[00:04] ok ... i am running now camera autopilot on 08 image === moocow is now known as moocow2 [00:08] \o/ [00:08] popey: http://paste.ubuntu.com/5860109/ [00:08] OK!!!! [00:09] camera autopilot works on 08 [00:09] no error [00:28] Saviq: do you need someone to trigger the build job for unity8? or should we wait tomorrow's build? [00:28] asac`: nice === Noskcaj10 is now known as Noskcaj [00:33] rsalveti, it should be daily releasing now [00:33] rsalveti, so nothing we can speed up, really [00:34] rsalveti, should be in tomorrow's image IIUC [00:38] Saviq: cool, that's fine === SkavenXXI is now known as SkavenXXI-[OFF] === _salem is now known as salem_ [00:57] Hello everyone! Is Ubuntu Touch in a more usable state now? Has it evolved a bit since the initial release of the preview? [00:58] CrusaderAD: Ubuntu touch is more usable than tomorrow and yesterday it will be more usable than today :) [01:00] Oranger, you are a true poet [01:00] *blink* [01:02] Ahah ^^ pmcgowan : I feel something like irony in your words ;) [01:02] veebers: thomi: who of you had a device? [01:02] asac`: currently veebers has both devices [01:02] not at all [01:02] asac`: but I can get one back as soon as he un-bricks it :) [01:03] It says in the wiki that nexus 7 is broken, is that still the case? [01:03] thomi: veebers: https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdDhEUy1nM1Jab3N4VGNQS0dQR19lTFE [01:03] started testing the first tab === chriadam|away is now known as chriadam [01:03] (canonical apps_ [01:04] CrusaderAD, nexus 7 is working with some issues related to camera and video afaik [01:04] thomi: i just run through the list from the pad [01:04] veebers: ^^ [01:04] Ok, it's 3 am here... Good night everyone ! [01:05] asac`: awesome. I think veebers may be at lunch at the moment [01:05] ok thanks! [01:05] kk [01:05] asac`: but I'll talk to him when he gets back [01:05] or, I guess he'll read this [01:05] thomi: whats his full name? :) [01:05] (so i can shared) [01:05] Chris Lee [01:08] hello? [01:08] hi [01:09] can you helpme out? [01:09] what's wrong? [01:10] i did a build for ubuntu touch but in the out directory i get a cm10.1 zip that says unofficial and some .img files [01:10] and im not sure what to do with them [01:11] which device? [01:11] mhall119: awake? [01:12] the device is a p769 its another variant of the p760 [01:12] Optimus L9\ [01:14] so no one can help me? [01:14] veebers: thomi: gallery-app autopilot is odd ... undeterministic. fails sometimes 23/23 [01:14] sometimes 5/6 [01:14] etc. [01:14] seems not all tests are run [01:14] is that a known prob? [01:14] asac`: not that I know of - maybe pastebin the output? [01:14] (e.g. that sometimes not all apps are run [01:14] ) [01:15] asac`: it should certainly run all the tests every run [01:15] wait [01:15] yeah that seems very broken :) [01:15] that was me being screwed :) [01:15] :) [01:15] did run a different test by hitting arrow key too many times [01:15] ahhh [01:15] dam [01:15] I do that all the time [01:15] True_unReal, are you following the porting guide? [01:16] yea i did but its not very helpful on what to do with the files after you build [01:16] them [01:17] True_unReal, the latest images are also not compatible with the guide and we are updating that [01:17] what do you mean? [01:17] not sure I can help you, most of the devs are offline [01:18] True_unReal, we made a big change to the container model [01:18] so now we boot to ubuntu and contain android rather than vice versa [01:19] ok so the build i made was useless then? [01:20] perhaps not but I am not sure, for sure the instructions will change [01:20] will be revised the next couple of days [01:20] Hi asac`, thomi: I got the phone 'unbricked' [01:20] nice :) [01:21] True_unReal: not useless, can still work with our tagged images [01:21] veebers: can you make one device remotely accessible to thomi? so we can split the load of filing the spreadsheet? [01:21] veebers: or, would you like to work around here this afternoon? [01:21] and with our current unflipped ones, will just not work yet with the flipped images (ubuntu booting first) [01:21] (just an idea) [01:21] but the porting is still useful [01:21] asac`: I can do one better and hand it off in person :-) [01:21] True_unReal: so if you build them, you just need to flash it :-) [01:21] asac`: veebers and I live about 5 minutes away from each other [01:21] veebers: oh NZ is so small :) ... or are you living the "geek district"? :) [01:22] lol [01:22] use the cm10...zip file first (via recovery) [01:22] cool [01:22] hah. I wish we had a geek district [01:22] and then the ubuntu.zip [01:22] mmm ok but can you answer me this question i flashed the cm10.1 zip file it build with some vsauce arm file zip and it boots but the screen is just black? [01:22] thomi: sure, let me finish this coffee and I'll pop around [01:22] veebers: by the time you get here I should have closed this bug I'm working on [01:22] ok ... i will continue testing... note that i didnt include the logs, nor filed bugs [01:23] but what ubuntu.zip? [01:23] want that owners really get used to do that on their own [01:23] True_unReal: the screen is just black because you're probably missing the ubuntu zip file now (the ubuntu image) [01:23] let me find you the link [01:23] alright [01:24] True_unReal: http://cdimage.ubuntu.com/ubuntu-touch-preview/raring/monthly-06/raring-preinstalled-phablet-armhf.zip [01:24] ok thank you [01:25] this is a tagged build, so should be stable [01:25] once you get that running you can try the saucy based image [01:25] asac`: for now [01:25] yay ... second app that has zero test failures is "share-app" :) [01:25] mhall119: how long? [01:25] a while still, probably [01:25] wanna help running through a few autopilot tests to seed info [01:25] alright [01:25] it's only 9:30pm [01:25] 6:25 [01:26] pm [01:27] is there a major difference between saucy and raring [01:28] True_unReal: a few things, but it's better to start with the raring based one [01:28] as it's known to be more stable [01:28] alright [01:30] is there anything else i need to flash after that? [01:30] nops [01:31] * rsalveti takes a break [01:36] rsalveti, thanks [02:04] asac`: colourful :) [02:05] scary :) [02:05] i will change the numbers to reflect success/total [02:05] currently its fail/total [02:05] but i think folks believe the high numbers are good [02:13] thomi: veebers: mhall119: i leave the "core apps tab" alone [02:13] continue on SDK ... feel free to split filling that out [02:14] i probably will drop to bed before getting to unity [02:14] err shell [02:14] asac`: ok, just getting set up [02:14] nice [02:14] i started to fill some of the application names and components in the "core apps" tab [02:14] but... there are more in the list in the pad ... so dont forget :) [02:15] also note that sometimes the component name doesnt follow the same naming scheme as all the others === chihchun_afk is now known as chihchun [02:16] e.g. phone-app-connected-autopilot turned out to be connected_tests rather than phone_app_connected [02:16] mhall119: thomi: veebers: ^^ [02:24] I'm running xubuntu saucy and whenever i open the clock app it says that the alarm part has not been made yet. is this meant to happen? [02:44] asac`: what build are you running? [02:44] asac`: This device seems to be running saucy-34 [02:49] thomi: 08 [02:49] that should still be /current [02:49] so just phablet-flash might do the right thing (TM) [02:49] hmmm... how come I have 34? [02:49] i refer to a date [02:50] 20130708 == 08 for me [02:50] i dont know about other image version scheme :) [02:50] asac`: ahh [02:50] http://cdimage.ubuntu.com/ubuntu-touch/daily-preinstalled/ [02:50] maybe you are still on ubuntu-touch-preview? [02:50] not sure what saucy-34 means :) [02:50] asac`: what happens when you run: cat /system/ubuntu_stamp | grep JENKINS_BUILD [02:50] on the device? [02:50] but if you know what i means i am happy :) [02:51] yeah... that i dont know :) [02:51] oh wait [02:51] -37 [02:51] but i think we dont use jenkins anyway [02:51] so not sure if that number is meaningful still... [02:51] hmmm [02:51] ok [02:51] i think jenkins is only used for part of the image right now [02:51] (android parts) [02:55] thomi: veebers: mhall119: oki ... crashing for today... core apps and shell/unity tab would be nice to get more stuff in... [02:55] ignore indicators/mir tab for now i guess [02:55] they dont have matches in the pad [02:56] hey guys so you have to download the ubuntu images? [02:56] will check in 15 minutes before completely dropping [02:56] True_unReal: i just run phablet-flash ... that does the right thing with my device to upgrade to latest daily image [02:56] asac`: will try, but having some device issues still [02:56] n7 or n4? [02:56] mmm ok thanks [02:57] fwiw, i run that on my laptop with phone connected to it [02:57] both :-/ [02:57] thomi: hmm. ok... not working after phablet-flash? [02:57] can't get the n7 to boot at all, suspect a battery issue, n4 needs flashing, but I'm having Internet issues with the flash process [02:58] asac`: I'll struggle on, and we can touch base tomorrow perhaps === salem_ is now known as _salem [03:02] hey guys any idea how to fix bootloop [03:15] thomi: yeah. try n4 ... if that also fails do something else and rather check with team her etomorrow what to do [03:15] mhall119 might do a few of the core apps [03:16] and with that we already have good data for this night effort : [03:16] ) [03:16] bye === asac` is now known as asac === chihchun is now known as chihchun_afk === chihchun_afk is now known as chihchun === Namidairo`bnc is now known as Namidairo [03:31] got it to boot hoorah!! [03:31] where do I file bugs against friends-app? it doesn't have a bugtracker configured in launchpad [03:31] do i file against manhattan, or....? [03:32] mhall119: if you're still around, I guess you'd know? === Namidairo is now known as Namidairo`bnc [06:31] hello? [06:32] hello? [06:34] hi [06:41] hello? [06:45] any one here? [07:21] ogra_, hi, I just phablet-flashed the latest image on the N$ and there's no display. Is this the current status with no Mir landed yet or did I do something wrong? [07:21] I mean N4 [07:24] hey didrocks [07:24] hey rickspencer3! [07:24] so, bug list for dog-food busting bugs? [07:25] seems like we need to be able to identify the bugs, and have a certain person decide if they are breaking dog-fooding [07:25] sounds like the most effective and immediatly actionnable plan to me (doesn't prevent adding more tests, but it's more a long term thing and take time to grow the list) [07:25] didrocks, indeed, having more tests is not useful if we aren't fixing the bugs we are finding already ;) [07:25] exactly :) [07:26] rickspencer3: so agreed, then, we switch the image if we are happy with current image state [07:26] didrocks, I thought that was the plan all along [07:26] like the image would be pending or something [07:26] and then when it proved itself, it would be pushed out [07:26] I don't know who/how the image switch is done [07:27] the most difficult part I guess is to have that list public and readable [07:27] hence the small tool proposal I have done in the past (but it's not the only solution, maybe others would have better ones) [07:27] good morning [07:27] didrocks, what does the tool do, exactly? [07:28] rickspencer3: producing the web pages I posted on the last email, like http://people.canonical.com/~platform/olddesign/upstream.html [07:28] then, upstream can see, on their components, which bugs needs to be addressed design-wise and release-wise [07:28] ok [07:29] didrocks, so how do we identify "stop the line" bugs on that list? [07:29] and design/release can "ack" when they confirm that the fix is working (and it cleans + archive) [07:29] didrocks, I think a lot of these bugs are not related to design at all [07:29] like the battery eating bug [07:29] rickspencer3: see "distro priorities" [07:29] (second part) [07:30] so basically, we have 2 master project [07:30] "design priorities" [07:30] and "distro priorities" [07:30] we attach that to a bug which seems important [07:30] then, some people can switch to a particular status to "ack, this is something important that needs to be addressed" [07:30] set a priority [07:30] didrocks, ok, so the tooling seems good [07:30] but who are the "people" that do the attaching [07:31] once the bugs is fixed for upstream and in distro [07:31] and how do they find out which bugs are blocking dog fooding? [07:31] (for the Ubuntu Touch case) [07:31] the "people" will ack it's fixed and switch to fix released, the bugs will be archive :) [07:31] rickspencer3: it was John for design and me for Unity until now [07:31] I'm happy to help, I think the foundation team should give a hand as well [07:32] didrocks, I think it's too much to ask you to do all of Desktop and Touch [07:32] rickspencer3: well, days are quite long already TBH, so yeah, can't just rely on me for sure :-) [07:32] didrocks, so, basically, you guys dredge through bug reports on lp each day and find bugs that you think need to be addressed? [07:32] rickspencer3: yeah, I have good filters in my inbox to quickly look over bugs [07:33] so, it sounds like we a need a person or people to do the same thing for Ubuntu Touch [07:33] right [07:33] do you think anyone considers themselves to be that person already? [07:33] didrocks: rickspencer3 the image switch is planned to happen once ogra_ and I have flashed multiple devices and done a shakedown test [07:33] * popey is catching up [07:33] popey, sorry, what "image switch" are you referring to? [07:34] pointing /current at the image du jour [07:34] rickspencer3: speaking of the devil, maybe popey? he knows how to find bugs and as he's talking to the community a lot, he can see which bugs are more a priority [07:34] rickspencer3: TBH, the French forum helps me a lot to see what bugs needs to be addressed [07:34] popey, so you are saying you will look at test results, confirm that the image has been manually validated, and then throw the switch? every day? [07:35] that was the decision we came up with yesterday after the awful image [07:35] popey, ok good [07:35] i think rsalveti mentioned this in the email thread [07:35] (just looking at the messages and seeing the number of complains on something is a good stats to see if something went bad :p) [07:35] as the image gets finished soon after ogra_ and I start our day [07:35] i look at all the new bugs each morning as part of my daily routine also [07:35] didrocks, so it sounds like popey and ogra_ are the "people" who identify the bugs [07:35] but I don't necessarily (currently) assign priority to them, but do an initial triage/confirm [07:36] yep, sounds logical to me :) [07:36] popey, so how do we make sure that people who "broke the image" fix the issues asap? [07:36] well initially we need some smarts to figure out what broke [07:36] in other words, if you "made the image awful" it needs to be your #1 priority to unaweful it [07:37] I agree. [07:37] "unaweful" -> /me marks on his dictionary :) [07:37] popey, traditionally on the desktop that happened by going to the logical maintainer for the area that seems busted [07:37] and then that person investigates [07:37] I agree with the thread where it's noted we have a wide spread of people across a number of components. Pat/Olli & Bill oversee AIUI [07:38] popey, ogra_, didrocks so how do we make unawfuling the #1 priority? [07:38] But we're likely to find these issues 4-5 hours before any of them start [07:38] so need to be able to reprioritse engineers when we get an awful build [07:39] popey, first of all, I just saw via tab completion that someone has the irc nick "poutine", which is the most awesome nick ever [07:39] Without individual engineers having numerous managers pulling in various directions [07:39] I keep seeing pictures of poutine online, never tried it, looks delicious [07:39] rickspencer3: harassing people was the only way until now, polling doesn't work… but I think the engineering manager should be the point of contact [07:39] which is an issue for timely fixes [07:39] didrocks, well, for the next 2 weeks, I am in Berlin [07:39] so I can harass people the moment they awake ;) [07:40] then I will trade places with asac [07:40] rickspencer3: engineering manager -> I meant, direct manager :-) [07:40] like x broke y, x's managers is responsible to get it fixed timely :) [07:40] didrocks, popey experience with 12.04 suggests that having someone beating the daily quality drum early in "the day" is helpful [07:40] +1 [07:41] well, that's what I did… but warning, it creates tensions :) [07:41] I was based in France for a lot of 12.04, so I could spend my mornings doing that [07:41] (especially when you enforcing tests for everything and having aggressive revert usage) [07:42] didrocks, well, maybe someone like asac would be better for that? [07:42] sounds good to me :) [07:42] ;) [07:42] I'll be based on the East Coast starting in August, so my schedule won't be terrible for it, like when I was West Coast [07:45] i would like a bot in here which announces when the build is done [07:45] we used to have that in an internal channel [07:45] oSoMoN: morning! [07:48] popey: btw, I think we want to discuss image build time [07:48] duration or start/end time? [07:48] popey: it should be just after we've dealt with daily releases, instead of the day after [07:48] start time* [07:48] what, so late in the eu evening? [07:49] popey: I've no idea when you build the image TBH :) [07:49] ogra_ sent a mail to the list, it happens ~9 am now [07:49] didrocks, "due to popular demand the daily image builds (flipped image) now run 5h [07:49] earlier (starting at 8:32 UTC)" [07:49] didrocks, ogra sent that yesterday [07:50] I'd rather it was at least ~2 hours earlier so my first activity of the day would be flashing [07:50] I think we should discuss that time [07:50] seb128: popey: ok, I think we didn't manually publish what we can publish at that time, knowing that sil2100 is starting his day now [07:50] I think 8:31 would be better [07:50] [07:50] so if there are packaging changes, we won't get latest [07:51] right, that's good for things that autoland [07:51] it's too early to get things that europeans need to land [07:51] I don't know how long an image build takes [07:52] the issue by pushing it later is that we don't have the images early in the day [07:52] seb128: yeah, but I think things that don't daily release (please stop use the term autolanding, everyone is confused by it :p) are less impacting/have big changes [07:52] like popey says, it's already late for him [07:52] popey: yeah, but you need latest content and if we are in manual publishing mode… [07:52] didrocks, we should maybe start thinking about bi-daily release? [07:52] you will get an extra day of daily [07:57] one in the u.s end of day [07:57] and one in the european morning? [07:57] so u.s end of day ones would have most of the day changes and be on the image [07:57] seb128: that doesn't change for manual publishing if nobody is here to manual publishing changes [07:57] well, the u.s guys would be [07:57] they aren't right now, so first, ensuring they can [07:57] and bi-daily release means that upstream has less time to land conflicting changes [07:57] (see my email about backward compatibility) [07:57] telling "everything should land in a coherent piece for 00 UTC" is already hard to get respected [07:57] so twice a day, not sure how it will plan [07:57] but in any case, it's not a power issues, daily release can still bi-daily, it's more a process/ensuring that everyone's on board issue [07:57] s/still/be/ [07:57] surely if it's building at ~5am for a 7am UTC test start nobody is up at 5am (sorry Aus/NZ) doing anything just prior to that are they? [07:57] or are we waiting on stuff building from 10pm utc the day before? [07:57] popey: the issue is not building, it's when they are packaging changes or a stack failed [07:57] the deal with daily release and other distro maintainers is to block publication if we have packaging changes [07:57] right, but nobody would look at that till you guys wake anyway [07:57] and someone with upload rights review the packaging diff [07:57] and "ack" by publishing it to distro [07:57] popey: exactly, and that's the issue [07:58] or a stack failing: for instance, the apps stack isn't published right now because we are getting 28 autopilot tests failing [07:58] (I guess that's why sil2100 is pinging oSoMoN ;)) [07:58] ☻ [08:00] Yes ;) Seems like some change in the UI toolkit I guess [08:00] Since there's an invalid assert going on with transperancy now [08:00] we really need the sdk to run apps AP tests [08:01] morning ... [08:02] Ubuntu touch на TurboPad 720 | http://askubuntu.com/q/318404 [08:02] Good morning ogra_ [08:02] hey ogra_! [08:02] rickspencer3, "unawfulling" would work if *any* tests would happen :) [08:04] theoretically freshly built images are supposed to come out at cdimage.u.c/$product/$type/pending .... where some automated tool is supposed to pick them up, run tests and if these are successful /pending should be moved to /current === schwuk_away is now known as schwuk [08:06] it is kind of a coincidence that we had a broken shell at the same time i moved the images earlier .... but it showed flaws in the process [08:07] so until the QA side actually happens popey and i will do manual smoketests (does it boot to shell, are all things i should see on the screen, can i start an app) .... with the hope that this automated testing starts to happen soon [08:09] i wouldnt mind to have bi-dailies (or even one every two hours or so) *if* the automation is in place and *if* we can get a dedicated livefs builder [08:10] at the point where we drop apt (soon), fixes will not get in through package updates anymore but your fix will have to do a whole turnaround through the whole process of landing in an image, that means fixes will take significantly longer to reach the user, doing more image builds definitely helps here [08:13] ogra_: the issues are that there isn't this backward compatibility/don't break ABI/API culture, so we need to be really careful in the way we builds and lands things [08:14] ogra_: which is what dailies are doing, ensuring we don't start building if everything that it build-deps on didn't finish to build [08:14] popey, i'd liked it 2h earlier too, but there are other builds running at that time, and we would just end up sitting in a queue [08:14] and not landing automatically if anything more on the "front" of the queue didn't land [08:14] (so every 2 hours is not feasable with those constraints) [08:15] didrocks, it is feasible if the CI jobs run in such a manner too [08:16] ogra_: CI doesn't pick latest of other changes and that won't fly with "we need to land those breaking changes together" [08:16] i would even go further and say trigger an image build every time an upload lands that closes a bug with critical status on LP [08:16] sil2100: hey, sorry I was out for an errand, what’s up? [08:17] didrocks, hold them back until they are unbroken then [08:17] can CI run in a constant loop and if all tests pass release ? [08:17] *do a release [08:17] ogra_: I don't think it can do that, and CI doesn't build with ppas, distro builders, only daily releases does it [08:18] ogra_: but yeah, I would be happy if we have that, but it means that upstream needs to stop breaking API and not being backward compatible [08:18] oSoMoN: hi! hm, I already poked the UI toolkit guys and it seems to be a problem in the toolkit [08:18] which isn't the target AFAIK for now at least [08:18] oSoMoN: since we're getting a lot of touch failures on apps stack [08:18] so with the no-packages model that means that i might have to wait two days for a fix ? [08:18] even a critical one ... [08:19] oSoMoN: a lot in webbrowser - but it's due to a rounding error in the UI toolkit, so I guess we need the guys to fix that... [08:19] sil2100: can you point me to those failures, so I can have a look, just for my info? [08:22] ogra_: that's orthogonal with package/no-package, isn't it? [08:22] no [08:23] oSoMoN: righto [08:23] once the android packages are in the archive (this month) you would be able to just dist-upgrade immediately once the fix lands [08:23] oSoMoN: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/395/testReport/ [08:23] oSoMoN: MismatchError: After 10.0 seconds test on Suggestions.opacity failed: 1 != dbus.Double(0.9999999999641627, variant_level=1) [08:23] Is the issue everywhere [08:24] ogra_: but the android packages are not daily releasing, so not linked to the current discussion? [08:24] with the no package model it will take the CI time and then i still have to wait for a new image build [08:24] it adds up [08:24] the android packages will be re-build per change [08:24] ogra_: defines CI, a lot of people means daily release or upstream merger or "autolanding" [08:25] hey guys can i ask a question? [08:25] sil2100: huh, that’s weird indeed, is it confirmed that it’s an issue in the toolkit? [08:25] if either something in the android git or in platform-api/hybris happens they will be rebuilt [08:25] oSoMoN: so it seems! Michał just now put up a bug: https://launchpad.net/bugs/1199662 [08:25] Ubuntu bug 1199662 in Ubuntu UI Toolkit "New easing causes animations not to reach their target values" [Undecided,New] [08:26] oSoMoN: since there have been some modifications made yesterday [08:26] but you agree that for that, we need to have all components not breaking API or ABI and having some kind of backward compatibility, right? [08:26] ok [08:26] didrocks, well, i refer to "run once a day to build PPA packages that i dump in the archive once i know the deps and tests pass" [08:26] guys how do you fix a bug? [08:27] didrocks, imho the "once a day" needs to be switched to "at every change (or at a bunch of these)" somehow [08:27] 10:26:09 didrocks | but you agree that for that, we need to have all components not breaking API or ABI and having some kind [08:27] | of backward compatibility, right? [08:27] ogra_: ^ [08:27] True_unReal: depends on what bug you have in mind, usually you find the problem causing the bug in the code and fix it there [08:27] wifi and network radio [08:28]