_Riki: everything you need is downloadable from https://wiki.ubuntu.com/Touch/Install [17:29] Alright, thanks for the link.# [17:29] ogra_: hmm, that's new to me, we used to do ssh to get a fully configured session in the proper namespace for autopilot [17:29] Is there an app-store? [17:30] not yet [17:30] sergiusens, ah, then i'm wrong [17:30] SO i can only use thhe pre installe dapps? [17:30] sergiusens, i thought it used adb all the way [17:30] So* Installed* [17:31] well, if you want to source the dbus addr you'd need to make the dbus.conf (I would do it in unity8.conf, which we ship) [17:31] sergiusens, then i dont get why the phablet user isnt used though [17:31] Strike_Riki: no, you can install your own apps or write your own [17:31] dump it into a file [17:31] you can then source in bashrc [17:31] pretty ugly, might as well just do the list-session and join it in bashrc, then parse and export [17:31] if we are going the ugly way ;) [17:31] ricmm, right, but it seems i dont need to, i have it in my env ... [17:31] ogra_: there's an phablet-test-run in phablet-tools [17:31] the point is that i still cant start apps [17:32] sergiusens, and thats whats run for the dashboard tests ? [17:32] DBUS session is your env? [17:32] how? [17:32] phablet@ubuntu-phablet:~$ env |grep DBUS [17:32] DBUS_SESSION_BUS_ADDRESS=unix:abstract=/tmp/dbus-NYQGkgcmzr,guid=04a8863d22e6ea4a8624ee5d51e4059c [17:32] dunno [17:32] lol [17:32] from the upstart dbus,conf i guess [17:32] nik90: whereabouts is the database used by the clock? [17:33] you exported the list-env from the session? [17:33] otherwise you shouldnt have it, I dont [17:33] Can i install a apk? [17:33] no, thats a fresh boot [17:33] fresh ssh session [17:33] Strike_Riki: you already asked if you could run android apps and I said no. so installing apk.. no. [17:33] ogra_, but is that the right address? [17:33] * ogra_ checks bashrc ... we might have made a mess in that in the past [17:33] or is that from a left over .dbus-session file? [17:34] kenvandine, good question [17:34] ir probably is [17:34] i had a stale file [17:34] yeah [17:34] ogra_: that's out of my reach [17:34] Can i convert an app to an installeable? [17:34] let me check if the file gets surced by bashrc [17:34] ogra_: do initctl list-env and export that address, you should be able to launch apps [17:34] doanac: hey, regarding tests, ogra_ and me would like to know how they were run [17:34] Strike_Riki: depends on the app [17:34] # export the phablet shell dbus session on normal login shells [17:34] source /home/phablet/.dbus-session [17:34] export DBUS_SESSION_BUS_ADDRESS [17:34] what I'm saying is lets join upstart and export the dbus session in bashrc [17:34] sigh [17:34] ricmm, well, lets just update the file :) [17:34] popey: hmm..in the desktop it is under .local/share/Qt Project/QtQmlViewer/QML/OfflineStorage/Database === vanhoof_ is now known as vanhoof [17:35] popey: here is where all core apps using sqlite storage store their databases in. [17:35] hmm [17:35] ogra_: if you are going to do that, do it in unity8.conf [17:35] not all ☻ [17:35] which we are currently shipping [17:36] and be done [17:36] nik90: /home/phablet/.local/share/webbrowser-app/ for example [17:36] popey: only core-apps though, clock, calendar, weather [17:36] right [17:37] * popey deletes and reboots === d_kessel is now known as dkessel === VargaD_ is now known as VargaD [17:37] ricmm, either that or ship a dbus.override ... i'm undecided [17:37] unity8.confis better than a new file [17:37] gah [17:37] ricmm, ogra_: no [17:37] every time I reboot, the clock flips back to 1970 [17:37] otherwise, do it in bashrc by joining upstart [17:37] popey: the system time? [17:38] seb128, what else ? [17:38] I'm still vouching for doing the right thing in bashrc that pam would do [17:38] ricmm, ogra_: do it in /usr/share/upstart/sessions/dbus.conf imho ... that's where we set it [17:38] which is join the upstart session [17:38] yes [17:38] and set up the env from there [17:38] it fixes after a few seconds [17:38] seb128, right, so i shoudl fix it upstream ? [17:38] seb128: you set the env there, but not write it to a file [17:38] that would be an override from us [17:38] guys... it is not broken [17:38] ricmm, right, just add the line to dump it to a file in there? [17:38] ogra_, "upstream"? do it in ubuntu... [17:38] the session is indeed present in the upstart session [17:38] ogra_, or who do you call upstream? [17:39] we just happen ot need it in our ssh sessions :) [17:39] seb128, does desktop still use ~/.debu/session/ ? [17:39] dbus.conf only exports it in the env, it sets up no file [17:39] seb128, upstream as in the dbus package [17:39] ogra_, desktop uses upstart yes [17:39] nik90: all working fine after deleting that database [17:39] seb128, so you likely want that file structure too ? [17:39] popey: you scared me for some time :) [17:39] ricmm, right, I was suggesting to add 1 line to write the env to a file as well [17:39] nik90: sorry ☻ [17:40] ogra_, well, I'm fine having a file with the env, maybe better to use .cache/upstart/dbus though [17:40] ok, i'll start with a simple file then [17:40] ogra_, no reason to pollute ~ [17:40] well, we have the reading stuff in bashrc already [17:40] shipping an override is pollution imo [17:40] thats why I said just do it in bashrc :_ [17:41] you dont need to read a file, just join the session [17:41] ricmm, no override [17:41] popey: anytime I make changes to the database, I am not exactly sure how to restart the user's database. So this might occur again as I update the database [17:41] we gain something from that too, you need to join the session to launch apps with upstart-app-launch [17:41] ricmm, directly in the .conf [17:41] we dont ship that .conf [17:41] ricmm, "fixing" dbus :) [17:41] I think exposing the session in a file requires broader discussion [17:41] this is the same dbus job used in desktop [17:42] root@ubuntu-phablet:/# dpkg -S /usr/share/upstart/sessions/dbus.conf [17:42] dbus: /usr/share/upstart/sessions/dbus.conf [17:42] fixing *dbus* [17:42] ;) [17:42] its not broken! [17:42] popey i would want to convert Tapatalk or the XDA App and Google Hangouts/Talk. Can i do that/ [17:42] ?* [17:43] ricmm, normally dbus-launch creates ~/.dbus/session/..... [17:43] Strike_Riki: if you're a developer, sure. [17:43] sure but our desktop no longer uses that [17:43] ricmm, we srtopped using dbus-launch [17:43] hmm, rild using 100% cpu [17:43] How would i go about doinng it? [17:43] anyways, up to you how you want to do it [17:43] but we *need* to be in the upstart session by default, for ssh and adb [17:43] so it might be better to just do it in bashrc [17:43] :) [17:43] Strike_Riki: You need to be a developer, and from your questions you don't sound like one. [17:43] two birds one stone [17:43] phone crashed after rild went wild [17:44] ricmm, all i want is to be able to have autopilot not fail all tests due to missing dbus address [17:44] Cheers ;) i have mad a few ROM's, never played with Ubuntu Touch though. [17:44] ricmm, and the dbus package owns the file we want ... i could make it conditional on the session that runs if the file is created or not [17:45] dbus spawns before the DESKTOP_SESSION is set, so no [17:45] Is there no way to sideload an app? [17:45] bachrc *has to* get rid of all these awful hacks [17:45] Strike_Riki: what kind of app? [17:45] Strike_Riki: Android ROM "development" is nothing compared to what is happening here [17:45] we dont want that file .dbus/session/ is legacy [17:45] The apps mentioned [17:45] sergiusens: getting back from lunch. you want to know how we launch the autopilot test in our smoke jobs? [17:46] ogra_: I agree, the test scripts need to set their env correctly by joining the session [17:46] ricmm, bashrc cant be the answer [17:46] Strike_Riki: dude. that's three times now. [17:46] ogra_: but seems like we have to do it [17:46] Strike_Riki: no, you can't install android apps on ubuntu touch. [17:46] hacking up shell defaults cant be an answer [17:46] no, I completely agree that the hack needs to go away [17:46] but it is the same way it would work on your desktop [17:46] ricmm, no, we have to do it right :) [17:46] ogra_: what happened to the dashboard stuff? [17:46] if you setup a ssh server and login to it, you wont see your upstart session [17:46] or dbus [17:46] did it turn out that we have big regression in dbus/image? [17:46] okay. Thanks. [17:46] who fixes the stuff? [17:46] asac, dbus issues due to the new upstart session [17:46] asac, i'm on it [17:47] once we have agreed on a fix :P [17:47] asac: this is related to the testing environment [17:47] right. [17:47] do you talk to doanac? [17:47] we removed hacks to facilitate people's life, now we are putting them back in somewhere else [17:47] ricmm, i think i'll go with seb128's suggestion and just do it in dbus [17:47] how about just making it work now and then landing a clean fix at best using a qa team maintained abstraction tool :) [17:47] :) [17:47] ok i am out for a bit [17:47] so no hurry :) [17:48] no idea what you mean with that tool ;) [17:48] ok :) [17:48] ricmm, i cant imagine that desktop doesnt want that as well [17:48] well frankly I'm waiting for lunch in my table, so do it in dbus [17:48] in which case the dbus job is just the right place [17:48] i am sure you know what i mean ... something that always just does the right thing [17:48] and that we use both for local dev as as in the lab to do the duty [17:49] of the main use cases [17:49] like phablet-network-setup is one example [17:49] phablet-test-run would be another option etc [17:49] ok out [17:49] 1-2h [17:51] anyone else see 1970 on the welcome screen on boot? https://bugs.launchpad.net/touch-preview-images/+bug/1201910 [17:51] Launchpad bug 1201910 in touch-preview-images "Date is set to 1970 on boot" [Undecided,New] [17:51] popey: yes. connect to network and reboot, to make ntpdate kick in. [17:51] ogra_: lunch almost here, will you do it in dbus then? [17:51] the hack will remain in bashrc either ways [17:51] xnox: actually it is updated by the time i shell in and run the date command [17:51] ricmm, yeah, and ask seb128 for review [17:52] xnox: but i dont like seeing 1970 when I boot the phone and demo to people ☻ [17:52] ogra_: does bug #1201865 make sense to you? I don't see any ubuntu-touch-session-phablet job in today's image [17:52] bug 1201865 in upstart "upstart on nexus4 is frequently writing :sys:power_supply-device-changed event messages" [Undecided,New] https://launchpad.net/bugs/1201865 [17:52] popey: if one flashes the device from cold boot (battery died to the point of not being able to boot the device) and boots, you will briefly see the wrong time. [17:52] ogra_: we need to join the default sessions to upstart somehow, or provide helper tools like asac said for launching apps [17:52] xnox: we should never show 1970 [17:52] ogra_: so either we do a helper or add the next hack to bashrc [17:52] we can talk about it after I have lunch [17:52] it's a known date that we could omit displaying [17:52] popey: that's why be default, ubuntu desktop images boot without clock visible / without datetime indicator ;-) === ricmm is now known as ricmm||lunch [17:52] hah [17:53] ricmm, well, lets discuss that later, i want to stop working after the upload ... [17:53] ricmm||lunch, and we should also involve foundations and desktop into that discussion [17:53] popey: it's small things, and typically people do not notice the things that are missing. hence it's the dirty little secret of live-cds ;-) [17:53] ogra_: I thought thats what we have been doing just now [17:54] ricmm||lunch, just because seb128 discusses with us doesnt mean cross team consensus :) [17:54] I'm all for 0 hacks in bashrc, nothing in dbus.conf and proper env setup by whoever wants to use the system [17:54] ricmm||lunch, i bet slangasek might have a different view :) [17:54] technically we are just enabling a hack for fast development, he already said there is no better way to do it [17:55] either profile.d or bashrc [17:55] slangasek, the bug is fine, thats the new upstart session [17:55] if you want to wait for the upload, ship it in dbus for now [17:56] ogra_: how? the only upstart job I see in today's image is /etc/init/ubuntu-touch-session.conf, which does not log to /var/log/upstart/ubuntu-touch-session-phablet.log [17:56] slangasek, http://upstart.ubuntu.com/cookbook/#non-graphical-sessions-ubuntu-specific ... seems when the session is called with the user the username gets appended to the logfile [17:56] ah [17:56] wait [17:56] how? [17:56] that only happens for jobs which specify an 'instance' [17:57] /etc/init/ubuntu-touch-session.conf does not [17:57] there is another one, forgot its name [17:57] regardless, this aggressive output to the log is a bug in whatever's logging [17:57] ogra's link points to what james hunt advised us to do [17:57] we have the exact same arch defined there [17:57] and yes, the problem is aggressive logging [17:57] we have had that one for a while to tackle ;) [17:58] ok lunch or my wife will dump me [17:58] none of that explains why cking is seeing a /var/log/upstart/ubuntu-touch-session-phablet.log [17:58] the ubuntu-touch-session job in ubuntu-touch-session 0.56 will *never* log to that log file [17:58] 0.58 is currently in the image [17:59] hmm [17:59] well, i have that logfile here and it is updated every time i start the session [17:59] why do I not have 0.58, if I just phablet-flashed [17:59] :/ [17:59] * slangasek tries again [17:59] slangasek: maybe image failed to flash somehow [17:59] apparently, grr [18:00] oh, "not enough space in /data", that's interesting [18:00] yeah that gets me all the time [18:00] it's also incorrect [18:00] more like annoying, has anyone got the defacto solution to when one sees that "not enough space in /data" error? [18:01] what do you mean by defacto? cleaning up old builds? [18:01] I delete the old zips [18:01] we could append something to remove old files, like the old images, but kind of scary [18:01] I don't have old zips; I just have a system that's lying :) [18:01] what we can do is self-erase the zips after the flash is done [18:01] /dev/mmcblk0p23 4.6G 1.9G 2.7G 41% /data [18:02] I think it requires 3G [18:02] slangasek: 3 is the cutoff [18:02] get a new phone [18:02] meh, why is the error message reporting the used size? [18:02] if you don't care about your files in there, phablet-flash -b will do the work for you [18:02] ogra_: it's an N4. :P [18:02] cking: logs use to pile up too [18:02] yeah, you need an N4.5 [18:03] :P [18:03] slangasek: a better error message would help indeed [18:03] like flash is a free resource to fill ;-) [18:03] and the space in /data is limited because of the move in progress to the system partition ;) [18:03] seb128, http://paste.ubuntu.com/5881662/do you think thats ok ? [18:04] ogra_, do you need to use that filename for compat reasons? [18:04] popey: regarding bug 1201910, do you get that at every flash? [18:04] ogra_, I would prefer it under .cache/upstart if possible, to not clutter ~ [18:04] bug 1201910 in touch-preview-images "Date is set to 1970 on boot" [Undecided,New] https://launchpad.net/bugs/1201910 [18:04] seb128, well, we source it from bashrc currently [18:05] ok [18:05] rsalveti: only just noticed it today [18:05] when a friend came over and I demo'ed the phone ☻ [18:05] popey: clean phone? after flashing android? flashing with -b? [18:05] not -b, just a reboot [18:06] it's been flashed many times [18:06] oh, then it might not be storing the date correctly [18:06] as it gets the right time later on once you're connected due ntp [18:06] just did it again [18:06] http://popey.com/~alan/device-2013-07-16-190638.png [18:06] seb128, and specific filename you would like ? [18:07] alan@deep-thought:~/phablet/scripts$ adb shell [18:07] root@ubuntu-phablet:/# date [18:07] Tue Jul 16 19:06:57 BST 2013 [18:07] ogra_, no [18:07] or is dbus-session fine [18:07] k [18:07] popey: right, might be an issue with mako specifically [18:08] will try to reproduce [18:08] popey: was that after running completely out of battery? [18:08] http://popey.com/~alan/device-2013-07-16-190734.png time updated in the indicator, then *just* after I took that screenshot, the welcome screen updated to http://popey.com/~alan/device-2013-07-16-190820.png [18:08] no [18:08] just after an "adb reboot" [18:08] hm, right [18:08] failed to copy '/home/vorlon/Desktop/phablet-flash/ubuntu-touch/20130716/saucy-preinstalled-touch-armel+mako.zip' to '/sdcard/': Is a directory [18:08] 20130716, 180801, Charging, Good, 98, 376, 4360000, 3200000, 4336155 [18:08] battery is pretty full [18:09] do I not have a current version of phablet-flash? [18:09] popey, what happens if you suspend/resume [18:09] pmcgowan: via what method? tap power? [18:09] sure [18:09] * popey reboots to test [18:09] slangasek: you seem to, that's just the result of an 'adb push /home/vorlon/Desktop/phablet-flash/ubuntu-touch/20130716/saucy-preinstalled-touch-armel+mako.zip /sdcard/' [18:09] I think it will re-read the time info [18:09] not reboot [18:09] sergiusens: yes, and I have no /sdcard/ [18:10] i need to reboot to get back to 1970 [18:10] so is my filesystem broken, or is my phablet-flash broken? [18:10] [18:10] slangasek: were you playing with the ro fs work? [18:10] back to the future [18:10] slangasek: filesystem [18:10] Do I need to bite the bullet and phablet-flash -b? [18:10] heh [18:10] rsalveti: no [18:10] no [18:10] that doesn't fix it [18:10] slangasek: you can also from recovery do a 'format userdata and /data/media' [18:10] ok, phablet-flash -b it is [18:10] slangasek: no [18:10] slangasek: -b might indeed help, but interesting that recovery might have failed to mount that partition for some reason [18:11] slangasek: it will potentially break your partitioning [18:11] rsalveti: oh. it wasn't in recovery. [18:11] http://popey.com/~alan/device-2013-07-16-191046.png <- thats after a reboot, then once the welcome screen appeared, I tapped power, then waited then tapped again to wake [18:11] when I 'adb shell'ed in, it was Ubuntu. [18:11] then 'adb reboot recovery' failed [18:11] slangasek: is your image old? [18:11] sergiusens: clearly ;) [18:11] might be he has the broken getprop [18:11] I don't have any getprop [18:11] so reboot into recovery [18:11] yeah, that's it [18:11] manually [18:11] and from there do [18:12] popey, will try on the latest image once its downloaded, have not seen that [18:12] phablet-flash -d mako [18:12] k, easy to reproduce [18:12] popey: flashing with -b to see here [18:13] * popey does it to grouper to see if it does same [18:13] http://popey.com/~alan/device-2013-07-16-191333.png [18:13] yes happens on grouper too [18:13] but goes back to 2000 not 1970 [18:15] popey: that's probably the hardware default for initial clock [18:15] yeah [18:16] popey: please run: adb reboot; adb wait-for-device; adb shell date [18:17] here I always get the right values (near the values used before reboot) [18:18] crw-rw---- 1 system system 253, 0 Jul 16 18:16 /dev/rtc0 [18:18] rsalveti, i think his RTC battery drained due to draining the system battery heavily [18:18] k [18:18] ogra_: yeah, but it seems to be happening with his grouper as well [18:18] so in case i am brave and boot with mem=512m (how do i do that?) ... and the system fails miserable (like shell gets killed because of OOM etc.) [18:18] its ~98% now though === barry is now known as bobdobbs [18:18] will i be able to get back to normal boot easily? [18:18] dbus fix uploaded ... should be fine with tomorrows build === bobdobbs is now known as barry [18:19] asac: just need to change the boot args, with abootimg [18:19] asac, abootimg ... [18:19] alan@deep-thought:~$ adb reboot; adb wait-for-device; adb shell date [18:19] Thu Jan 1 07:29:27 BST 1970 [18:19] rsalveti: ^ [18:19] popey: mind pasting your dmesg? [18:20] sure [18:20] http://paste.ubuntu.com/5881707/ [18:20] [ 2.538989] rtc-pm8xxx rtc-pm8xxx: setting system clock to 1970-01-01 06:29:23 UTC (23363) [18:20] yeah [18:21] ogra_: is that in archive? [18:21] asac: yes [18:21] asac, uploaded a second ago [18:21] abootimg -x [18:21] hmm [18:21] so i do: [18:22] oh, abootimg you mean [18:22] change the config to include the mem=512 argument [18:22] i dont know :) [18:22] rsalveti, huh ? [18:22] boot image is /boot/an*.img ? [18:22] then create it with 'abootimg --create saucy-preinstalled-boot-armhf+maguro.img-new -f bootimg.cfg -k zImage -r initrd.img-touch [18:22] seems not [18:22] ogra_: just to change it in his host machine [18:22] abootimg -i $(find /dev -name boot|head -1) [18:22] its only on my phone [18:22] do it on the device [18:23] no need for that bootimg nonsense [18:23] what shall i hack then ogra_ ? [18:23] you can just change it on the fly [18:23] isw there a config file + update-* [18:23] no [18:23] you mean on the phone? [18:23] ogra_: just so if in case he wants that to be auto generated when testing the images [18:23] interrupting boot? [18:23] yeah [18:23] asac, adb in ... then run the above command [18:23] well i want to see what happens for now :) [18:23] that shoudl spit our the current config [18:24] (from running ubuntu) [18:24] abootimg -i $(find /dev -name boot|head -1) ? [18:24] how do i change the kernel line? [18:24] first run that command [18:25] (patience ... ) [18:25] do you get proper output ? [18:25] ok [18:25] have it [18:25] abootimg -u $(find /dev -name boot|head -1) -c "cmdline=mem=512m" [18:25] http://paste.ubuntu.com/5881725/ === A-TCP-BreAKDoWN is now known as An-IP-BreAKDoWN [18:25] that sets mem=512m [18:25] then reboot [18:26] abootimg -u $(find /dev -name boot|head -1) -c "cmdline=" [18:26] that sets the empty cmdline back [18:26] popey: wonder if that is the side effect of the udev rules [18:26] 70-mako.rules:ACTION=="add", KERNEL=="rtc0", OWNER="system", GROUP="system", MODE="0660" [18:26] which changes the permission and such [18:26] under /lib/udev/rules.d [18:27] there's no need to change stuff for rtc (in theory) [18:27] asac, and rsalveti is indeed right, if you want to autogenerate that for test images you better use abootimg on a boot.img file before flashing it [18:27] for changing it on a running img the above is better though [18:27] Hello , [18:28] * ogra_ goes afk for a while [18:30] popey: can you check what happens after running reboot from the device? [18:30] instead of using adb [18:30] rsalveti: looks good... rebooting :) [18:31] ogra_: errr you [18:31] adb reboot will just reboot the device directly, without setting anything up [18:31] guess it will fail miserably given that almost 300M are GPU [18:31] can we force making that smaller? [18:31] asac: I'm not sure the gpu memory is shared with the system there [18:31] i guess an SoC WITH 512m would rather ship a bit least aggressive GPU [18:31] rsalveti: free shows 710m [18:31] I believe it might be a pre-allocated part of it [18:31] so i guess its GPU [18:31] right, that's why [18:32] rsalveti: wonder... if you have a 512m SoC how much would GPU eat at max? 128? [18:32] 512m would probably give ~400mb to the system [18:32] if so we could boot with adjusted mem [18:32] to get a better approx about our main mem [18:32] that depends, just get an android device which has 512mb [18:32] so the log-in screen is there [18:32] let me see [18:32] if that actually did anything [18:33] mem=... didnt work [18:33] http://paste.ubuntu.com/5881750/ [18:33] it was seenb by kernel [18:33] but free is not believing us [18:33] hackish kernel? [18:34] mem=1G vmalloc=768M [18:34] hmm. [18:35] not sure how to get that away [18:35] not sure if that's already coming from the bootloader [18:35] guess we want vmalloc=512m or 386m maybe [18:35] guess so [18:35] would kernel parse last or first mention of a parameter? [18:35] not sure, would need to see the code [18:36] apw: rtg: might be able to help you on that [18:36] asac, i would say that most parameters just set some global [18:36] so in the common case the last one most likely [18:36] of course some are special and the first locks, so it cannot be guarenteed either way [18:37] yeah i try using vmalloc=512m now [18:37] lets see [18:37] no luck [18:37] also we have swap [18:37] which i dont really want [18:38] popey: so in theory rebooting via 'reboot' should set up your clock correctly [18:38] are we waiting for lifecycle before we kill swap? [18:38] ok, will try now [18:38] popey: adb reboot breaks that logic and makes the hardware to keep whatever it was using before [18:38] swap isnt that big anyway... can we just kill it? [18:38] rsalveti: reboot as root? [18:38] popey: your hardware is not setting to the same date/time at every reboot, right? [18:38] takes a while.. [18:38] popey: yes [18:38] ok, rebooting [18:38] popey: I know, but at least once :-) [18:39] yes, that came up with the right time [18:39] asac: not so sure, swap was useful when using browser [18:39] swap fundamentally affects the kernels ability to cope when we have too little memory, even if it is tiny [18:39] but that was a workaround for the mwc demo [18:39] ideally we'd have no swap [18:39] rsalveti: well, we certainly don't want to hide that browser is going beyond budget :) [18:40] popey: cool, then just run adb reboot to see [18:40] maybe a demo flag :) [18:40] k [18:40] asac: sure [18:41] rsalveti: adb reboot came back with the right time [18:41] popey: right, then the reason is that /etc/init/hwclock-save.conf is only called during reboot [18:41] with timed that might need to change [18:42] popey: so you can either close the bug or change that to be 'date/time not set right after setting the system clock' [18:42] the hardware one, I mean [18:42] k [18:42] brb [18:43] Hey all. I have a short question for the IRC. I asked yesterday but i wasn't on long enough to get an answer... [18:43] Will U Touch support x86? [18:45] rsalveti, asac, i think the reserved ram is ducati [18:46] yeah, but for me neither adding vmalloc nor mem to the end of cmdline has any effect [18:46] i dont know how to change it [18:46] see http://paste.ubuntu.com/5881750/ === ricmm||lunch is now known as ricmm [18:47] i think the device has a) a bootloader hardcoded cdmline, b) the kernel cmdline and c) the cmdline we hand over in bootimg [18:47] ogra_: back, with a full belly [18:47] ptobably your mem arg doesnt override something that was set before [18:48] is there a trick ? [18:48] ogra_: what did you end up doing re upstart? [18:48] ricmm, well, dbus change is uploaded to read from ~/.cache/upstart/dbus-session ... the barchrc mangling stuff updated to ues that file instead ... [18:48] something that i coyuld do? i guess we would have to hack the kernel :)? [18:48] you mean write to .cache/upstart/ ? [18:49] ricmm, that bit i will pix by putting a snippet into /etc/profile.d/ instead ... so slangasek doesnt need to suggest it all the time (and we can finally say we use it) :) [18:49] ricmm, right, write [18:49] s/pix/fix/ [18:49] awesome [18:49] and desktop can use it from there too [18:50] while at it, how would you feel about a profiile.d hook to make all terminals join the phablet UPSTART_SESSION by default [18:50] for now, at least [18:50] mdeslaur had some securooty concerns though [18:50] *security [18:50] so peopne can ssh -> start application APP_ID="foo" [18:50] i guess we need to show all our evil hacks to the security team at some point :) [18:50] yes [18:51] ogra_: shall I start assigning the CVEs now? :) [18:51] haha [18:51] mdeslaur, after FF [18:51] :P [18:51] hehe [18:51] frankly I dont approve of any hacks just to ease app launching [18:51] ricmm, sure, go ahead with the UPSTART_SESSION stuff [18:51] not even the dbus one :) a default ssh login shouldnt be in the context of the running ui session [18:52] ricmm, well, this was more to unbreak autopilot [18:52] actually this is to perpetuate autopilot's broken behaviour [18:52] we are on a phone after all ,.... and starting X apps on a machine you ssh in with the same use works too [18:52] shouldnt autopilot set up its running environment? [18:54] ricmm, what's the alternative to launching from ssh? [18:54] it probably should [18:55] ricmm, my point is that X has exactly the same behavior [18:55] for development it's a bit hard to avoid [18:55] ogra_: yup I know [18:55] ricmm, through ~/.dbus/session/$sessionid [18:55] kalikiana: no option, just set up your correct environment [18:55] if you need to launch apps in the running session, join the session [18:55] so i dont think what we do is massively wrong [18:56] or we put in the hack, but if we have the hack then we dont take complaints ;) [18:56] its not wrong for now, in our single user scenario and heavy development [18:57] ricmm, would it be possible to output a message with the right command when ssh'ing in? [18:57] knowing exactly what to do is the biggest hurdle [18:57] I'll send an email today explaining the upstart session to the list [18:57] with ogra's changes you will be in dbus tho [18:57] so should be transparent to launch them as you always have [18:57] with the latest updates it "half" worked when I did what I used to do [18:58] wrong dpi and no uitk [18:58] if it was very obvious what to do without reading and asking, I wouldn't mind the exact invokation much [18:59] that's from my pov anyway [18:59] tomorrows image will work as usual [19:00] kalikiana: I heard from Kaleo that you are working on a search box widget. [19:00] kalikiana: Do you plan to implemented instant search as part of that so that it searches as you start typing? [19:00] kalikiana: what device do you have? [19:01] popey: is https://bugs.launchpad.net/ubuntu-clock-app/+bug/1172869 still valid? [19:01] Launchpad bug 1172869 in Ubuntu Clock App "[Clock app] Timer label updates before the timer hand reaches destination" [Medium,Triaged] [19:02] nik90: lemme see === racarr is now known as racarr|lunch [19:02] ricmm, galaxy nexus [19:03] pixel ratio shoudld be fine as long as you are phablet then [19:03] the default fits gnex [19:04] nik90: hmm, arguably it's not a problem... e.g. as the second hand goes from :31 to :30 the text switches from :31 to :30 as the hand leaves :31. which is true, because at that point it's 30.9999, then 30.9998 etc [19:05] I login as phablet user, yes [19:06] popey: tbh both text and hand start moving at the same time. However text updates instantly while the hand takes less than a second to reach its new destination. [19:06] popey: I do not think it is an issue. [19:06] I guess it could be that qtcreator's spawning ssh changed in some other way. I'll try updating the image again tomorrow and see if it works [19:06] popey: do you mind if I close it? === Aww_ is now known as EvilAww [19:07] nik90: sure [19:07] nik90: _however_ it _is_ an issue for clocks which go forward ☻ [19:07] e.g. stopwatch [19:08] nik90, search box widget is a bit ambiguous… in app search. your app defines "I am searchable", you can trigger it from the dash, the search view(s) replace(s) the header with a searchbox. the dash integration part is WIP === schwuk is now known as schwuk_away [19:09] kalikiana: oh. I thought you were making a widget which is like a textfield and search a listview for an item. [19:09] my mistake [19:09] popey: true, but the difference in updates between the text and the hand is very very small that I do not think it can be noticed [19:10] *difference in update frequency* [19:10] ☻ [19:10] I notice ☻ [19:11] but I am annoying like that, and happy to let it lie as there are other more important things to do [19:11] alrite I will keep it open for the stopwatch and clock. But I do not see a way to fix it. [19:11] if I start asking the text to wait until the hand has finished its transistion, I will accidentally add more bugs [19:12] yeah, it should be low priority [19:12] rsalveti: the glib commit for the nm-glib bug is this one right? https://git.gnome.org/browse/glib/commit/?id=31fde567a95ff8f50b6b0e75d4010da9b73514ed [19:13] * cyphermox_ is confused as to how this would affect the bug; it doesn't seem to be touching any related code [19:13] nik90, essentially yes. it goes further and adds tight shell integration, and have you not worry about all the details [19:13] so: you define your search resuls view [19:13] you don't care how it is invoked [19:13] nik90, in what context are you asking about search? [19:13] kalikiana: I need such a widget in the clock app to search world cities that I present in a listview to the user. [19:14] kalikiana: currently I have implemented it myself, but thought a universal search textfield to search through a list would be useful for some of the core apps [19:15] ogra_: https://code.launchpad.net/~ricmm/livecd-rootfs/join-upstart-session-default/+merge/175101 [19:16] what do you think about that? simple [19:16] doanac: so there will be a new image for today [19:16] with the fix so we can hopefully run stuff :) [19:16] we need to punch stuff through [19:20] ricmm, but from a new file in /etc/profile.d/ ;) [19:20] oops right [19:21] ogra_: any preferred number? [19:21] ricmm, ogra_ what does it take to enable the nexus 7 in tablet mode? [19:21] ricmm, as you like :) [19:21] pmcgowan, what does "in tablet mode" mean ? [19:21] ogra_, make it behave like a tablet [19:22] go landscape, havea sidestage [19:22] nik90, you'll want to use the search stuff then. I'm not sure exactly when it'll be ready but during this week or next dash integration will be implemented, maybe we can catch up again when there's something to try [19:22] kalikiana: sure. [19:23] pmcgowan, dunno, i have no clue how tablet mode gets enabled :) happz to learn about it :) [19:23] nik90, actually I'd be interested if you can point me at your current code, to see what requirements you have, what api you'd need etc [19:23] ogra_: pmcgowan we would need to translate input if we want it to default like that [19:23] must be in the boot somewhere [19:23] technically 0,0 is top left when looking at the n7 in portrait [19:23] so it would take some hacks, as the default hw orientation is actually portrait [19:23] rsalveti: sergiusens: I've created an MP to add the last of the Core Apps to the daily touch builds, if you can review them when you have time: https://code.launchpad.net/~mhall119/ubuntu-seeds/touch-remaining-coreapps/+merge/175105 [19:23] pmcgowan: i think we said that sidestage etc. would only be on larger devices... so maybe its a unity runtime check or something? [19:23] ogra_: refresh the MR [19:24] ricmm, is that the magic you never figured out? [19:24] pmcgowan: e.g. based on screensize etc.? [19:24] asac, I have gotten feedback from the head of design to perhaps make it do sidestage [19:24] also remember N7 is kind of in maintenance mode, we prefer to not look too deep there :) [19:24] yeah [19:24] pmcgowan: I stopped because design discarded the idea, you know that :( [19:24] ricmm, I never did ;) [19:25] ricmm, hmm, looks good, now i wonder if we shouldnt just ship that file in ubuntu-touch-session :D and thats actually one hack less then ;) [19:25] asac: ogra_: I think it'd probably be good to have a kernel patch to always use the latest argument, if possible [19:25] so we could overwrite the ones passed by the bootloader [19:26] ogra_: right, good idea [19:26] rsalveti, definitely ... i also wonder if our packages actually ship the android defined cmdlines [19:26] will re-MR [19:26] cyphermox_: yes, found with git bisect [19:26] rsalveti, i was suspecting before that we dont [19:26] since we have packages === Guest44134 is now known as Zic [19:27] (but havent looked at the kernel config yet) [19:27] ogra_: what do you mean by our packages? [19:27] kernel you mean? [19:27] I believe that's just from the bootloader [19:27] but yeah, need to check [19:27] rsalveti, mfisch: MR for emergency low-battery shutdown: request state [19:27] oops [19:27] I meant https://code.launchpad.net/~sforshee/powerd/low-battery-shutdown/+merge/174861 [19:28] ogra_: should I take https://code.launchpad.net/~mhall119/ubuntu-seeds/touch-remaining-coreapps/+merge/175105 or want to do it yourself? [19:28] sforshee: cool, thanks [19:28] rsalveti, feel free [19:28] ogra_: ok [19:28] rsalveti, i have to catherd another image build [19:28] for the dbus fix [19:29] ogra_: right, cool [19:29] sforshee: I can look in a bit [19:30] ogra_: https://code.launchpad.net/~ricmm/session-manager-touch/join-upstart-session-by-default/+merge/175106 [19:32] ricmm, drom the .sh :) [19:32] builds and works [19:32] *drop [19:33] ah, no, dont [19:33] seems thats common schema [19:33] how ugly :P [19:33] :) [19:33] yup [19:33] looks fine [19:34] rsalveti: can you take a look at ^ [19:36] do we need to change anything on the infra side now that the fix landed? or will autopilot stuff just work again? [19:36] ricmm: -su: initctl: command not found [19:36] when adb shell; su - phablet [19:36] asac: ? [19:36] initctl is not under phablet's path [19:37] sergiusens: autopilot tests failed completely in todays image [19:37] sergiusens: http://reports.qa.ubuntu.com/smokeng/saucy/image/3010/ [19:37] ricmm: might want to use /sbin/initctl instead [19:37] i was told fix is in archive or close to be there [19:37] and then ogra wanted to spi a new image so we might hit something with a useful dashboard [19:38] ogra_: changed back to needs review [19:38] rsalveti, oh, why ? [19:38] asac: oh, I'm not sure of what doanac is using in utah, if he's using phablet-test-run from phablet-tools, no, and it's me who needs to update [19:38] :( [19:38] thought you guys had talked about it [19:38] doanac: ^^ [19:38] ogra_: #adb shell; su - phablet [19:38] -su: initctl: command not found [19:38] asac: pinged a couple of times but didn't get a reply [19:38] rsalveti, hmm, so we need to fix adb [19:38] doanac: !! :-P [19:39] ogra_: well, it's under /sbin [19:39] ogra_: wouldn't hurt to use the entire path there [19:39] traditionally ubuntu has /sbin in the users path [19:39] sergiusens: lets see [19:39] we might be able to see it [19:39] in logs [19:39] http://reports.qa.ubuntu.com/smokeng/saucy/image/3010/ [19:39] going to gallery aopp [19:39] so yeah, we should fix it there, why is that not set properly though ? [19:39] going to build [19:40] it worked before [19:40] ogra_: right, indeed, we need to fix the path as well [19:40] it surely was a change in behaviour [19:40] if behavioyur is exactly like before [19:40] rsalveti, you adb in as root and sudo to phablet, no ? [19:40] it willw ork [19:40] ogra_: right, but with su - [19:40] well, shouldnt make a difference ... [19:40] /usr/local/bin:/usr/bin:/bin:/usr/local/games:/usr/games [19:41] ogra_: it does, the - makes it use a clean env [19:41] oh, its the /etc/environment mangling i guess , we dont use d-i to create it, yeah [19:41] sergiusens: Running command: utah-autopilot run -v gallery_app.tests.test_album_editor.TestAlbumEditor.test_add_photo [19:41] in https://jenkins.qa.ubuntu.com/job/saucy-touch-maguro-smoke-gallery-app-autopilot/7/consoleFull [19:41] rsalveti, oh, wow, not even /usr/sbin [19:41] ogra_: yeah [19:42] bzr branch lp:utah :) [19:42] now how that gets set [19:42] livecd-rrotfs ? [19:42] there's a user config hook [19:42] 01- [19:42] live-build/ubuntu-touch/hooks/48-setup-env.chroot [19:43] god [19:43] we need to clean that up [19:43] *whine* [19:43] maybe /etc/login.defs [19:43] yup [19:43] sergiusens: i think thats where the code lives [19:43] ogra_: the path there is fine [19:44] only matching path I found locally was /etc/login.defs [19:44] asac: http://bazaar.launchpad.net/~ubuntu-test-case-dev/ubuntu-test-cases/smoke-touch-apps/changes [19:44] whats that? [19:44] ah the definition [19:45] how it does the jobs [19:45] ogra_: yeah, gets from /etc/login.defs [19:45] asac: sergiusens: i don't use phablet-test-run. [19:45] doanac: yeah just saw [19:45] it required ssh and such we can't always use for our tests [19:45] rsalveti, strange [19:45] sergiusens: so will our fix make infrastructure just work? [19:45] e.g. a new image can succeed again? [19:45] :) [19:45] why do i get a fine path on a desktop then [19:45] sergiusens, asac: I run autopilot stuff with: [19:45] 'export $(grep SESSION /home/phablet/.dbus-session)', [19:45] 'sudo -i -u phablet {} \\'.format(env), [19:45] ' DBUS_SESSION_BUS_ADDRESS=$DBUS_SESSION_BUS_ADDRESS autopilot $*', [19:46] asac: not sure, doanac do you stop the shell? [19:46] doanac: if you do, that code needs to be updated [19:46] sergiusens: i launch the script listed above via an upstart job [19:46] ogra_: at your host: echo $PATH; sudo su - ; echo $PATH [19:46] why woudl i do sudo su - :P [19:46] sudo -u user -i [19:47] right, whatever :P [19:47] just to reproduce the issue [19:47] yeah, k [19:47] sergiusens: so what do i need to update specifically? [19:47] bit thats the sudo (or su) invocation ... [19:48] *but [19:48] ogra_: yeah, seems related with su [19:48] we cant fix user behavior indeed [19:48] su will not set the env correctly [19:48] well, you tell it to do that with the - [19:48] well, that just tells it to use a clean env [19:49] Provide an environment similar to what the user would expect had the user logged in directly. [19:49] that's a lie [19:49] :P [19:49] ah, yeah [19:50] so feel free to fix login :) [19:50] doanac: so you never stop the shell and the tests just work? [19:50] ogra_: more interesting: [19:50] root@ubuntu-phablet:/# sudo -u phablet -i [19:50] initctl: Unable to query session directory [19:50] yeah [19:50] see ricmm's fix above [19:50] that's with his fix [19:50] sergiusens: what do you mean by "stop the shell"? [19:51] doanac: for running apps, like the gallery [19:51] rsalveti, oh, hmm [19:51] i run "adb shell initctl emit " and the upstart job will launch a script that runs the commands i listed above [19:52] ogra_: * Removed unity8 from touch [armhf i386] [19:52] lol [19:52] after running ./update in the meta package [19:52] doanac: ok, just interesting that it works since the shell would be locked [19:52] root@ubuntu-phablet:/# sudo -u phablet -i [19:52] phablet@ubuntu-phablet:~$ env|grep XDG [19:52] phablet@ubuntu-phablet:~$ [19:52] doanac: can I see the full code instead of the snippet? === racarr|lunch is now known as racrr [19:52] rsalveti, so you miss the XDG stuff for the session dir [19:53] rsalveti, hmm ? thats weird, it should just use the ppa [19:53] right :-) [19:54] ogra_: got it, not using the next ppa there [19:54] ogra_: will add that there [19:54] it did before [19:54] sergiusens: we have a "utah-autopilot" script here: http://bazaar.launchpad.net/~utah/utah/dev/view/head:/examples/run_utah_phablet.py#L131 === Neo31` is now known as Neo31-desktop [19:54] are you sure? [19:54] we ship unity8 since a while [19:54] that's what i use to run autopilot stuff. [19:54] ogra_: sure, but that was because it was also part of the phablet-team/ppa [19:54] it wouldnt be shipped if it wasnt used by meta [19:54] not anymore [19:54] oh [19:54] ok [19:54] yeah, makes sense [19:55] sergiusens: we then have typical upstart code like: http://bazaar.launchpad.net/~utah/utah/dev/view/head:/examples/run_utah_phablet.py#L119 [19:55] * ogra_ wasnt aware the ppa was cleaned up :) [19:55] yay [19:56] rsalveti: I pushed the change [19:56] where is ricmm haha [19:56] oh, there you go [19:56] :-) [19:56] * ogra_ pokes https://launchpad.net/ubuntu/saucy/+source/dbus/1.6.12-0ubuntu2 === Neo31-desktop is now known as A7med [19:56] whay does it now promote ... hmpf [19:56] doanac: this is the shell restart part http://bazaar.launchpad.net/~utah/utah/dev/view/head:/examples/run_utah_phablet.py#L142 === racrr is now known as racarr [19:57] sergiusens: yes [19:57] doanac: unity8 is an upstart user session now [19:57] sergiusens: okay. I need to update this ASAP then [19:57] ogra_: you say '~/.dbus-session' in your changelog, but the code is using $HOME/.cache/upstart/dbus-session instead [19:57] ogra_: is that expected? [19:57] can you just do a "stop unity8" or something? [19:58] rsalveti, yeah, mdeslaur pointed that out too ... [19:58] doanac: so new imagese are probably coming out in 1-2h [19:58] :) [19:58] doanac: technically you can [19:58] i was to lazt to just upload for a changelog fix :) [19:58] doanac: in theory stop/start unity8 as the phablet user should work once that new package is built [19:58] doanac: if you have joined the upstart session, which with my MR wshould happen by default [19:58] rsalveti, $HOME/.cache/upstart/dbus-session is right [19:58] when su - phablet or ssh'ing as phablet [19:58] ricmm: okay thanks [19:58] -bash: /home/phablet/.dbus-session: No such file or directory [19:59] ogra_: right, then we need to change ^ as well, right? [19:59] doanac: is there an easy way for sergiusens to try utah-autopilot to really see if his updated system will suceed? [19:59] to source from the right file [19:59] rsalveti, yeah, already done, see livecd-rootfs [19:59] that is changed [19:59] yup [19:59] just saying that we would know for sure then [19:59] and just can sit back :) [19:59] and click buttons [19:59] ogra_: great [19:59] rsalveti, that should go into a profile.d snippet as well [19:59] just pushed [20:00] yeah [20:00] asac: install and run the "run_utah_phablet.py -N lp:~doanac/+junk/autopilot-utah-example/master.run" [20:00] that would be even better [20:00] but i'll keep that for tomorrow since we want to build a new image [20:00] hate to see stuff coming into .bashrc [20:00] ++ [20:00] well /etc/environment looks worse than our bachrc stuff [20:00] sergiusens: could you run the above line? [20:01] on your local/patched to latest system? ... of course just if you have it at hand [20:01] ogra_: details [20:01] :P [20:01] heh [20:01] sergiusens: you might add "--skip-install" so it doen't install a new image your system [20:01] doanac: asac ack [20:02] asac: I'll need to update, but not to worry [20:03] ogra_: how long for your dbus package to hit archive? [20:03] ogra_: just pushed new meta [20:04] mhall119: merged [20:04] ricmm, well, it sits there https://launchpad.net/ubuntu/saucy/+source/dbus/1.6.12-0ubuntu2 [20:04] in proposed ... [20:04] he usual rodeo [20:04] the* [20:04] ogra_, you screwed up the changelog entry, but thanks for actually fixing the location ;-) [20:04] well, livecd-rootfs was uploaded afterwards [20:05] seb128, yeah, you are not the first to point it out [20:05] * ogra_ will make a t-shirt for the next sprint "i screwed up the changelog entry" [20:05] :P [20:05] ricmm: happroved [20:05] \o/ [20:06] ogra_: lol [20:06] ogra_: order a few [20:06] M for me [20:06] "yes, we used a ppa" [20:06] :) [20:06] haha [20:06] we trust jenkins [20:06] we got a lot [20:07] doanac, so for the next image you might want to source the file from the new location (see the code, ignore the changelog) http://paste.ubuntu.com/5881676/ [20:07] ogra_: add something in bashrc that symlinks to that file [20:07] j/k j/k ;) [20:07] HAHA [20:07] are the launcher apps still hardcoded somewhere? [20:08] pmcgowan: yes, in unity8 [20:08] ogra_: ack. [20:08] pmcgowan: if you assign to me, I can update [20:08] sergiusens, ok, I think calendar got renamed or something [20:08] sergiusens: can you make it dynamic and secure as well? :P === thomi_ is now known as thomi [20:09] remembers the entire secure talk when demoing it [20:09] secure because hardcoded ? [20:10] rsalveti: sure, can we enable rw on /? trust me [20:10] sure :-) [20:10] :-P [20:16] boiko: still around? [20:23] sergiusens: yep [20:23] rsalveti, sergiusens, whats the prob, we even have it world writable now :) [20:24] ogra_: hahah [20:24] guess ssh complains all the time here [20:24] yeah [20:24] popey: I know this is late, but can you quickly check if you are able to see the correct sunrise/sunset times for your location? [20:25] and by correct you indeed mean that he needs to stay up to verify ? [20:25] ogra_: :-) [20:26] ogra_: he can just check with google [20:26] :) [20:28] * ogra_ sighs ... dbus seems to have quite some deps to check ... of britney went on vacation without telling us ... [20:28] ☻ [20:28] s/of/or/ [20:28] nik90: ok [20:30] yes, its accurate nik90 [20:30] 05:01 / 21:10 which agrees with me googling for sunset/sunrise times [20:31] popey: I just fixed it and released the bug, but I got a comment saying it was wrong. Thanks for checking [20:31] I realise it is getting late [20:34] nah, its fine [20:39] and finally, upgrade using gpg: http://paste.ubuntu.com/5882103/ (not very visible as the files are valid ;)) [20:39] ogra_: I was being sarcastic ;-) [20:39] sergiusens, no, really ?!? [20:39] :-) [20:39] :) === Gh0sT is now known as mahabal [20:40] sergiusens, rsalveti: can one of you commit http://paste.ubuntu.com/5882107/ to the recovery git branch and kick a rebuild of android? [20:41] (yeah, that contains a tar.xz as base64, hopefully git am won't mind) [20:41] stgraber: sure [20:42] rsalveti: thanks! [20:42] sergiusens: can take care of the git push, got my hands at the phablet-saucy stuff [20:42] stgraber: yeah, give me a couple [20:42] sergiusens: I can push it, will have to push in both branches [20:42] sergiusens: ping you back to trigger a new build [20:43] in theory we'll be feature-complete after that one. There may be some way of improving performance (in the worst case the .tar.xz is currently read 4 times...) but for the normal update (first flash and delta updates from there), it's not too bad (read twice and smaller files) [20:43] xnox, binary blobs for qemu ? [20:44] even in the worst case, a full reflash took a bit over 5min on my nexus4, though getting measurements on maguro would be nice (should be much easier once I finish my phablet-flash changes) [20:45] ogra_: i believe there are google vendor files and blobs for qemu yeah.... unless they are simply prebuilts and source is available somewhere. [20:45] some scary greps all over, hope it's all behaving properly :-) [20:45] xnox, i would expect the latter [20:45] xnox, they wont be in our tree indeed :) [20:46] and i doubt in CM trees either [20:46] ogra_: rsalveti: at the moment I'm looking at consolidating & hopefully removing "prebuilts/ndk/" from the tree. Some thinks compile to API level 9, some to 8, our libc is 10 and AOSP is at 14 [20:46] /o\ [20:46] (there once was a cyanogenmpd 2.x emulator or so .... but i havent seen something newer) [20:46] ogra_: yeah, it seems dead, that one. [20:47] xnox: what would the the benefit of that? [20:47] just saving disk space? [20:47] smaller sourcer package [20:47] -r [20:47] I wonder if that will cause issues with our porters [20:48] might need to check a port coming from ICS [20:48] I'm kind of against just removing most of stuff for now, prefer doing the clean up later [20:49] rsalveti: well, it's mimimum versions of the ndk, so things should be forward portable. our blobs do work with libc at 10, so that should be an ok baseline. but our tree at the moment compiles stl/libc yet other packages use includes from and link against prebuilts/ndk/* [20:50] xnox: right, that's why I was interested to see if there would be any other dependency with ports coming from ICS (using ics blobs but in a jb - 4.2.2 image) [20:51] prebuilds/ndk is 472MB of legacy stuff (many levels) and I can't be rebuilding them all, nor uploading blobs into the ubuntu-archive. [20:52] rsalveti: right this is not removed from the repo forest, but not used/excluded in the archive build. [20:52] ogra_: can we turn off syncing of jenkins ubuntu-touch-preview? [20:52] i feel that: brunch $foo should just-work(tm) and generate ready-to use images. [20:52] xnox: right [20:53] sounds fine then [20:53] sergiusens, do ports work with flipped ? [20:54] ogra_: no, but the livebuild is broken for unflipped [20:54] sergiusens, oh [20:54] my package correctly creates zips for both flipped and unflipped. [20:54] yeah, i'll remove the job from nusakans crontab [20:54] and system/boot images. [20:55] xnox, your package wont be used on modotola droid :) [20:55] sergiusens, hey! will you be working for messaging-app and dialer-app CI jobs ? [20:55] om26er: yes [20:55] sergiusens, I could if you want [20:55] om26er: don't worry, focus on getting good autopilot tests ;-) [20:56] ogra_: ah, gotcha..... i obviously do not have the unflipped ubuntu tarball...... [20:56] yeah, and it should be long dead, but we havent verified a port ourselves yet [20:56] i'll do that this week [20:56] sergiusens, sure, it kind of falls into my job so wanted to make sure if you were comfortable. [20:57] there is just always another emergency holding me back [20:57] sergiusens: one thing, we never tagged a saucy build for flipped, right? [20:57] sorry, unflipped [20:58] xnox: where are you pushing your stuff? [20:58] debian/ dir is at: git@github.com:xnox/android.git [20:59] and manifest.xml is just locally editted to remove stuff. [20:59] and uploaded into the ppa [20:59] <om26er> sergiusens, I could if you want [20:59] arh. [20:59] ignore. [20:59] https://launchpad.net/~ubuntu-toolchain/+archive/android [20:59] * xnox needs new batteries for my mouse. [21:00] *his [21:02] om26er: no sweat [21:04] sergiusens: hey, did you by chance had time to look into the CI and autolanding for those new branches? [21:05] boiko: where are your MRs, need to get to it now [21:06] sergiusens: https://code.launchpad.net/~boiko/telephony-service/packaging/+merge/175140 [21:07] sergiusens: for the other two branches I think salem_ didn't do the MRs yet [21:07] stgraber: pushed [21:08] sergiusens: can you trigger a new android build? [21:08] boiko, for future ref, I am your QE ;) [21:08] rsalveti: err, I was reviewing the diff! [21:08] sergiusens: I said I'd take care of it [21:08] rsalveti: I'll create [21:09] rsalveti: must of missed it since I thought I read phablet-saucy [21:09] om26er: oh, sorry, didn't know you were also in charge of setting that up, good to know [21:09] sergiusens: right, just said that I'd push it because I had my hands in the phablet-saucy branch already :-) [21:09] sergiusens: sorry for that, I should have asked om26er instead [21:09] so I just pushed in both branches [21:10] boiko: don't worry, we have different plans for these projects === dandrader|afk is now known as dandrader [21:17] boiko: so the telephony-service would still depend on indicator-messages in the phablet-team ppa [21:18] sergiusens: yes, that didn't change :/ [21:19] boiko: good thing i checked :-P [21:19] asac, looks like dbus might still take a while, the dependency checks are still running (http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html) [21:20] sergiusens: yeah, the only thing is that the UI is not there anymore, but the rest is still the same [21:22] sergiusens, https://code.launchpad.net/~tiagosh/dialer-app/dialer-app-noautopilot/+merge/175141 [21:22] rsalveti: thanks! [21:22] sergiusens, https://code.launchpad.net/~tiagosh/messaging-app/messaging-app-noautopilot/+merge/175142 [21:24] sergiusens: let me know once you kick the build, I'll monitor and test once it's done [21:26] stgraber: kciked [21:26] *kicked [21:26] thanks [21:26] salem_: thanks === MrDHat is now known as MrDHat|afk === salem_ is now known as _salem [21:56] i have a question on the default user's folder names in ubuntu touch. on normal ubuntu, certain folder names, like "Desktop" or "Music" are automatically when a user is created. and the thing that happens is that the folder names are translated to the user's (or system's) language. [21:57] should an ubuntu touch app expect user folders to be translated this way? if so, how would it find the correct folder for music, for example? [21:59] dkessel: apps should expect translated folders, yes, I think for things like ~/Music/ there is an XDG method of retrieving the actual name [22:04] dkessel, you want XDG_MUSIC_DIR (look at the file ~/.config/user-dirs.dirs)... i looked into this for the music-app but couldn't figure out how to query XDG from javascript [22:05] ahayzen, yeah. I found bug 1198909, which is about this issue. [22:05] bug 1198909 in Ubuntu Music App "Fails to scan Music folder for non-English locales" [Low,Confirmed] https://launchpad.net/bugs/1198909 [22:09] hello [22:09] dkessel, yh i was looking at tht yesterday... [22:13] ogra_: :/ [22:13] ok [22:24] ahayzen, mhall119 , thanks for your help so far. i am in contact with danielholmen to fix my problems. good bye, see you the other day [22:24] dkessel, thanks cyas :) [22:51] hello everybody [22:52] asac, dbus is there, build started [22:53] * ogra_ vanishes [22:53] asac: that's normal. if it's not considered for >>1h than there are things to fix. otherwise it's typically <<30minutes after fully build across all arches to be copied over into release pocket. [22:53] hmm... it was 4h. [22:53] https://launchpad.net/ubuntu/+source/dbus/+publishinghistory [22:58] how can i debug a broken recovery? I installed it on my device but it's giving me a error installing is there anything like a serial connection on desktops for the phone? === Gh0sT is now known as mahabal === _salem is now known as salem_ [23:12] sweet. latest build is awesome [23:13] can someone remind me/correct me, is there a way to 'exit' an application that your in besides from the home screen [23:20] asac, build failed, something seems wrong with the buntu-touch-coreapps-drivers PPA i'll check tomorrow [23:22] rsalveti, sergiusens , ^^^ in case one of you wants to look (no idea whats wrong there ... w3m cadejo.buildd/~buildd/LiveCD/saucy/ubuntu-touch/ from nusakan has the live builder logs) [23:23] will check [23:23] (nothing changed in the code that could have impact on that ... ) [23:24] Fetched 11.4 MB in 31s (364 kB/s) [23:24] W: Failed to fetch bzip2:/var/lib/apt/lists/partial/ppa.launchpad.net_ubuntu-touch-coreapps-drivers_daily_ubu [23:24] E: Some index files failed to download. They have been ignored, or old ones used instead. [23:24] thats the error [23:25] right, probably a sync error when fetching the data from launchpad [23:25] ogra_: we should trigger it again, want me to do it? so I can watch once it's done [23:25] well, i'll trigger a new build but wint be around in care it fails [23:26] running [23:26] ogra_: rsalveti seems to be a common ppa issue, jenkins sometimes fails due to this [23:26] cool [23:26] yeah [23:26] it is the first time i see it [23:27] saw that a few times already [23:27] even when running apt-get update [23:27] k [23:27] we'll see (well, you, i'll be off :) ) === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader