[00:17] ok, so I'm latested mir + keyboard [00:17] still getting unity8 + maliit + hud crashes right now [00:18] having a hard time getting the unity8 UI [00:18] but eventually got it [00:18] have to press power button at right time [00:18] rebooting to confirm [00:18] haha I think I know what makes it work on second boot [00:18] it's because it's not writing the crash file [00:19] that takes long [00:19] ah lost it again [00:20] can't wake up the screen anymore [00:21] reboot not even working :-( [00:22] lool: what are you running? [00:23] #89 + mir and ABI rebuilds from PPA + keyboard [00:23] it comes up [00:24] ok well it's super simple to crash [00:24] just press the power button like crazy [00:24] if I go slowly, it's ok [00:24] asac: this might be an existing condition [00:24] asac: want to give it a try? :-) [00:25] asac: if you're on #89, does it remain black if you press power button a bunch of times? [00:25] reboot worked this time [00:25] lool: my phone is busted ... i can install latest again [00:25] I guess I was a bit unlucky on first and second boot [00:26] lool: so you say now its stable? [00:26] asac: I think I was a bit unlucky in tsting on my previous install and on second boot here [00:26] I also understand why second boot works [00:26] it's not apparmor related, it's just that there's an unity8 crash file in place [00:27] so a second one isn't generated and it restarts quickly enough in the startup sequence [00:27] I just did a phone call [00:28] that worked [00:28] ok [00:28] asac: so I tested basic functionality [00:28] it looks like #89 [00:29] will land these parts after testing keyboard just a bit [00:29] ok [00:30] == Publishing mir == [00:31] win 75 [00:32] lool: leave a few channels :) [00:39] I have super good news [00:39] unity8 doesn't crash once I take hud in [00:39] and the hud crash is indeed gone [00:39] so we're likely NOT going to see this race on start where it writes the crash [00:39] thanks to th enew hud [00:40] Last stable build (#43), 1 min 52 sec ago [00:40] just a maliit crash now [00:41] \o/ [00:41] lool: so everything in now? [00:41] or ready for copying voer ? [00:43] asac: slowly getting into unapproved [01:03] sergiusens: landings 168 and 169 -- are the clicks in place? (dropping-letters stockticker sudoku) and was music-app built? [01:03] checking music-app [01:04] lool, music app is not supposed to go in the image as click [01:05] will we regress tests when converting them to click? :P [01:05] sergiusens: I know, just checking whether it will be in [01:05] asac: we're not converting anything this time [01:05] asac: just fixing existing tests [01:06] sergiusens: can we resume click conversion on friday only? [01:06] asac, I'm just staging things for tomorrow, not applying anything [01:06] as to leave thursday green for user experience fixes/tweaks [01:06] lool, sure [01:07] lool, stock ticker didn't make it but you can skip it [01:07] rsalveti: you're still landing libhybris? [01:08] rsalveti: spreadsheet says for image 90? [01:08] lool, it seems trunk is broken or the jenkins instance failed badly [01:10] lool, and the music app MR from mine never merged; you can skip it too [01:10] fginther, do you know why? https://code.launchpad.net/~sergiusens/music-app/manifest/+merge/190250 [01:11] fginther, this also failed for no known reason https://code.launchpad.net/~sergiusens/sudoku-app/click/+merge/190231 [01:12] fginther, from the MR there are no changes made after the approval :-/ [01:15] fginther, seems stock ticker has been skipping jenkins as trunk doesn't even debuild :-/ [01:16] fginther: ping? [01:17] cyphermox: anything I can try to help you with? [01:18] fginther, if you can re trigger https://code.launchpad.net/~sergiusens/stock-ticker-mobile-app/click/+merge/190238 would be great [01:18] cjohnston: I just wanted to make sure whether someone was working on "fixing" the autopilot tests [01:18] cjohnston: as in, more specifically, the lxc containers that run this with otto [01:19] cyphermox: is this the same issue that vila was looking into this morning? [01:19] cjohnston: so as not to interfere if I'm about to look into it and someone is already hard at work [01:19] cjohnston: maybe? I don't know what he was working on this morning [01:19] sergiusens: that MP should automagically retrigger when someone approves it [01:20] cjohnston, so stock ticker is auto triggered again? [01:20] cjohnston, the emails were a bit confusing [01:21] cjohnston, what about https://code.launchpad.net/~sergiusens/sudoku-app/click/+merge/190231 ... the message is kind of like from a ci issue [01:21] sergiusens: I'm not 100% sure.. however the first step would be that the MP needs to be approved [01:21] cjohnston, as in infra; might be the launchpadlib api returning no approved revno as usual [01:21] cyphermox: im looking for the context from this morning.. [01:21] cjohnston: what I'll be looking into is this failure case: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2484/console [01:22] cjohnston: however I'll need help to fix this: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=qa-nvidia-gtx660/2485/console [01:23] cyphermox: http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=autopilot-intel/2465/console is what he was looking at this morning [01:23] which looks to be quite similar [01:23] cjohnston, anyways, this is for future research, but http://91.189.93.70:8080/job/sudoku-app-autolanding/14/parameters/? has revno 137 and that's the last one in the MR [01:24] sergiusens: ya.. im looking at that now [01:24] cjohnston: I wouldn't say so conclusively, there isn't enough information to tell either way [01:25] sergiusens: let me see if I can kick that one again [01:26] kgunn: trying https://code.launchpad.net/~lool/hud/readd-desktop-bdep/+merge/190283 [01:26] cyphermox: do you know anything about hud? [01:26] lool: yeah, why? [01:27] ah, I misread, sure I'll review [01:27] cyphermox: Actualy was looking for some recommendation on testing it on desktop [01:27] cyphermox: is it enough if I see file menu from e.g. gedit in hud? [01:27] lool: ah. well that should probably be enough, yeah [01:28] cyphermox: review welcome then :-) [01:28] * cyphermox wonders why that b-dep was ever dropped [01:28] sergiusens, hey [01:29] doing a CI run [01:30] fginther, hey, so you can read the backlog if you like or just ignore; mhall just reapproved the failures [01:30] | is a valid candidate in http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html so, again iiuc, should [01:30] | be available soon and fix Mirv issue [01:31] fginther, but I was wondering if you ever reenabled the music app stuff [01:31] grr [01:31] valid candidate only means it passed the first stage of two [01:32] sergiusens, music-app is disabled intentionally [01:32] fginther, ack [01:33] to bad I can't get the regressed fix in :-/ [01:33] sergiusens: http://91.189.93.70:8080/job/sudoku-app-autolanding/15/console [01:33] cjohnston, yay [01:33] thanks [01:40] lool: looks fine but I'd close a bug with it, and I'd make sure the resulting package properly has bamf in its binary depends. Seems to me like it could not be there [01:43] lool: nevermind, that... worth checking but I revise my assessment, it's probably in the depends [01:44] approved, but I heard you may have to wait before ci / autolanding happens [01:44] cjohnston: so should I just leave the lxc containers alone and let vila handle it when he's online? [01:46] cyphermox: yup [01:46] cyphermox: I'll test the binaries on my desktop [01:46] ok [01:46] cyphermox: now that fginther is around, he may have more info from vila's findings [01:48] cyphermox, vila mentioned the fix was in proposed, but failed to mention the package [01:48] it's lxc [01:48] just a matter of upgrading that package and rebooting, and intel should probably be fine [01:49] cyphermox, ack [01:49] fginther: nvidia shows a different issue but it will still need to be applied there [01:51] cyphermox: I had to fix a bdep [01:51] cyphermox: qt was needed on !armhf :-) [01:51] cyphermox: mind top approving? [01:52] oh, right yeah qtbase would have been required everywhere [01:52] sure [01:53] thanks [01:54] done [01:57] fginther: you handling the package update? (just making sure, I have access to do it, but there's another issue with the nvidia autopilot machine I don't have access to fix) [01:58] cyphermox, I don't think I have access to those machines [01:58] ah [01:58] then I'll just go ahead and do it [01:58] cyphermox, thanks, I'll look into that tomorrow [01:58] for next time [01:58] sure [01:59] it's actually super easy, perhaps we can hangout tomorrow and I'll walk you through it [01:59] and you most likely already do have access [02:00] that said, if you have root or jenkins access on 10.97.0.1, could you prod the qa-nvidia files roughly to fix http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=qa-nvidia-gtx660/2486/console ? [02:01] cyphermox, let's see... [02:07] fginther: also, you aware of anything broken with jenkins-slave and why the version would have to stay at 1.509.2+dfsg-1 ? [02:08] arf, whatever [02:08] cyphermox, does it need to match the master [02:08] possibly [02:09] not sure, it's just a debian revision though [02:09] I doubt it does, I think the master supplies all the relavent jar files [02:09] I'll update, and there will be a trace here that it's what I did, so if things still don't work, I'll know to revert back to -1 [02:09] ack [02:10] have you seen that file permission issue before (it's owned by jenkins, not sure why it's giving it problems) [02:11] nope [02:12] I assume someone did something manually that caused this permissions issue [02:12] rebooting dx-autopilot-intel and qa-nvidia-gtx660 now to finish fixing lxc [02:13] go ahead, I hopefully fixed the perm issue [02:13] ack [02:13] I'll test it as soon as the machines are back up, I'll have a new container created and try to run some tests through [02:13] was that fix just an update of lxc? [02:17] yes [02:17] that and reinstalling the jenkins-slave package which was in a very inconsistent state, and blocking all updates of packages [02:17] ack [02:18] all the changes are in /var/log/apt/term.log / history.log [02:23] fginther: confirming, tests now actually do run again [02:23] cool [02:23] they still fail, but that's just because some packages are actually broken or something [02:23] :-) [02:25] lool: /var/log/upstart/otto-setup.log: libmirserver5:i386 Depends on libmirplatform [ i386 ] < none -> 0.0.14+13.10.20131010-0ubuntu1 > ( libs ) (= 0.0.13+13.10.20131003-0ubuntu1) can't be satisfied! [02:25] /var/log/upstart/otto-setup.log: Installing libmirserver6 as Depends of libubuntu-application-api-mirserver1 [02:25] ^ mir itself or some related package it actually broken, for realz [02:25] something built with an older version, I'm guessing [02:26] cyphermox: ah we need to update the list I guess [02:26] perhaps [02:26] cyphermox: didn't know / had forgotten this [02:27] I don't think it's the case though, that seems like broken versioned deps more than something wrong in the list [02:27] cyphermox: pushing [02:27] cyphermox: No I think it's explicitly listed [02:27] I've done the rebiulds in the PPA already, should be installable [02:27] yeah, I see, you're right [02:27] should I file the merge? [02:28] nope [02:28] pushed [02:28] ah, ok [02:28] I do like Didier, I push directly to cupstream2distro-config :-P [02:28] sorry [02:28] ok [02:28] well we still need to update the jobs [02:28] I will do a mp next time [02:28] cyphermox: I'm doing one [02:28] alrighty then :) [02:28] second one coming [02:28] lool: ubuntu-keyboard autopilots seem to have problems still (especially with mir), was talking to bfiller about that earlier [02:28] stacks/saucy/platform.cfg done, stacks/saucy/unity8.cfg in progress [02:29] plars: Yes, see #qa [02:29] plars: they are known borken [02:29] lool: yeah, I was hoping to add them to the daily runs, but he recommended holding off until after 13.10 [02:30] lool: if they stabilize before then though, I'd be happy to put them in [02:31] lool: planning to kick off an image tonight? [02:31] lool: ok, so I'll have platform run tests now, so we can check that something possibly goes green [02:32] ok [02:32] cyphermox: thanks [02:32] plars: yes [02:32] plars: trying to squeeze hud in [02:33] lool: going to be unity8 instead, platform goes green regardless [02:53] bfiller, ping [02:54] fginther: hey [03:42] == Building #90 == [03:43] bfiller, fginther: Would be nice to get this to land https://code.launchpad.net/~om26er/camera-app/disable_video_button/+merge/189566 -- I failed [03:43] I've put it on the plan for later this UTC today [03:46] lool: not sure why the autopilot is failing [03:47] lool: I tried it on the maguro and worked for me [03:50] lool, bfiller, those same tests failed on the image test [03:51] at least on mako, hmmm [03:54] lool, rebuilding that failed maguro test here: http://10.97.0.26:8080/job/generic-mediumtests-runner-maguro/2280/ [04:00] vila: awesome! thanks a lot, and good to learn again about the system === bfiller is now known as bfill_afk [04:01] Mirv, morning [04:01] fginther: "evening" :) or night. [04:04] ok, so https://launchpad.net/ubuntu/saucy/+source/lxc/1.0.0~alpha1-0ubuntu10 is in release pocket so otto should get fixed [04:20] launched unity8 + indicators builds, should arrive probably a bit after #90 is ready so I can start testing those [04:20] Mirv: don't expect autopilot tests to work [04:21] Mirv: seems like both autopilot machines are still having issues due to dbus access through apparmor [04:21] (through lxc) [04:21] Mirv: I've already updated both to the new lxc [04:23] this needs to be further looked at, but I really need to go to bed -- could you file a bug or poke the security team and/or stgraber, and provide them with /var/log/upstart/lightdm.log from the lxc containers? you can get that by starting and logging into the containers manually [04:23] cyphermox: oh, so it was not enough :( well, I'm used to skipping the check jobs [04:23] cyphermox: I know how to unpack a recent .otto file at least, possibly also can find how it was to log in [04:23] cyphermox: I'll contact stgraber [04:24] on dx-autopilot-intel, sudo lxc-start -n [04:24] on the console you'll see init complain about lightdm failing to start, and there will be logs in /var/log/upstart/ [04:25] stgraber is in the same timezone as I am though so he might not answer fast :) [04:25] ttyl [04:25] yep, I'll wait for these newest runs to finish [04:25] ok [04:26] we get builds, and we do manual testing on devices anyhow, so funnily enough it's not critical [04:26] unfortunate, though, getting green light on AP on desktop side is a very useful indicator [04:28] merging lool's manual unity-mir update [06:24] FYI I needed to do cu2d-skip -r saucy indicators indicator-messages arm64 [06:26] urgh, arm64? [06:26] Mirv: hey, it was waiting on it? IIRC, I put that in the ignore list [06:27] has indicator-messages built or published in saucy on arm64? That would be the only explanation why it's monitoring it [06:29] didrocks, FYI, check jobs because lightdm fails to start [06:30] didrocks, I honestly didn't really had time to look at the details, vila thinks it's because of apparmor [06:31] jibel: so vila is looking at it? (good way to dive into otto) [06:36] didrocks, he was but was blocked on this apparmor issue [06:36] didrocks, I'm quickly reviewing build 90 and will have a closer look [06:43] jibel: thanks a lot! still catching up on emails [06:52] jibel: didrocks: so there was supposed to be lxc update that would fix the otto problem, but cyphermox reported it did not help, and suggested to login to lxc and check /var/log/upstart/lightdm.log and provide that to security team / stgraber [06:52] didrocks: welcome back, also :) [06:52] didrocks: there were some recent arm64 churning around, but I don't know the details. maybe related. [06:55] Mirv: thanks! ok, I'll ask cjwatson abour arm64 ;) [07:23] didrocks: the arm64 port is spinning up [07:23] didrocks: so yes, it's possible you'll need to cause some things to skip it [07:24] didrocks: indicator-messages/arm64 has indeed built and published [07:24] cjwatson: ok, basically the logic is "check if a binary is available for that arch at the destination and wait for it if there" [07:24] ok, I'll hack something around for the transition [07:24] thanks for confirming cjwatson :) [07:24] Mirv: I'll handle that once I've finished with my emails ^ [07:25] as infinity put it, "Apple's told the world we all need 64-bit phones", so you folks may care about arm64 at some point rather than it being an irritation :) [07:28] cjwatson: oh, it's not an irritation at all, just need to handle the fact that the ppa doesn't build arm64 while distro seems to have done so [07:28] cjwatson: so cu2d was waiting forever [07:29] cjwatson: if you can enable arm64 on our ppas, that's fine as well [07:29] (if we have enough builders of course) [07:29] didrocks: (a) I can't (b) we're chronically short of builder time (c) I suspect the failure rate will just annoy the shit out of you right now :) [07:30] cjwatson: ok, so better for us to skip it, got it [07:31] we have three builders which we've temporarily (?) stolen from PES, which occasionally throw random faults, and which are desperately trying to get through as much of the backlog as possible before release now that we have a mostly-working toolchain, and one sim which takes hours to even get out of bed but which can be used in desperation cases [07:31] yeah, so sounds the saner path is to skip it for saucy in the code [07:31] * didrocks adds a note [07:38] main goal is to have enough in saucy that we can demonstrate that we're making an effort to catch up with other distributions, and that users and customers/partners (we have some already, this is a funded port) can experiment [07:53] morning [07:54] Mirv: oh wow thanks for noticing and doing it! [07:54] * ogra_ dances around bug 1234743 [07:54] bug 1234743 in linux (Ubuntu) "omapfb module floods system with udev events on samsung galaxy nexus" [Undecided,Incomplete] https://launchpad.net/bugs/1234743 [07:55] seems CM has a patch for it [07:56] i guess i'll have to build a kernel today [08:00] ogra_, heeyyyyyyy there buddy! can you take a look at this super-quick? https://code.launchpad.net/~robru/ubuntu-seeds/ubuntu.saucy/+merge/190262 ;-) [08:01] robru, wow, we seed it on desktop ? [08:01] ogra_, yes, it is going to be the new container for webapps. we desperately need this in 13.10 [08:02] ogra_, it means we get to shuffle off a lot of crappy firefox/chromium distropatches [08:03] robru, is there an approved MIR for it ? [08:03] ogra_, yes, somewhere. let me find that... [08:03] seems it sits in universe atm [08:04] didrocks: could you do a packaging ack? http://10.97.0.1:8080/view/cu2d/view/Saucy/view/Unity8/job/cu2d-unity8-saucy-3.0publish/lastSuccessfulBuild/artifact/packaging_changes_unity8_7.82+13.10.20131010-0ubuntu1.diff - both dependency additions are from adding the albumart support [08:04] or ogra_ if he's quick [08:04] ogra_, https://bugs.launchpad.net/ubuntu/+source/unity-webapps-qml/+bug/1206268 [08:04] Ubuntu bug 1206268 in webbrowser-app (Ubuntu) "[MIR] unity-webapps-qml" [Undecided,Fix committed] [08:06] robru, lol, thats the same bug, but i was blind and didnt notice the [MIR] in the title :P [08:06] ogra_, haha, yeah [08:09] ogra_, did I do the seed correctly? some seed items are in (parentheses) and some aren't, it wasn't clear to me what the distinction is [08:10] robru, parentheses mean its a recommends (you can uninstall it without tearing ubuntu-desktop out with it) [08:10] ogra_: you seeing icons in "more suggestions" on #90? [08:10] popey, havent started on 90 yet [08:10] Mirv: sounds sane enough, +1 [08:10] ogra_, oh, ok. yes, then my MP should be good then [08:11] robru, already merged, regenerating ubuntu-meta now [08:11] ogra_, THANK YOU! and goodnight ;-) [08:11] sleep well :) [08:11] robru: good night! :) [08:24] hi! [08:24] :) [08:25] psivaa: are we retrying 90? :) [08:25] in particular notes-app feels like it is a regression again [08:25] asac: i'll do that now.. just starting [08:25] hi didrocks ! [08:26] didrocks: welcome back! :) [08:26] congratulations didrocks! Welcome back! [08:26] hey asac! thanks ;) [08:26] thanks ev ;) [08:26] condolenc^Wcongrats didrocks ! [08:26] asac: what's wrong with notes app? [08:27] ogra_: would you think you could do the click list thing easily in the manifest? [08:27] :) [08:27] ogra_: would be nice to see the click updates in your diff then :-) [08:27] popey: in build 90 the AP test failed or didnt run. guess we need to retryu that job [08:27] lool: are you still awake? [08:27] asac: http://reports.qa.ubuntu.com/smokeng/saucy/touch_ro/mako/90:20131010:20131010/4646/notes-app-autopilot/ [08:27] pass there? [08:27] ogra_: ahah, thanks a lot :) [08:27] lool, will need a few changes, livecd-rootfs needs to spit it out on the livefs builder and cdimage needs to learn about copying them [08:28] lool, i'll try to do that today [08:28] popey: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/90:20131010:20131010/4644/ [08:28] popey: we are looking at touch_mir only nowadays [08:28] ah [08:29] oh, hmm, wait, i can probably inject it directly into the manifest file [08:29] oof [08:29] * ogra_ will think about this after the call [08:31] ogra_: cool [08:31] asac: I had a long night! [08:33] boooo [08:33] (server disconnects) [08:33] ogra_: yeah, sticking it in manifest directly in the same format is probably sane enough [08:34] ogra_: as long as anything generating changelogs is prepared to at least tolerate the packages being missing, and possibly later fetch changelogs from the app store or whatever [08:34] cjwatson, well, thats jibel's prob, my changelog only is a manifest diff [08:34] he does the detailed stuff [08:35] it may be his problem but I'm sure you don't want to break it at this point ... [08:37] no, indeed [09:06] hi, would anyone know what's up with the jenkins jobs for core apps? It seems Jenkins reviews don't get triggered again? Here's an example -> https://code.launchpad.net/~rpadovani/ubuntu-clock-app/09Oct2013/+merge/190139 [09:06] == Building #91 == [09:06] dpm: Some seem to have been disabled, not sure why [09:08] lool, no idea why, either. Would someone be able to help to reenable them so that we don't have to wait until fginther is online? [09:12] dpm: I technically might be able to [09:13] dpm: do you have a list? [09:13] dpm: I'm not sure whether it's a good idea though [09:14] dpm: we disabled a bunch of upstream merger jobs because the tests are failing everywhere [09:14] dpm: so it's not helpful to send failures [09:14] dpm: fginther reenabled some known good ones [09:15] dpm: let's wait til he's up to confirm these? [09:15] lool, ok, in that case, sounds good, thanks [09:15] dpm: I tried to run a music-app one by hand overnight, and got some failures [09:15] popey: hey! did you start dogfooding a little bit 90? can you try the usual tests and keep us posted? [09:16] asac: notes on mako_mir has run (24 pass 1 fail) waiting in transit to dashboard [09:17] psivaa: 1 fail == systemsettle? [09:18] asac: no, it's a test fail different ones during different ones.. flaky tests. [09:18] didrocks: i have started, yes [09:18] didrocks: icons missing in more suggestions. [09:18] asac: rerunning may make the passrate low being that flaky [09:18] popey: turn on the Internet! :) [09:18] psivaa: i thought we fixed notes-app flaki tests [09:18] inet addr:192.168.1.120 Bcast:192.168.1.255 Mask:255.255.255.0 [09:18] it is on [09:19] popey: on touch_ro ... e.g. this is a new flaki one [09:19] 64 bytes from 8.8.8.8: icmp_seq=1 ttl=46 time=34.3 ms [09:19] this is on mir [09:22] asac: not sure about that.. imho test may appear flaky due to other changes as well.. e.g mir etc [09:23] psivaa: maybe starts with the 0% ones [09:23] popey: oh, good, I was about to ask about that since I was testing a new click version and icons are missing :) [09:23] so not my fault [09:23] asac: psivaa: is 0% on unity8 expected, even without the staged fix? [09:23] (missing for newly-installed apps too) [09:23] didrocks: yes [09:24] didrocks: my understanding is that its supposed to get much better with the stale socket fix [09:24] didrocks: yea we have not made unity8 run fully with mir yet [09:24] didrocks: 167 [09:24] http://imgur.com/uoOfi0D <- missing icons [09:24] asac: yeah, ok, I'm up on that one :) [09:25] popey: do you know if after enabling wifi (I had no network at first), I need to reboot to see the more suggestions? [09:25] (I never noticed that) [09:25] maybe because of my script making immediately the image rw + reboot [09:25] didrocks: usually locking / unlocking may show them [09:25] lool: the powerd fix that is supposed to improve performance is in 90? [09:26] i have seen more suggestions not show up on first boot [09:26] * didrocks tries [09:26] popey: usually maybe… not now then :p [09:26] * didrocks reboots [09:26] lock / wait / unlock then ☻ [09:26] too late, trying to boot now [09:27] ah better [09:28] popey: interestingly, the same icons are not appearing for me, but the other are [09:28] maybe a server side issue? [09:28] bug 1237875 [09:28] bug 1237875 in unity-lens-applications "Many applications have no icons" [Undecided,New] https://launchpad.net/bugs/1237875 [09:28] popey: thanks! [09:28] np [09:28] popey: let's continue dogfooding and check in few minutes if you think we can promote that image [09:28] oooh! [09:29] more suggestinos just disappeared [09:29] then re-appeared [09:30] wish unity-scope-click.log had timestamps [09:33] popey, confirmed it's new on 90, did you file a bug or should I? [09:34] doing now [09:37] jibel: bug 1237886 [09:37] bug 1237886 in Unity 8 "More suggestions scope disappears/re-appears" [Undecided,New] https://launchpad.net/bugs/1237886 [09:39] * popey wonders if someone gave the backend a kick, the icons appear now [09:41] popey: you did lock/wait/unlock? [09:41] popey: apart from that one, any other worrying issue that should prevent us from promoting it? [09:41] jibel: ^ [09:42] its quite a bit slower [09:42] than previous Mir image? [09:42] or compared to SF? [09:42] no, than SF [09:42] yeah, that's an expected (and acceptable from the executives) regressions [09:42] ok [09:43] * didrocks isn't on the same server than popey maybe… :p [09:43] popey, James tait fixed the icons some minutes ago [09:43] ah [09:44] didrocks, 90 LGTM, not too many crashes (don't enable/disable BT from u-s-s) and the main regression found is acceptable [09:45] jibel: main regression being bug 1237886? [09:45] bug 1237886 in unity-scope-click (Ubuntu Saucy) "[regression] More suggestions scope disappears/re-appears" [Critical,Confirmed] https://launchpad.net/bugs/1237886 [09:45] didrocks, yes [09:45] popey: ack for you as well? ^ [09:45] yeah, agreed. it seems pretty good here too. [09:45] jibel: thanks! [09:45] ok, let's promote it then [09:45] didrocks, it is also a bit more responsive than previous build but I didn't stress the device [09:45] jibel: don't stress it too much IMHO (j/k) ;) [09:45] where is ogra? ;) [09:46] popey: do you kno who else can promote the image? (apart from stephane) [09:47] I agree, a bit more responsive than 89, probably thanks to the powerd part [09:47] ogra_ was having freenode issues, he's on other irc networks ☻ [09:47] my bip server went crazy [09:47] didrocks, lool, rsalveti, serguiens, infinity, cjwatson, stgraber and me at least [09:47] ogra_: ready to promote #90 please? :) [09:47] I think popey need to send an email? [09:47] i shall [09:48] thanks guys! [09:48] did qanyone test on maguro ? [09:48] * ogra_ didnt yet [09:48] i have not [09:48] ogra_: I don't think that we are going to block on maguro anymore (now that we have Mir) [09:48] is 90 == 20131010 ? [09:48] didrocks, ?? [09:48] ogra_: we know we'll regress (or did the world change?) [09:48] why would Mir change anything [09:49] like, it was going to be slow as hell [09:49] its slow, like it is on mako ... but beyond that it works [09:49] popey, yes [09:49] and i think i have a fix for the slowness later today ;) [09:49] (for the driver side at least) [09:49] didrocks, in any case, many of us have a maguro and need it to work [09:50] ogra_: oh great! do you think you can give a 5 minutes tests? [09:50] I was thinking that we were going to call it unsupported and be done [09:50] for development, fixes and testing ... probably not as a daily phone though :) [09:50] (the fact that for development people are using it is a right pick) [09:50] ogra_, what about grouper? [09:50] yeah, its downloading ... give it 20min [09:50] ogra_: ok, keep us posted, thanks! [09:50] grouper at least booted but I haven't tried tests [09:51] apart from installing a couple of click packages [09:52] seb128, grouper and manta are nice to have [09:52] second prio atm [09:52] k [09:52] grouper should at least work again though [09:52] apart from the bluetooth issue [09:52] that's a good news ;-) [09:53] thanks jibel [09:53] ogra_: http://people.canonical.com/~ogra/touch-image-stats/20131010.1.changes, is that automated or you kicked it manually? [09:53] (the image build) [09:54] thats automatic, there is a cron job checking cdimage for new images every 5 min [09:54] ogra_: I'm speaking about starting the build [09:55] oh, thats manual [09:55] ok, you did it :) [09:55] ? [09:55] for #91 [09:55] i didnt build any images in three or four days [09:55] hum [09:55] maybe lool did then? [09:55] i left that to loic the last days [09:55] ok [09:55] http://people.canonical.com/~ogra/touch-image-stats/20131010.1.changes is that 91? [09:55] i.e. not 90 [09:56] popey: yeah, 90 is 20131010 [09:56] kk [10:02] ogra_: ping when published pls [10:10] I'm back [10:11] didrocks: I started a build after the call as noted here [10:11] didrocks: the 10 one is from "late yesterday" aka 4:30 this morning or so, and the .1 one is from just after our call to pick up unity8 and run the image tests [10:11] asac: powerd fix is in #90 yes [10:12] asac: it's improving performance when resuming from deep sleep (you have to wait some minutes to reach this state) [10:13] oh, crap, i think it installed 91 now [10:13] ah, no [10:24] 62 0 0 0 100% [10:24] ups [10:24] odd thing :-) [10:24] cleaning the landing spreadsheet [10:24] I'm about to step out for lunch [10:25] image #91 works nicely here [10:25] will play with it some more [10:30] saving contacts doesnt work in 90 ... [10:30] diapler app doesnt come up with 90 [10:30] (i can take calls and can tap the dialer app to make it start with the currently running call) [10:31] So the -pthread change is confirmed NOT to fix maliit crashes [10:31] ogra_: dialer-app worked for me in #90 this night [10:31] err morning [10:31] * lool tries outgoing call [10:31] well, it just doesnt come up if you accept a call [10:31] yeah works [10:32] ogra_: ah [10:32] i can make the call and all [10:32] it just doesnt go on screen [10:32] trying [10:32] ogra_: oh is it meant to? [10:32] i have to start it via the icon druing the call to see it [10:32] yes [10:32] on SF it used to come to the front [10:32] right after tapping accept [10:32] ogra_: But I think we're using snap decisions now? [10:33] ogra_: Anyway it was always like this for me :-) [10:33] let's see if it's per design or not [10:33] it never was like this before [10:33] right [10:33] you still can recieve calls [10:33] so at most its a UI regression [10:34] 90 looks good otherwise ... the click lens seems a bit crashy [10:34] (it vanishes and re-poüpulates quite often) [10:34] popey, ^^^ [10:34] didnt you also see that on mako ? [10:34] ogra_: bug already filed about that [10:34] bug 1237886 [10:34] bug 1237501 in unity8 (Ubuntu Saucy) "duplicate for #1237886 [Regression] Expanding Installed category does not show all the installed apps at once" [High,Confirmed] https://launchpad.net/bugs/1237501 [10:35] oh, no thats the other one [10:35] no, that is it ☻ [10:36] the category completely vanishes and re-populates if i close an app [10:36] oh [10:37] the close button goes away again if you swipe left/right [10:37] nice [10:38] yes [10:38] ok, 90 looks fine to me [10:38] well [10:38] FSVO fine [10:38] :P [10:39] :D [10:40] ok, 90/20131010 promoted ... give it a few to mirror ... [10:40] lool, didrocks, asac ^^^ [10:41] 71.6MB downloading \o/ [10:42] mail sent [10:43] thx [10:43] \o/ [10:43] thanks ogra_, popey [10:44] * ogra_ goes to play with maguros graphics driver now [10:44] guys, but 90 is not the image we want... we want 91 for todays release candidate :) [10:44] or even 92... the fixes [10:44] asac, ergh ... [10:45] yeah, let's cross fingers for 91 AP results [10:45] 89 was better than 90 [10:45] now we cant promote 89 anymore :) [10:45] but nice to have a first Mir image promoted ;) [10:45] yeah thats good [10:45] right [10:45] and it isnt that bad [10:45] * asac just sees rick coming screaming that the click scopes etc. are regressed [10:45] psivaa: you are watching the AP results for 91? [10:45] at leas using it here [10:45] well, 89 was better from the user experience :) [10:45] anyway [10:45] ttyl [10:46] didrocks: yes [10:46] didrocks: we really need to get the scope regressions under control somewhat [10:46] thanks psivaa [10:46] yw :) [10:46] didrocks: ralsina is off... alecu have to help out [10:46] asac: yeah, I'm dealing with it right NOW! ;) [10:46] ok [10:46] thanks [10:46] yw! [10:47] ogra_: cool [10:47] asac, didrocks: is it the missing icons? That was a server side issue from what I saw yesterday (and yes, I am off but I cheat a little) [10:47] ok, trying a dbus workaround for maliit [10:48] ralsina: dont do that. we also have apps disappearing from app lens etc. [10:48] asac: oh, fun [10:48] ralsina: BTW sorry I was wrong, another u1-credentials is in the pipe, but we can live without it today and land it between friday and wed [10:48] ralsina: get some rest! [10:48] ralsina: the other one is about more suggestions category randomly disappearing [10:48] ralsina: regression from 89 to 90 [10:48] didrocks: probably crashing? [10:48] yeah, and go away ralsina, get some rests ;) [10:49] * asac thinks that if stuff disappears its usually that the scope crashed etc. [10:49] lool, didrocks, asac: ok, alecu can handle that, I am off now :-) [10:49] asac: nothing is /var/crash [10:49] ralsina: ttyl! :) [10:50] 12:50:15 mhr3_ | didrocks, fix for that was merged in click scope yesterday [10:50] on maguro i have a good bunch of crash files [10:50] asac: ^ [10:50] * didrocks looks [10:50] locally i mean [10:51] i always have crash files [10:51] yes [10:51] i only have them since Mir [10:51] asac: ogra_: I have 2 of them, but not related to this, I meant [10:51] yeah since mir i have even more [10:51] http://paste.ubuntu.com/6217497/ [10:51] sil2100: it seems there are work for you :) [10:51] um [10:51] I updated my daily phone from 89 to 90 [10:51] system 705 1.3 2.8 67728 54072 ? Sl 09:59 0:43 /system/bin/surfaceflinger [10:51] no mir [10:52] sil2100: https://code.launchpad.net/~ubuntuone-hackers/unity-scope-click/trunk, hence we have rev 73 at least, let me add a landing plan [10:54] hum, the history tab I took great care is completely removed now (of the landing plan) [10:56] didrocks: the DONE one? [10:56] didrocks: it was completely broken this morning, I just removed it this morning [10:56] ok [10:57] all my updates, snig! [10:57] snif ;) [10:57] didrocks: Sorry, I should have checked with you first on this, I thought it was still my broken query [10:57] didrocks: everything was squashed in one thing, it was unreadable [10:57] didrocks: I can try fishing it back though [10:57] didrocks: really sorry if I dropped something useful, it looked so much like my broken page :-) [10:58] lool: no worry, I wanted to follow what you landed, but if it was broken [10:58] bah, got some error in unity 7 testing, surely not from indicator-messages so I guess I didn't reboot or something [10:58] so.. chaps.. why am I still on surfaceflinger on #90? [10:59] sil2100: before you land the qtorganizer5-eds, can you please check with me? [10:59] popey: hmm, I flashed #90 this morning and .display-mir was there and no flinger process [10:59] oh balls, my phone is on 83!? [10:59] didrocks: lool: so 92 should continue to be only on regressionns from mir [10:59] i see eds is now in flight [10:59] popey: forgot --devel-proposed? [10:59] I updated to 90 though. system settings lied [10:59] asac: yeah, noticed that, hence the ping to sil2100 [10:59] ok thanks [11:00] if its not in archive yet its good [11:00] didrocks: double check that that the other item for 92 is wanted [11:00] 174 [11:00] Mirv: no, updated from system settings [11:00] asac: I think 174 is wanted for Mir crashes (from what I read on #phablet) [11:00] didrocks: you can untick the INIMAGE #90 to see it in the main sheet [11:00] right [11:00] * popey reflashes [11:00] didrocks: do we know whats going on with landing 131 [11:00] that one is big RED still :) [11:01] bbiab school [11:01] asac: I first want to get the scope thing under control (and no other landing), wait a minute :p [11:01] didrocks: sounds good. [11:04] * ogra_ twiddles thumbs while the kernel builds [11:04] so exciting ! [11:07] ogra_: ah sword fighting practice time? [11:07] dpkg-deb: building package `linux-image-3.0.0-3-maguro:armhf' in `../linux-image-3.0.0-3-maguro_3.0.0-3.18_armhf.deb'. [11:07] :D [11:09] kerne build? [11:09] why another one [11:09] sil2100: not seeing the pings? ;) [11:09] asac, testing the ultimate fix for the maguro driver ;) it will be a lot more performant with this i suppose [11:09] ok... put in a landing ask though [11:10] asac, once i know it works :P [11:10] ah ok [11:10] thought it was already in archive :) [11:10] thats just my local build ... [11:10] nah, i would never do that [11:11] * ogra_ reboots and crosses fingers [11:16] bah, black screen :( [11:17] hmm, SF works [11:17] with one frame per second :( [11:20] sounds like a failed landing attempt [11:20] a tip from Savi_q that service stop powerd before trying to test unity8 [11:20] asac, sounds like Mir doesnt do what i was told it does too though [11:20] trying that with my http://pastebin.ubuntu.com/6217575/ to see if some success occurs [11:21] ogra_: never believe what people tell you :) only believe in code [11:21] asac, well, according to the Mir guys they dont use the VSYNC uevent ... [11:21] the patch i'm trying moves VSYNC from uevent to a sysfs node [11:22] SF still starts ... Mir doesnt [11:23] aha [11:23] W/SurfaceFlinger( 659): Timed out waiting for hw vsync; faking it [11:23] so SF knows how to fake it :P [11:35] * ogra_ guesses he needs rsalveti to take a deeper look [11:44] dpm, lool, automatic upstream merger was disabled for apps which have failing tests after the switch to mir. The list was sent to ubuntu-touch-coreapps@lists.launchpad.net [11:46] fginther, thanks. Just came back from a trip and it took me a while to catch up on e-mails. I've just now seen the e-mail and indeed the clock branch is one of the ones being still disabled. [11:47] fginther: thanks [11:48] * fginther will be back in an hour [11:49] * Mirv got sort-of 7/22 success with Unity8 on mir with http://pastebin.ubuntu.com/6217659/ [11:49] * ogra_ is sad ... i had high hopes for that patch :/ [11:51] hi didrocks, lool, asac [11:51] I'm following two bugs that you mentioned above: [11:51] one, icons missing from more suggestions [11:52] two, installed apps being hidden and reshown for a few seconds [11:52] Mirv: still a start :p [11:52] alecu: excellent! I think we have a good understanding on both, but please track them :) [11:52] yep, that's a start [11:52] alecu: so, mhr3_ told me the second one will be fixed in switching to latest scope-click? [11:53] didrocks: yes, dobey worked on a fix for that. Let me fetch the bug numbers, to see if we are talking about the same. [11:53] didrocks: I forgot to highlight this in the meeting, but mediascanner FTBFS in PPA and the previous version was too but differently [11:54] will ping thost-r [11:54] lool: ok, thanks! (but get some rest first ;)) [11:54] * didrocks is going to handle the clicky scope thing himself as sil2100 is still not around :/ [11:54] popey: free for some tests with me? [11:55] ya [11:55] didrocks: you want to try landing click stuff? [11:55] lool: the click scope for fixing the more suggestion [11:55] right [11:55] popey: let me check the status of what's in the ppa, uno momento [11:56] there are various small regressions since some unity8 updates that triggers more refreshes I think [11:56] didrocks: do i need a RW phone for this? [11:56] popey: latest image + RW yeah [11:57] popey: ok, the idea is to remove unity-lens-application from your phone [11:57] popey: and add ~ubuntu-unity/daily-build ppa, installing unity-scope-click [11:57] (to uprade to latest version) [11:57] then reboot [11:57] ok [11:58] didrocks: I don't think the latest scope fixes "more suggestions". That's a server issue [11:58] popey: for the tests, I wrote those I think that are useful (line 175 in https://docs.google.com/a/canonical.com/spreadsheet/ccc?key=0Au6idq7TkpUUdGNWb0tTVmJLVzFZd0doV3dVOGpWemc#gid=0) [11:58] alecu: oh? mhr3_? ^ [11:58] didrocks: https://bugs.launchpad.net/ubuntu/+source/unity-scope-click/+bug/1237834 [11:58] Ubuntu bug 1237834 in unity-scope-click (Ubuntu) "[regression] Not all click apps in 'More suggestion' have icons" [Critical,Fix released] [11:58] alecu: there are 2 things [11:58] icons not showing up [11:58] (yeah, seems to be server-side) [11:58] didrocks: kk [11:58] and whole "more suggestions" not appearing [11:58] didrocks: the latest scope fixes another thing, which is momentarily disappearing installed apps [11:58] (the category itself) [11:58] didrocks, that's about the icons [11:59] didrocks, it will fix the "flickering" [11:59] mhr3_: flickering being, no "more suggestions" at all, right? [11:59] didrocks, it disappears, reappears... etc etc [11:59] didrocks: no: flickering means half the apps (the click ones) dissapearing for a second, then reappearing [12:00] hum, seems different then [12:00] 12:49:54 didrocks | mhr3_: more seriously, it seems that between image 89 and 90, we have "more suggestions" category randomly [12:00] | disappearing [12:00] 12:50:00 didrocks | (for the apps) [12:00] popey: it's what you saw as well, right? ^ [12:00] i put a video on the bug showing what I saw ☻ [12:00] this is what I'm seeing too; some pieces of the UI disappear and appear fairly often [12:00] the camera never lies [12:01] and I get some display glitches, but feel more like presentation / logic issues than low level mir stuff [12:01] same here lool [12:01] not seen graphical glitches as I'd define them [12:02] popey: what's the bug you put that on? [12:02] popey: I lack a name for that, but basically I'm sometimes seeing graphical elements overlap [12:02] bug 1237886 [12:02] that shouldn't be [12:02] bug 1237501 in unity8 (Ubuntu Saucy) "duplicate for #1237886 [Regression] Expanding Installed category does not show all the installed apps at once" [High,Confirmed] https://launchpad.net/bugs/1237501 [12:03] bah [12:03] I was on lunch, super early today [12:03] https://www.youtube.com/watch?v=QnW5fkqZpKc alecu [12:03] thanks [12:03] didrocks: the qtorganizer stuff is tested and ready to go, I didn't publish it as I didn't get the consensus if we should publish anything non-mir-fixing or not [12:04] sil2100: no, hold on on that one (but great to have it staged though! ;)) [12:04] we only take mir-fixing stuff first [12:04] didrocks: is it breaking anything ;p ? [12:04] Ah [12:04] sil2100: I don't care, it's a risk :p [12:05] btw. anyone else encountering https://bugs.launchpad.net/unity8/+bug/1237838 ? [12:05] Ubuntu bug 1237838 in Unity 8 "Application scope icons not loading sometimes on Mir" [Undecided,New] [12:05] didrocks, popey bug is what dobey was working on yesterday. We suspected it was the old "infinite loop" bug, so we reopened that: https://bugs.launchpad.net/unity-scope-click/+bug/1233756 [12:05] Ubuntu bug 1233756 in unity-scope-click (Ubuntu) "Possible infinite loop in scope search" [Undecided,New] [12:06] alecu: I'm still worried on rev 72 though [12:06] why refactoring so close to the release? [12:06] sil2100: already opened, server-side issue (and under fix) [12:06] * alecu looks === alan_g is now known as alan_g|lunch [12:07] didrocks, the more suggestions disppear, reappear from popey's video is fixed with 73, the icons missing is server issue (when more suggestions is actually there), anything else is yet another bug [12:07] mhr3_: yeah, I didn't see it at all for several seconds, so unsure [12:09] didrocks: ah, by rev 72 you mean the moving of the old app scope to inside the click scope. [12:10] yeah [12:10] do we see more unity8 crashes beyond in webbrowser? [12:10] it seems quite a risky move few weeks before the release [12:10] didrocks: yes, it's a huge change, but afaik it has already landed, right mhr3_? [12:10] alecu: it didn't [12:11] well, it's in trunk [12:11] but not in archive/image [12:11] hud is still crashing in 91 [12:11] what does this fix? can we back that out and do that next cycle? [12:11] http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/91:20131010.1:20131010/4658/share-app-autopilot/ [12:11] alecu, it's not in the image yet [12:11] afaik [12:11] my phone is flickering like mad now.. [12:11] alecu, it is in the daily ppa though [12:11] thostr_1: hey, hud is still crashing in 91: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/91:20131010.1:20131010/4658/share-app-autopilot/ [12:11] popey: with the changes? [12:12] didrocks: so, can I help with something? ;) [12:12] sil2100: I'm trying to decipher if we do that scope transition or not TBH ^ :) [12:12] didrocks: ok, looking into it right now [12:12] didrocks: yeah [12:13] mhr3_: mind helping popey? ^ [12:13] if it's due to the code change, I would say, let's back out rev 72 [12:13] popey, video pls :) [12:13] it doesn't seem to be the best time for those changes [12:14] it won't have anything to do with 72 [12:15] mhr3_: uploading [12:16] ev, ping [12:16] Saviq: pong [12:16] ev, hey, is it normal I had to comment out -proposed and -updates ddebs and add deb-src for the retraces with daisy? [12:17] ev, retracer would complain about 404s and no sources otherwise [12:17] oh interesting [12:17] mhr3_: it's like running a 60 hz game on a 50hz telly [12:17] there is a bug when -updates is used too early, but this doesn't sound like it [12:17] which ones were 404ing? [12:18] sil2100: do you mind trying this as well? [12:18] popey: funny, that's what the software-center icon has always reminded me of [12:18] sil2100: apt-get install unity-scope-click on latest image from the ppa [12:18] sil2100: removing u-l-a + reboot [12:18] I genuinely thought the lines across the middle of it were a graphical glitch for the longest time [12:18] didrocks: is 90 the latest? [12:19] mhr3_: https://www.youtube.com/watch?v=Vqq7ag7QxDg - can't see it [12:19] hah [12:19] sil2100: 91 is the latest, but I think 90 is fine enough [12:19] bah, stopped flickering now [12:20] popey: I always see your videos as marketing one with your voice :p [12:20] haha [12:20] i have about 50 videos which only show bugs [12:20] they're all unlisted otherwise my channel would look like youtubes best ubuntu haters channel [12:21] didrocks: trying, rebooting now [12:23] popey, hmmmm, don't think it's in any way related to the click [12:23] yeah [12:23] mir thing? [12:23] no idea, could be [12:24] popey, mhr3_: once I get my phone booted up again I'll inform if I get the same flickering [12:24] popey, can you rep it with each reboot? [12:24] ev, -updates and -security ddebs [12:27] popey, mhr3_: on image 90 with u-l-a removed an u-s-c 0.1+13.10.20131010-0ubuntu1 installed I see no flickering [12:27] didrocks: ^ [12:28] sil2100: ok, can you please try launching an app/installing/removing, see previews, check that all apps (especially defaults one that are .deb are shown?) [12:28] Saviq: when was this? They were missing, but slangasek pointed it out and pitti fixed it this morning. [12:28] ev, yesterday [12:28] (I'm presuming this is 13.10 armhf) [12:28] ah right, do try again :) [12:28] ev, so that might be fixed already :) [12:28] whoop [12:28] didrocks: ok, so testing the click scope altogether? Sure thing, doing [12:29] ev, btw, feels like this email could go to ubuntu-phone at least - if not ubuntu-devel even [12:30] Saviq: asac asked me to dup it to ubuntu-phone and I did yesterday [12:30] sil2100: please [12:31] ev, you're ahead of me... [12:31] :D [12:32] mhr3_: can't repeat it now [12:34] popey, so... mir :) [12:34] probably [12:35] didrocks: seems okay here. rebooted a few times, installed apps, icons still there, apps still run [12:35] popey: and default apps are shown? [12:36] (the one which are still .deb) [12:36] yes, i see browser, camera, contacts, dialler, file manager, friends, gallery, messaging, music, system settings, terminal, update manager, stock ticker, sudoku, dropping letters... etc [12:36] didrocks: it seems fine here, installing and uninstalling works - it seems to be a bit flacky, but as everything with Mir enabled ;/ [12:37] didrocks: by 'flacky' I mean, sometimes I need to scroll the screen up and down to see 'More suggestions' category ;p [12:37] popey: sounds good! sil2100 confirmining seeing those? ^ === alan_g|lunch is now known as alan_g [12:38] sil2100: seems more Mir-related, right? it was already the case? [12:40] No application seems missing [12:45] sil2100: ok, great! [12:45] sil2100: let's get that scope click in please ;) [12:45] psivaa: ok, next issue, it seems that we still have webbrowser-app tests results being as low as we had before [12:45] didrocks: just the click scope for now, or the whole click stack? ;) [12:45] psivaa: can you work on that/check with Saviq? [12:46] 14:45:33 Saviq | didrocks, ok, I did stop powerd, running now with powerd runnin [12:46] for context ;) [12:46] sil2100: what we tested, meaning click scope [12:47] ogra_: I'm uploading the seed transition for it FYI, nothing else was pending? [12:47] not from my side [12:48] morning [12:50] didrocks, psivaa, that's mako, mind you [12:50] Saviq: it is [12:50] didrocks: are you also taking ask 135 with the seed change? [12:51] lool: yeah, I'm talking about that one [12:52] didrocks, lool: https://code.launchpad.net/~vila/cupstream2distro/star-log/+merge/188875 can get some love ;) === gatox is now known as gatox_brb [13:01] didrocks: u-s-c in proposed \o/ [13:01] vila: ack [13:01] vila: approved but not top approved as I can't really run the tests [13:01] sil2100: great! [13:04] didrocks: Saviq: was talking to tvoss in #phablet, in the webbrowser-app test runs in mako we see unity8 crash : [13:04] http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/91:20131010.1:20131010/4658/webbrowser-app-autopilot/ [13:06] let's see what Saviq will say [13:16] * psivaa needs to pop put for an some time.. will be back a bit later === psivaa is now known as psivaa-afk [13:25] lool, didrocks, erm, could you guys drop the duplicated "stop on" line ? [13:25] (that seems to be a merge error from some time before ... the bottom one needs to go) [13:26] lool, didrocks (talking about the maliit upstart job change) [13:27] Good news everyone [13:27] tvoss confirmed we could start keyboard *after* unity8 and this fixes the maliit-server crash in my testing! [13:27] Fix uploaded [13:27] will build an image with it once in archive to get cleaner results [13:27] ogra_: Oh right [13:27] ogra_: it seemed bogus to me too [13:28] ogra_: urgh, I need coffee as well [13:28] ogra_: I've already uploaded, but please send a mp for it and I'll approve it [13:28] ogra_: stop on stopping unity8 is enough IMO [13:28] ogra_: other good news: dropped a sleep 1 :-) [13:28] lool, can i just upload without MP [13:28] yay ! [13:28] ogra_: let's not upload right now but let that one go through quickly [13:28] yeah, delighted with that one :p [13:28] ogra_: but once it's in archive, please do [13:28] k [13:29] * ogra_ wishes hs omapfb change would work better [13:29] all my hopes ... trashed ... life is so mean [13:30] huh [13:30] didrocks, Mirv, jibel: nvidia hanged up again... [13:31] Will try to terminate the job and see if things move on [13:31] sil2100: hum "again"? I probably missed an episod :p [13:31] or is it the lxc issue that was discussed? [13:31] (I think jibel is looking at the issue) [13:31] didrocks: we still seem to be having DNS issues from time to time again [13:31] * didrocks *sighs* [13:31] didrocks: well, it hanged up just as it did in the past, not sure what's the currently most trendy way of things hanging up :D [13:32] * sil2100 likes trendy hang-ups [13:32] ev: do you think we can really have all those done as a priority? ^ (as it started near our sprint in Boston), even if things are better, seems it's not fully fixed [13:32] didrocks: sil2100: it's not time to time but every time since Tuesday morning, although last night with the lxc fix it seems to have changed [13:33] didrocks: I think I CC:d you on the e-mail that was the latest on the topic, basically the new errors seen after the lxc fix went in [13:33] didrocks: the DNS issues? [13:33] Mirv: yeah, when sil2100 told about nvidia hang, I thought this was specific to nvidia (so another issue) [13:33] so it's somewhere between apparmor <-> dbus <-> lxc, apparmor and dbus saw big uploads Mon-Tue night === bfill_afk is now known as bfiller [13:33] ev: well I don't think there anything left that's actually a DNS issue [13:34] ok, so those issues are not DNS-related? [13:34] asac, plars : are we sure we're unlocking the greeter before running the tests? omer did some debugging and found the messaging-app failure due to greeter not being unlocked [13:34] sil2100: do you have the trace? [13:34] this http://10.97.0.1:8080/job/autopilot-saucy-daily_release/label=qa-nvidia-gtx660/2494/console is about the X not starting [13:34] indeed, the whole DNS resolution happened [13:34] and the lightdm log was http://pastebin.ubuntu.com/6216936/ [13:34] sil2100: why did you talk about DNS issues? [13:35] asac, plars : can't stop unity8 anymore like we used to, but assuming you know that already. greeter needs to be unlocked before running tests [13:35] bfiller, we are sure that there are hacks to unlock the screen in utah that work with SF ... [13:35] bfiller: we had that discussion just a few minutes ago [13:35] wether they still work in Mir is another story [13:35] bfiller: we always unlocked greeter [13:35] bfiller: for all app tests [13:35] bfiller: yes, we unlock the greeter but not for unity8 [13:35] bfiller: at least for a couple month :) [13:36] asac, are we sure that still works after the Mir switch [13:36] the "stop unity before app tests", never worked well [13:36] ogra_: we had that discussion as well [13:36] ogra_: yes [13:36] k [13:36] plars just triple checked afaik [13:36] also i see the same numbers locally after unlocking [13:36] that we see on the dashboard [13:37] asac: specifically with the messaging-app tests we're only able to get it to fail locally when greeter is locked, just one data point [13:38] bfiller: with mir? [13:38] bfiller: i think messaging-app was not one of the flaki ones [13:38] plars: did we have problems with messaging-app and needed to retry that? [13:38] didrocks: right, it's not DNS, been too quick with judgement [13:38] ok [13:38] asac: yes with mir [13:38] sil2100: please read logs :p [13:38] Mirv: what's up with X not starting? I might have missed some e-mails, but now intel is stuck it seems on the same thing [13:38] asac: and all the autolanding jobs are failing the same with that [13:38] Mirv: for platform stack [13:38] hmm [13:39] bfiller: let me try as well [13:39] plars: can you quadruple check? [13:39] asac: if we did it would have been earlier, psivaa-afk might know but I can look and see if it was restarted [13:39] would be nice to see that it unlocks in the log [13:39] bfiller: is messaging app converted to click? [13:39] asac: no [13:39] kk [13:40] sil2100: so if you haven't followed, all check jobs have failed since Tuesday morning (just look at http://10.97.0.1:8080/job/autopilot-saucy-daily_release/). you were CC:d in the "Re: cu2d check job failure" [13:40] plars: so where is the script that unlocks the phone? how can i run it locally? [13:40] plars, are you seeing any problems with the screen unlock code? It looks like all the upstream merger tests are failing due to what looks like a failure to introspect unity8 [13:40] fginther: thats what i want triple checked [13:40] Mirv: I quietly thought that's fixed already ;p [13:40] fginther: can you find that code and see if it works? [13:40] asac, working on it now [13:41] Mirv: then the only way is to just abort the jobs for now, yes? [13:41] fginther: thanks... keep us posted. very importrant to know 100% that we are doing it right [13:41] sil2100: yes, currently that's the only way (timeout after 2h as well) [13:41] Bummer [13:42] asac: looks like the messaging tests worked on maguro, just not on mako. It had not been restarted but I'll restart it now, and also try it locally so I can watch what it's doing on mako [13:42] bfiller: ^ [13:42] bfiller: so running it twice makes it fail [13:42] * asac reboot [13:42] plars: ok. still more and more people report that we might not have unlocked [13:42] so we should take that super serious :) [13:42] at best using the camera in the lab to observe [13:43] asac: I just ran a test with friends-app (because it's quick) and watched it unlock just fine here [13:43] plars: run it 20 times :) [13:43] asac: I am [13:43] plars: how do you run it? [13:43] asac: locally? I'm running it with the same script we use in the lab [13:43] and are you 100% sure that the code you try locally is the same as in the lab [13:43] from the lp:ubuntu-test-cases/touch branch [13:43] ok [13:43] asac: 100% [13:44] without looking? :) [13:44] asac: with or without - I'm running from the exact same branch [13:44] k [13:45] asac: there could, of course, be some corner case where it doesn't unlock - if we believe messaging app has shown this, I'll run the messaging app test lots of times in a loop here... that's what I'm doing now [13:45] asac: and I'll watch while it's doing it [13:46] yeah [13:46] plars: i am not sure... dashboard says 4 tests, with 2 fail 2 succeess [13:46] locally i get 2 success and 0 fail i think [13:46] asac, bfiller: messaging app passed on the retry for mako [13:46] plars: with 2 success? [13:46] or 4? [13:47] otherwise we have a dashboard bug i think [13:47] messaging-app-autopilot 2013/10/10 11:37 4 2 2 0 2 50% [13:47] asac: 4 - 2 of those are systemsettle [13:47] asac: where are you looking? I don't think it's shown up for maguro yet [13:48] looking at last run: [13:48] http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/91:20131010.1:20131010/4658/ [13:48] didnt know that we have 2 for ss [13:48] thats fine then and makes sense [13:48] asac: that's the old run [13:48] asac: and on mako [13:48] asac: those failed both of the messaging tests indeed [13:49] asac: but the rerun that I just did, they all passed fine [13:49] ogra_: for the vsync one disabling in upstart is fine, isn't it? [13:49] plars: do we see crashes int he rerun too> [13:49] it was already disabled [13:49] didrocks: any objection for me to kick an image build when session enters release pocket? (some minutes) [13:49] asac: same at home, but I'll keep trying to reproduce some case where it does not unlock [13:49] lool: no objection as well, I guess the seed change is already in (didn't check) [13:49] asac: yes, still crashes on unity8 and maliit-server [13:49] let me look still [13:49] ah unity-scope-click | 0.1+13.10.20131010-0ubuntu1 | saucy-proposed/universe | source, amd64, armhf, i386 [13:49] rsalveti, well, its a gross hack ... and we need the same for udev ... and the traffic will still be there and put load on the system, even if we hide it [13:50] * didrocks looks [13:50] didrocks: seems so [13:50] lool: the seed change is, not the scope [13:50] didrocks: ah no, I dont see your upload [13:50] didrocks: where's the seed change? [13:50] ogra_: right, but I believe that's fine anyway [13:50] rsalveti, i found a patch for the espresso arch that switches from uevent to a sysfs node ... this works fine on the kernel side but i cant get the userspace to make use of it [13:50] well, binary blobs :-) [13:50] didrocks: dont see an ubuntu-touch-meta upload [13:51] rsalveti, i can get SF to start but it falls back to faking vsync [13:51] lool: I didn't dput, but I wanted to see the scope in the archive first [13:51] ogra_: right [13:51] rsalveti, and Mir doesnt start at all [13:51] lool: let me look at why it's not in [13:51] sil2100: did you track it? [13:51] rsalveti, you dont think we could somehow make it read from sysfs instead of uevents ? [13:51] unity-scope-click/i386 unsatisfiable Depends: libunity-protocol-private0 (>= 7.1.2+13.10.20131010) [13:51] lool: were we under any transition? [13:51] ogra_: nops [13:51] * didrocks looks [13:52] hmpf :( [13:52] mhr3_: hum, you did add a new symbols and used it? ^ [13:52] xnox, so we will need your udev filter [13:52] seems there is no way around [13:52] didrocks: Bad! unity-scope-click/i386 unsatisfiable Depends: libunity-protocol-private0 (>= 7.1.2+13.10.20131010) [13:52] Right! [13:52] how much cpu is udev consuming? [13:52] Shit, it needed libunity, just now I remember it pulled it in [13:52] rsalveti, constantly 5-10% while the screen is on [13:53] But we'll have to do more tests for that I guess [13:53] sil2100: you should have seen that when apt-get install :/ [13:53] ogra_: right, when rendering stuff in there [13:53] Let me upgrade the desktop [13:53] rsalveti, what bothers me a bit that this hasnt been an issue before [13:53] didn't it force it? [13:53] didrocks, asac: That said, it will just drop the number of crashes on dashboard, it will not help any of the AP runs I think [13:53] since maliit respawned immediately [13:53] lool: don't you think things timed out? [13:53] ogra_: did you try filtering in udev to see what happens in the userspace side? [13:53] ogra_: ack. [13:53] rsalveti, and i dont know when it started ... i used to be able to use my maguuro for a day without running out of RAM ... bow its like 20-30min max [13:54] lool: maybe there is a long tail of cause->consequence? [13:54] well, the uevent was always there [13:54] who is leaking memory now? [13:54] rsalveti, but was it always exposed ? [13:54] always [13:54] rsalveti, upstart is ... [13:54] didrocks: possibly [13:54] didrocks: if the test starts real quick [13:54] but I kind of doubt it [13:54] ogra_: it wasn't before we flipped the container [13:54] didrocks: landing asks has a fix for EDS from renato.. is that likely to land? [13:55] in the image [13:55] we will see I guess [13:55] ogra_: still? [13:55] rsalveti, well, upstart collects it and doesnt free it, jodh works on fixing the freeing [13:55] oh, ok [13:55] but he doesnt work on fixing the root cause [13:55] popey: not today [13:55] didrocks: tomorrow? [13:55] let's try to filter it in udev and see [13:55] popey: probably, there is still this Mir blocking everything [13:55] rsalveti, well, i dont think it was there when we flipped [13:55] right, will that be final last? [13:56] rsalveti, it definitely only caused issues for me since a week or two [13:56] ogra_: it was always there :-) [13:56] the uevent [13:56] it might be because we got a newer upstart along the way [13:56] was it always on the system dbus ? [13:56] popey: depends, it's hard to know when things are getting fixed and getting ETAs TBH :) [13:56] or in udev ? [13:56] udevd didnt consume much CPU for me in the past [13:57] well, I assume udev was getting all the events in there [13:57] didrocks: trying to figure out our absolute deadline for core apps updates ㋛ [13:57] sil2100: so, you are installing that on your laptop and testing quickly? [13:57] rsalveti, could it be that hybris shielded it before ? [13:57] can't, it comes from the netlink [13:57] so that we didnt actually see it on the ubuntu side [13:57] weird [13:57] so kernel -> udev was always there [13:58] popey: the earliest is tomorrow, but plan Monday [13:58] now what happens via udev -> upstart is what caused it [13:58] well, upstart/udev/bridge isnt so new either [13:58] we got a newer udev and upstart since we flipped [13:58] *upstart-udev-bridge [13:58] didrocks: did that already, both desktop and phone are fine, preparing things for release and double checking [13:58] ogra_: we were using the old udev before [13:58] then we switched over to use the systemd based one [13:58] hmm [13:58] from the same source and such [13:58] No packaging changes [13:58] we always used the systemd one in saucy though [13:58] and we got a new major version of upstart [13:59] it switched very early in the cycle [13:59] right [13:59] sil2100: ok, please do it now [13:59] and i'm sure udevd wasnt always under the top consumers in top [13:59] it now is [14:01] lool: right. as expected [14:01] lool: i believe its still due to something weird [14:01] like: what fixes the webbrowser APs for saviq [14:01] unity8 crashing while running certainly could be one reason [14:02] but he also has those crashes [14:02] didrocks: it's in the unapproved queue now [14:02] so i believe there is more into it [14:03] rsalveti, what i find strange is that CM10.2 on espresso seems to use that sysfs change just fine (though i assume they have a newer PVR or whatever in espresso) [14:03] can someone update me on latest trickses to run unity8 testsuite? [14:03] sil2100: great! pinging the release team? [14:03] I'd like to test socket removal here [14:03] ogra_: just different pvr [14:03] pvr is a mess [14:03] k [14:06] fginther, how can we get this branch for clock merged? -> Can we enable the Jenkins merger or do we do it manually? [14:06] https://code.launchpad.net/~rpadovani/ubuntu-clock-app/09Oct2013/+merge/190139 [14:07] dpm, one moment === gatox_brb is now known as gatox [14:10] dpm, clock is passing in the latest image, I'll turn automatic merges back on [14:19] perfect, thanks fginther [14:21] asac, when I had 100%, nothing crashed - or at least not unity8 or maliit-server [14:22] asac, confirmed jibel's crash in maliit-framework https://bugs.launchpad.net/ubuntu/+source/maliit-framework/+bug/1237870 [14:22] Ubuntu bug 1237870 in maliit-framework (Ubuntu) "maliit-server crashed with SIGSEGV in QHostAddress::clear()" [High,Confirmed] [14:24] Saviq: ok. so the crashes will most likely fix it [14:24] then lets focus on the unity8 crashes... do you see that? [14:25] Saviq: can you reproduce the unity8 crashes that we see during AP tests? === alan_g is now known as alan_g|tea [14:28] asac, yeah, some - we're on it [14:28] asac, unity8 crashes - send them to me [14:28] asac, got any .crash files? [14:33] lool: thanks for the review, yes the tests expect you to sudo [14:34] lool: ask didrocks for rationale but roughly the integration tests needs to really exercise pbuilder [14:38] asac: dude === psivaa-afk is now known as psivaa === ricmm_ is now known as ricmm [14:39] asac: http://people.canonical.com/~ricmm/powerd_0.13+13.10.20131009-0ubuntu1_armhf.deb [14:39] try with that [14:48] Saviq: is there a bug for the top/bottom + left/right scrolling creating UI glitches [14:48] lool, first time I've heard of it === alan_g|tea is now known as alan_g [14:55] asac, fginther : we believe many of the failed tests are caused by this bug https://bugs.launchpad.net/mir/+bug/1238062 [14:55] Ubuntu bug 1238062 in Mir "Multiple apps autopilot tests fail on revealing the toolbar" [Undecided,Confirmed] [14:56] this is only occuring under mir and not SF [14:56] didrocks: do we have to skip arm64 builds manually in cu2d? [14:56] something causing toolbar to think it's still animating after 10 seconds [14:56] sil2100: oh right, I didn't have the time to workaround this [14:56] didrocks: since I see indicators is blocking on arm64 right now [14:56] I'll skip it for now [14:57] sil2100: yeah, please skip :) [14:57] remind me tomorrow about it [14:57] Sure :) [14:58] bfiller: awesome... letrs fix it!! [14:58] what do you need to fix it? [14:58] any poneys? [14:58] asac: think that's a question for the mir/autopilot guys :) not a problem in the apps or their tests [14:58] at least not for those failures [14:58] something in the input stack maybe? [15:03] asac, fginther : can we turn auto merge back on for messaging-app and webbrowser-app? all of messaging-app failures due to this bug and most of browser's due to this bug and https://bugs.launchpad.net/webbrowser-app/+bug/1237472 [15:04] Ubuntu bug 1237472 in unity-mir "[browser] Autopilot test cases fail due to ap tap not being recognized" [Undecided,Fix committed] [15:04] we have a bunch of MR's that need to land in each [15:05] well, i expect that app developers are looking in the stuff up and down [15:05] so i hoped they could just go there and fix it [15:05] if that doesnt work then we can wait [15:05] didrocks: so lets take the powerd bandaid by ricmm [15:05] bfiller, at the moment nothing is running correctly on mako in upstream merger, appears to be an issue with unlocking the screen [15:06] bfiller, I know that doesn't answer your queston :-( [15:06] bfiller: they dont pass the tests [15:06] fginther: haha, just tell me you'll enable them :) [15:06] bfiller: if you dont want tests, you can directly push to trunk [15:07] but we cant land stuff easily anyway [15:07] asac: we can? [15:07] wello, you can whatever you want. i dont recommend it [15:07] if the apps thejn fail APs at the staging state we will not land them [15:07] bfiller: wait till the tests are fixed and help getting them fixed [15:07] thats the best way i swear [15:08] asac: we're trying to help fix them, believe me. but we have real app critical bugs that need to be fixed as well [15:08] bfiller: make MP ready, discuss those with us individually then [15:08] bfiller: the upstream merger will not succeed [15:08] it will always fail [15:08] so it doesnt help you [15:08] ok [15:08] you will jsut retry till forever or force it into the trunk [15:10] brb, modem restart [15:18] didrocks: can you work ricmm [15:18] powerd bandaid into the landing plan? [15:18] it will allow unity8 AP to run [15:18] asac: yeah, I read it, still on the HO with lool first [15:18] with failures, but at least we will see it [15:18] (and checking why britney isn't running) [15:18] didrocks: ok sorry. [15:20] no worry [15:22] sorted [15:24] thanks cjwatson [15:25] ricmm: ok, ready now [15:25] ricmm: so, do you have your powerd package handy? [15:26] let's get that in the incoming image [15:26] Saviq: asac told me you knew about it because it was alreayd there in the past :-) [15:27] Saviq: so you pull down home scope, and while it's slowly getting up (elastically?) you swipe left/right [15:27] Saviq: than there are visual glitches in the scope [15:27] Saviq: it is the same issue that we had 2-3 month ago while i was still in miami :) [15:27] Saviq: e.g. pull the screen up or down beyond and while it bounces back move it to the next homescfreen [15:27] if you go back its corrupted [15:28] like elements floating over each other etc. [15:28] asac, heh, still busy with your pet bug ? [15:28] * ogra_ can confirm it btw [15:29] its not my pet bug :) [15:29] it just came up today as being annoybing and folks couldnt reproduce it [15:29] so i showed one way how :) [15:29] * didrocks still thinks we have bigger issues today :p [15:30] we have issues ? [15:30] asac: do you know where ricmm's code is? [15:30] didrocks: ask him [15:30] ogra_: I heard of an ungreen dashboard :p [15:30] ricmm: where is your bandaid code? [15:30] didrocks, that can be easily fixed [15:30] https://code.launchpad.net/~ricmm/powerd/suspend-if-mir-alive/+merge/190173 [15:30] final: bip,gccxml,glew,kubuntu-debug-installer,kubuntu-firefox-installer,kubuntu-patched-l10n,libisoburn,libpwquality,pnm2ppa,ubuntu-touch-meta,x11-apps,x11-session-utils,x11-xfs-utils,xinetd,xorg,zsh [15:30] using CSS [15:30] asac: what I did here several minutes ago, trying to speed this up while he's away ;) [15:30] (that'll need another publisher run of course) [15:31] didrocks: https://code.launchpad.net/~ricmm/powerd/suspend-if-mir-alive ? maybe? [15:31] hmm. [15:31] seems thats an old one [15:31] asac: we have the link now :) [15:31] oh its not merged [15:31] didrocks: good [15:31] ricmm: nobody reviewed/tested? [15:31] i installed it [15:31] see #phablet [15:31] and it fixes unity8 AP [15:32] it was tested over there [15:32] err fixes that it runs [15:32] not suceeds [15:32] ricmm: mind pushing that to trunk so that we win some cycles? [15:32] ricmm: I approved it logically [15:32] then, let's handle a build with it [15:33] asac: it's an improvment [15:33] asac, lool, can we have a bug pleae? [15:34] didrocks: can you do that for me please im really neck deep in Mir code right now [15:34] ricmm: sure, doing then [15:35] Saviq: not my discovery, happy to file on [15:35] Saviq: (could you reproduce? [15:35] Saviq: for what? [15:35] Saviq: you can file bugs on your own. we test and tell you. and we constantly see it on dashboard [15:36] if we feel the ball gets dropped with use bugs for escalation [15:36] asac: powerd fix in trunk and preparing in the ppa [15:36] nice [15:36] Saviq: i dont even know what you want a bug about [15:36] still no luck on webbrowser? [15:36] or unity8 having test failures [15:37] Saviq: https://bugs.launchpad.net/ubuntu/+source/unity8/+bug/1238094 [15:37] Ubuntu bug 1238094 in unity8 (Ubuntu) "Graphical UI glitches when swiping left/right after a pull down" [Undecided,New] [15:37] asac: ^ [15:37] ahj that one [15:37] well there a bug makes sense [15:37] jfunk should have this on his list i think [15:51] didrocks: I still see u-s-c in -proposed :| [15:51] sil2100: autopkgtest for deja-dup 27.3.1-0ubuntu1: RUNNING (Jenkins: public, private) [15:51] http://people.canonical.com/~ubuntu-archive/proposed-migration/update_excuses.html#libunity [15:51] Ah, autopkgtest I see [15:51] did you look at this? [15:52] Yes, looked at this, but still makes me wonder why it takes so long! [15:52] asac, Saviq: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/91:20131010.1:20131010/4658/unity8-autopilot/ was the best run I could get with unit8 [15:53] plars, whoa! [15:53] didrocks: do you know where landing 174 is currently? [15:53] plars, that's better than expected! [15:53] sil2100: backlogging? britney was done [15:53] didrocks: is that inflight or inarchive and do we plan to take it? [15:53] Saviq: yeah, I've seen it pass a lower amount also [15:54] asac: I didn't get ping about it, I think it's pending next image, lool didn't get positive results this morning about it [15:54] plars: yeah. but doesnt unity normally hav more tests? [15:54] asac: also, not merged to trunk yet [15:54] didrocks: can we put it to WAITING FOR CODE or something? [15:54] asac: done already [15:54] INFLIGHT previously meant someone is working on landing it :) [15:54] ah cool [15:55] asac: I don't think so [15:55] indeed [15:55] nice one [15:56] asac: I'm going to retry webbrowser in ci - running locally with doanac's change to turn the display on earlier I'm getting a pass on webbrowser [15:57] plars: turn display on is also missing? [15:57] powerd in UNAPPROVED [15:58] Saviq: how can we ensure that the display is on? can we add that in autopilot? [15:58] (well, in 2 minutes) [15:58] asac: it was normally turned on in the unlock code, but we think that was happening too late and sometimes failing [15:58] plars: can you try the two hacks with other autopilots? [15:58] plars: maybe loads now go green? [15:58] asac, I thought was possible through powerd-cli ? [15:58] asac: which two hacks are you talking about? [15:58] hmm, lool i think we should land the mallit stop on fix too, it produces crash files for me on shutdwon without it [15:58] plars: i cant see a pass though with powerd stopped and the screen is surely on while i start the test [15:59] plars: powerd hack and screen on [15:59] Saviq: yeah, but can you call that from the test? [15:59] didrocks: I'm positive about the change, but it wasn't affecting maliit [15:59] Saviq: its aweful to add that to utah, because then people will debate and not be able to reproduce exactly what is done [15:59] awful [16:00] asac: let me finish trying webbrowser in the lab on mako with the display on code moved up, iirc we discussed earlier that that's really all that should be needed [16:00] asac: and that seems to be doing well for me when I tried locally [16:00] plars: right [16:00] plars: go fo rthat and run a few other apps as well that fail [16:00] asac: if that works, and if there's enough time before the next build, sure we can kick off the whole thing again and see what happens [16:00] asac, no idea [16:00] I can not type an SMS in messaging-app anymore [16:00] plars: those that regreessed and were stable before [16:01] asac, if I can call powerd to turn display on, sure - but as lool mentioned it sometimes happens during long boot [16:01] asac, so we really just need to fix the underlying bug [16:01] plars: i think we have a couple hours to play with this [16:01] Saviq: there is a fix in powerd in the pipe to do the right thing? [16:01] Saviq: but whats the underlying bug? I think i haven't heard what the real, real fix would be [16:01] lool, asac, in that case we don't need to do anything [16:01] didrocks: meeting? ;) [16:02] Saviq: maybe try it out [16:02] Saviq: it's in PPA I think [16:02] lool: i dont knoow this addresses that [16:02] Saviq: powerd build [16:02] lool, I'm trying out 50 things at any given time... [16:02] ricmm: does your powerd also help for cases where the screen needs to be forced on etc.? [16:02] Saviq: and you complain about a 2% difference in the number of things you're trying! :-) [16:02] yes! [16:02] lool: afaiui the powerd fix just ensures that we can start unity8-autopilots after we kill unity [16:05] asac: joining? [16:06] why not :) [16:12] asac: the powerd MR just makes powerd *prevent* suspend if the display server is unavailable [16:12] for the Mir case === mhr3__ is now known as mhr3 [16:17] https://code.launchpad.net/~mterry/session-manager-touch/socket/+merge/186391 [16:24] ogra_: I commented on the branch [16:24] thx [16:25] == Building platform-api (platform stack) == [16:25] ogra_: It's kind of great and sad that mterry was already preparing a fix to not use /tmp! [16:26] too bad we dropped the ball there [16:26] yeah [16:26] should we upload the second half of the maliit fix ? [16:26] so it goes into thei image too [16:26] *this [16:36] asac, i just added an item to the landing asks sheet for bugs fixes related to online accounts... but none of it can be tested on Mir images because signon-ui doesn't work on Mir yet, greyback is looking at that [16:37] asac, but i'd rather not wait to land these, they've been tested thoroughly on SF and the packages in the current image don't work either because of the same Mir issue === bfiller is now known as bfiller_afk [16:42] didrocks: was "unity-lens-application" removed from 92, when adding the new scope? [16:42] alecu: yeah, I didn't dive into the detail, but that was handled [16:42] great [16:42] so no duplicate :) [16:42] (normally :p) [16:46] didrocks, hand is on the trigger, should i shoot ? [16:46] ogra_: please do! [16:47] * ogra_ pulls the index finger [16:47] *BOOM* [16:47] lool: adding another land request for ofono, bugfixes [16:47] (building) [16:47] lool: multimedia is still not ready to go, blocked by a bug on mako [16:47] sil2100: FYI, so landing 179 and 180 [16:48] didrocks, I think I'm missing something. I'm looking here for pending merge-backs but don't see any: https://code.launchpad.net/~ps-jenkins [16:50] fginther: unity8 at least should have one? === ChickenCutlass is now known as ChickenCutlass_l [16:52] didrocks, there was one merged 8 hours ago, should there be a newer one? [16:52] kenvandine: sounds like not ready :) [16:52] fginther: oh, if it was merged, my bad then; well, good then! :) [16:53] thx! once the rest is there we can tackle it [16:53] kenvandine: we might also take it after we go out of special mir-only fix mode [16:53] hopefully tomorrow [16:53] but better get the complete thing in so you can tell us what to expect and we can avoid bad landings that dont work [16:53] asac, i'm mostly worried about the fixes in signon-ui and the archive freezing [16:54] asac, since it's in the desktop too [16:54] didrocks, "unity-scope-click transition (currently moving to the release pocket) to handle debs and click packages " [16:54] didrocks, debs ? [16:54] the rest is touch only [16:54] kenvandine: when is it freezing? [16:54] tonight [16:54] ogra_: yeah, the click scope is handling debs file [16:54] ogra_: fine :p [16:54] ugh [16:54] and fun [16:54] i dont think that should mipact our ability to land stuff properly [16:54] kenvandine: we have to try get a FE then [16:54] you might want to start doing that :) [16:54] or check with folks before [16:54] maybe they give you a day :) [16:55] (without paperwork) [16:55] asac, you know how things are with respins [16:55] ok, will be back a little bit before 10pm [16:55] landing it will cause respin and extra work for iso testing [16:55] i hate doing that to folks [16:55] kenvandine: i dont anticipate that there will be no respin [16:55] kenvandine: also we could upload it to saucy-updates if you dont need it on the desktop image [16:55] asac, basically i saying it doesn't make things any worse than it is... it's completely broken now [16:56] asac, ok, lets hope greyback makes some progress quickly :) [16:56] no pressure :) [16:56] lol [16:56] hehe [16:56] tell him the deadline [16:56] that will help [16:56] !! [16:56] at least he might defer his dinner [16:57] greyback, ^^ that signon-ui problem on Mir is blocking a bunch of fixes :) [16:57] greyback, any leads yet? [16:58] lets hope greyback isnt working on a critical mir crash :) [16:59] kenvandine: I'm working on something else right now, but this is a workaround for unity-mir you could try: http://pastebin.ubuntu.com/6218843/ [17:00] I'm working on something else right now, so that's all I can contribute at the moment === alan_g is now known as alan_g|EOD [17:00] greyback, ok, i'll give it a try [17:01] kenvandine: the correct fix is much more invasive, so that workaround might have to do until >1.0 [17:01] .../probably/ much more... [17:02] and I suspect that workaround has bad consequences [17:02] asac: do we promote #91? [17:03] lool: not sure [17:03] lool: is it strictly better? i guess given the dashboard we need QA input [17:03] jfunk: can you assess if 91 is strictly better (in some terms :)) on manual testing side yet? [17:03] than #90 [17:04] plars: did you rerun webbrowser with the "hacks"? [17:04] its still at 7.7% ... just wondering [17:04] * asac guesses he is still prepping things for this trial [17:04] asac: yes - on mako it all passes, not on maguro [17:05] ok so waiting for dashboard propagation? [17:05] asac: you probably missed the message when I was going out of the landing call [17:05] yeah [17:05] i dropped out also pretty quick [17:05] asac: it's IMO strictly better [17:06] asac: ping Oliver and Alan to confirm if you want to promote [17:06] right. i believe the best team to say not-imo is QA :). lets see if they can give us more unbiased data [17:06] asac: no, they are on the dashboard === gatox is now known as gatox_lunch [17:06] lool: that might also work [17:06] but... :) [17:06] i think we are at a point where we want more organized sign offs [17:06] plars: oh i see now [17:06] lool: lets check in #qa etc. if anyone is around [17:07] plars: Could you drive the ubuntu-keyboard testsuites from veebers into landing and add them to the dashboard? [17:07] plars: or would you expect he knows how to do it? [17:07] plars: Oh you didn't get it [17:07] plars: fwding [17:08] lool: I haven't had a chance to take a look yet, but I saw the mail and the merge proposal, I think it was just waiting on review signoff which I can't do [17:08] plars: sent [17:08] plars: who can do? [17:09] is there anyone != new zealand? [17:10] if someone suddenly has sudden outbreak of free time (yeah, right) and knows how to launch an app from command line under mir, please update bug https://launchpad.net/bugs/1238031 so that me and bzoltan have something to start from in the morning for the SDK [17:10] Ubuntu bug 1238031 in qtcreator-plugin-ubuntu "'Run on device' stopped working" [High,Confirmed] [17:11] asac: I can change the tests locally and give it a try for me, but I can't topapprove the patch so that it lands [17:11] asac: it doesn't give me an option [17:12] asac: someone needed from phablet team I would guess [17:14] sil2100, still around? [17:17] lool, asac: are landing still happening by "stack"? or do we have more granularity? [17:17] e.g indicators, can we land one alone? [17:17] sergiusens: plars might need you rhelp [17:17] sergiusens: we can land stuff alone if it is technically possible (e.g. can be copied in theory alone) [17:17] seb128: ^^ [17:17] seb128: per source package is possible [17:18] sil2100: could you fasttrack https://code.launchpad.net/~indicator-applet-developers/hud/fix-window-stack-crash/+merge/190401 [17:18] * lool is off & [17:18] can we land indicator-keyboard (desktop only)? [17:18] sil2100: this is hud crash fix [17:18] lool, sleep well ! [17:18] there is a fix important for saucy desktop in there [17:18] and dont dare to show up here tonight ! [17:18] cyphermox, sil2100, ^ can you get indicator-keyboard landed? [17:19] lool: have a good and long night of sleep :-) [17:19] thanks [17:20] asac, plars help on what? [17:23] asac: getting an MP pushed through - https://code.launchpad.net/~veebers/ubuntu-keyboard/update_autopilot_emulators/+merge/190319 [17:26] sergiusens: what plars says [17:27] ==== image 92 is done ==== [17:27] (on cdimage ... give system-image a few to pick it up) [17:29] asac, plars can you add a mr vote saying that you tested at least? [17:29] seb128: sure. [17:30] lool: ^ so I'll push the buttons to land indicator-keyboard [17:30] sergiusens: I'm just about to give it a try here [17:31] cyphermox, he's off for the evening, but that indicator is not on touch so I doubt they care [17:32] ah, good point [17:32] yeah, fight with the release team over it :) [17:33] seb128: 0.0.0+13.10.20131010.1-0ubuntu1 ? [17:34] cyphermox, correct! [17:34] alright. we have the go from release team? :) [17:36] cyphermox, well, it's a bug fix, I'm going to nag them for review once it's the queue [17:36] but it's an important fix and it's a trivial change [17:36] ok [17:37] lool: so i got at least one confirm from QA that 91 is good [17:37] ogra_: popey: wanna check if things are bgetter than 90? [17:37] and then release? [17:37] plars: anything from dashboard that isnt copmlete etc.? [17:37] asac, so since we only have half the maliit fix in (it will still crash on shutdown) should i upload the second half so we have it in tomorrows image ? [17:37] i dont know [17:38] i think thats untested as well [17:38] so might be still only half [17:38] well do you want to see .crash files after reboots or not ? :) [17:38] it was tested by loic and me [17:38] asac: there are some things that could use rerunning, but we are about to get another image right? [17:38] i dont want to see them, but i dont want to see other stuff [17:38] (i'm still running it here) [17:38] ogra_: which landing entry missed the half? [17:38] and it was actually introduced by a merge failure [17:38] asac, maliit.conf has two "stop on" stanzas [17:39] * ogra_ looks for the MP [17:40] asac, trivial but very effective ... https://code.launchpad.net/~lool/session-manager-touch/drop-duplicate-stop/+merge/190373 [17:40] (upstart defaults to use the second one which means maliit stays up after unity is gone and falls over) [17:41] * ogra_ would like to have that in, will definitely prevent .crash files on the dashboard [17:42] could I get an account on https://jenkins.qa.ubuntu.com? fginther perhaps you have the power? [17:42] * cyphermox -> lunch [17:43] asac, the first half is landing 177 [17:43] balloons: not many people do - it would have to be an IS thing most likely [17:43] asac, it fixes the crashes on startup .. [17:43] but not the shutdown ones [17:43] balloons: if you need something removed or changed though, your best bet it to talk to retoaded or rfowler [17:43] ogra_, asac, fginther: what's the plan to land indicator fixes to the touch image after saucy is hard frozen tonight? (since those are shared sources between desktop and touch)? [17:44] ogra_, asac, fginther: do we plan to land updates as SRUs and build the touch image for -updates or something? [17:44] seb128, i suspect we need to obey to the release team for these [17:44] plars, yea I kind of figured I was getting into some big fish [17:44] seb128, no [17:44] :-( [17:44] so basically "we are screwed"? [17:44] asac: ogra_ want me to test 91? [17:45] popey, yes [17:45] kk [17:45] popey, are you running -proposed always or no? [17:45] balloons: i have more than one device, so yes. and no. [17:45] haha, well said [17:49] error: device not found [17:49] ERROR:phablet-flash:Command 'adb reboot bootloader' returned non-zero exit status 1 [17:49] ☹ [17:49] popey, don't overheat them [17:50] popey, where were you on that device when that happened? [17:50] popey, recovery? [17:50] ubuntu [17:50] cpu on the device isnt busy [17:51] adb is running [17:51] doesn't show up at all [17:52] popey, dmesg? [17:53] asac, bfiller_afk, sergiusens: well, with the patch at https://code.launchpad.net/~veebers/ubuntu-keyboard/update_autopilot_emulators/+merge/190319 the keyboard tests run ok for me but I get 5 failures (2 passes) [17:53] sergiusens: http://paste.ubuntu.com/6219033/ [17:53] bfiller_afk: is that consistent with what you would expect to see? [17:56] popey, from a quick look 90% of those messages are related to your cable/socket [17:57] i moved the phone to a different usb port to test [17:59] sergiusens: other phone shows up okay [17:59] * popey reboots [17:59] popey, can you reboot it? [17:59] ☻ [18:01] fine now === gatox_lunch is now known as gatox === bfiller_afk is now known as bfiller === ChickenCutlass_l is now known as ChickenCutlass [18:19] cyphermox: <3 [18:19] ogra_: \o/ [18:19] in #91 the OSK disappeared when typing the 2fa in U1 and now I can't get it back in any app [18:20] lool, go away !!! we'll send handcuffs to your wife so she can tie you up somewhere !! [18:20] Where do I sign up for that service? [18:20] plars: they all passed under SF, haven't tried them on mir [18:20] * ogra_ puts on a serious face and points to lool's bed [18:20] bfiller: let me try with sf [18:20] ogra_: yup [18:21] asac: ^^ [18:21] ogra_: just before I do, did you have a bug for dialer-app doesn't come to front when in a call? [18:21] popey, just work a few nights in a row and dont stop to work even tough you said you'd do ... [18:21] had good dinner and about to watch Borgen and have a good night of sleep :-) [18:21] popey, you are into that handcuff thing, eh ? [18:21] newp [18:22] just thinking I could put the handcuffs on the wife and then come back down here and hack [18:22] * ogra_ hugs lool ... enjoy [18:22] system settings wont respond now, all gone white [18:22] * popey kills it and restart [18:23] yeah, OSK aint coming back. [18:23] * popey reboots [18:25] bfiller, asac: yeah, they all pass with sf - let me retry with this latest build though [18:26] asac: also getting random characters appearing in fields.. had `` in one field and a øø in another [18:27] popey, yeah, thats there since a while [18:28] k [18:28] try and press the power button when in terminal app ... thats also funny [18:28] oh that's always been there [18:29] really ? i only noticed it recently [18:29] heh [18:29] popey: are those regressions? [18:29] or noise? [18:29] noise i think [18:30] asac, the two chars if you focus a text input field are relatively new [18:30] ogra_: relatively is a relative term [18:30] :) [18:30] but not critical [18:30] not consistent either [18:30] asac: since lool is rightly taking the night off....who can kick a ci/merge on unity8 for this https://code.launchpad.net/~nick-dedekind/indicator-network/simunlock.dialog/+merge/185810 [18:30] ogra_: can you be more specific since when we see this? [18:30] i think i noticed it for the first time about a week ago [18:30] asac: thats for sim pin unlock [18:30] for instance: before or after mi? [18:30] before [18:30] kgunn: that we pick up tomorrow [18:31] asac: ok. so do we just manual merge on our own ? [18:31] it dumps one or two chars in the text field ... you can just use backspace [18:31] not release critical ... just ugly [18:31] kgunn: no levave it as is at best so we can somehow asess and test before landing if needed [18:31] oh, but it might affect tests ... hmm [18:31] and kick off a CI job if possible [18:31] asac: we've already tested it on a device with a locked sim, its been doubly reviewed [18:32] right. not saying its not safe [18:32] kgunn: right. thats fine. but we need to really look closely at best and might use CI for the landing manually etc. [18:33] asac: when does this happen ? [18:33] kgunn: autolanding is running on this branch already! [18:33] tomorrow morning euro time we will look [18:33] is it? [18:33] then its autolanding :) [18:33] hehe [18:33] lool: woohoo [18:33] asac, some phones ahve SIMs in the lab ... we need to make sure they are PIN free [18:33] ogra_: so you didn't have a bug? [18:33] kgunn: well its because you dont have tests [18:33] ogra_: will just log one now thne [18:33] else that might interfere [18:33] lool, ! [18:33] so you are going unprotectd anyway, so it doesnt help to turn your jobs off [18:33] ....didn't match the list in the ginther mail [18:33] go to bed [18:34] kgunn: the reply included only those that are not getting enabled [18:34] yup, off now [18:34] thats just two handfull or so [18:34] if you are not in there its autolanding [18:34] Hello! I have one more dbus upload, which fixes the desktop autopilot tests, that I'd like to land [18:34] the debdiff is small (once you look past the quilt refresh churn): http://people.canonical.com/~tyhicks/fixes/dbus/source/dbus_1.6.12-0ubuntu10.debdiff [18:35] and I've done extensive manual tests on it: http://paste.ubuntu.com/6219140/ [18:35] I'm in the process of running a slew of automated tests against it with QRT's test-dbus.py [18:36] (added a landing for the sim unlock fix with link to mp. off now) [18:37] The lowlevel details are that it, along with a recent lxc upload by stgraber, fixes dbus-daemon inside of an LXC container breakage when /sys/kernel/security/apparmor/ is mounted inside the container [18:39] Back from shopping [18:40] lool: should I release HUD? [18:40] (after testing) [18:42] the automated tests in QRT's test-dbus.py passed, so it is safe to say that the code is ready for the dbus upload [18:46] lool: when you get a moment, can you give me an ack/nack on the dbus upload? [18:47] tyhicks, he isnt around (or shouldnt be) asac is the one giving acks/macks tonight [18:47] *nacks [18:47] thanks [18:48] asac: Can I get an ack/nack from you on the dbus upload mentioned above? [18:50] asac: ^ [18:50] I wrote the kernel code that populates /sys/kernel/security/apparmor/features/dbus/mask, jjohansen (AppArmor kernel maintainer) agreed that this is the right thing to do in dbus-daemon, and jdstrand will be sponsoring the upload [18:50] so all of the right people are involved in this change [18:54] fyi, I added a landing ask just now [18:55] jdstrand: thanks! test-dbus.py passed so you can update column G if you'd like [18:56] done [18:56] tyhicks: thanks! :) [19:01] tyhicks: tyhicks why do you have all those null op changes in the diff? [19:01] * asac guesses they didnt apply anymore [19:01] asac: I changed a patch that was 3 patches down in the series file [19:02] asac: I'm trying to keep all of the apparmor mediation patches in an upstreamable form [19:02] asac: so changing that patch meant refreshing the two on top of it [19:02] sorry for the ugliness, but I think it is for the best [19:03] yeah dont worry, just for manual review its better without refresh [19:03] agree, but then the two patches above had line offset warnings [19:03] ogra_: so this goes into the android container right? [19:03] dbus ? [19:03] no [19:04] err [19:04] its talking about lxc container in the comment [19:04] so wonder if it might impact our side there [19:04] oh [19:04] ogra_: e.g. do we start an lxc container? [19:04] :) [19:04] yes, but not using the standard lxc startup stuff [19:05] we use the lxc-start command with a preconfigured setup ... [19:05] tyhicks, jdstrand, can that affect our container ? did someone test on a phone [19:05] note this change only affects dbus-daemon [19:05] we have dbus-daemon on the phone, rigth? [19:06] "dbus-daemon, running inside of an LXC container" [19:06] yes, but you don't run dbus-daemon in the container, do you? [19:06] wont affect us [19:06] * asac checked and its running a few times [19:06] well, just think out loud. ultiamtely checking what impact it might have or not have [19:06] not in the container [19:06] do we have this mask file on our phones? [19:06] right, the lxc bit should be fine [19:07] the change should affect the phone the same as the desktop [19:07] ls -la /sys/kernel/security/apparmor/features/dbus/mask [19:07] -r--r--r-- 1 root root 0 Jan 1 1970 /sys/kernel/security/apparmor/features/dbus/mask [19:08] so instead of stat()'ing /sys/kernel/security/apparmor/features/dbus, we're open()'ing /sys/kernel/security/apparmor/features/dbus/mask [19:08] yes i am seeing that [19:08] the file seem to exist at least :) [19:08] apparmor doesn't mediate stat(), but it mediates open() [19:08] /sys/kernel/security/apparmor/features/dbus/mask is part of the apparmor kernel backports in all our touch kernels [19:09] dbus-daemon doesn't run confined on the phone or on the desktop, so it will still be able to open() the mask file without any problems [19:09] right, looks safe to me overall [19:09] yeah. i feel its as safe as it can be from reading the code, but as everything we want to see it run on the phone :) [19:10] if we pump it in now and tomorrow we have a busted image we loose like 12 hours or so :) [19:10] which is probably like 20% of the time left :) [19:10] asac, and it will be hard to roll back [19:10] dbus is a package that falls under final freeze [19:11] it is [19:11] this issue only came to our attention this morning [19:11] but the change is stat to open [19:11] yeah. if we knew earlier we could have worked in the plan [19:11] I started an armhf local compile, but I doubt it'll finish in time [19:12] so if we do this i also have to get kenvandine's stuff in [19:12] tyhicks: you could toss it in ubuntu-security-proposed [19:12] he has the very same argument [19:12] asac: if testing HUD proves OK, can I release? [19:12] jdstrand: will that build an armhf deb? [19:12] please now without notice and testing because we have final freeze in a couple minutes [19:13] sil2100: HUD is in the plan right? [19:13] tyhicks: I hope so-- it is one of our native ppas :) [19:13] ok [19:13] tyhicks: "yes" [19:13] ogra_: what is the latest image we produced? [19:13] asac: it's mentioned in the Summary e-mail, yes [19:13] asac, 92 [19:13] ogra_: btw, #92 still has malit crashes [19:13] all the time [19:13] http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/92:20131010.2:20131010/4662/ [19:13] asac, i told you [19:14] ogra_: didnt this image pick up the powerd package? [19:14] it crashes on shutdown without the fix [19:14] powerd from 0.13+13.10.20131009-0ubuntu1 to 0.13+13.10.20131010-0ubuntu1 [19:14] http://people.canonical.com/~ogra/touch-image-stats/20131010.2.changes [19:14] yeah [19:14] its in [19:14] plars: can you retry unity8? [19:14] plars: that should have worked now with the powerd fixes by ricmm [19:15] ricmm: ^^ ... this is the log https://jenkins.qa.ubuntu.com/job/saucy-touch_mir-mako-smoke-unity8-autopilot/15/console [19:15] e.g. after powerd we still see not running unity test [19:15] plars: doanac: lets check on that... also check how we really run this unity8 testsuite [19:15] e.g. which commands to run [19:17] kenvandine: where is your debdiff? [19:18] kenvandine: you could have linked the important things like code/merges/etc. in the landing ask :) [19:19] * ricmm installs unity8-autopilot [19:19] let it fly [19:19] ... [19:20] lol [19:20] asac: I am, the job timed out [19:21] asac, it's 3 packages, all in trunk already [19:23] sergiusens: ok, I approved https://code.launchpad.net/~veebers/ubuntu-keyboard/update_autopilot_emulators/+merge/190319 and got the autolanding job to pass [19:23] asac: ^ sergiusens: I see failures with the tests under mir that I don't see with SF, but the point of the patch was to make the tests run at all - which it does [19:25] give me a concrete failure [19:25] I dont want to read through a thousand line log [19:25] with results I dont necesarily understand [19:25] plars: which one is the weirdest in your opinion? [19:27] ogra_, landing please https://code.launchpad.net/~sergiusens/ubuntu-seeds/share-app--/+merge/190452 [19:27] depends on old hud that needs to be removed from the images [19:27] ogra_, description has more ;-) [19:27] asac: any objections for getting HUD in ;) ? [19:28] It doesn't seem to crash for now! [19:28] ricmm: well, I'm actually somewhat hopeful that the problems I'm seeing are a result of the maliit crashes on mir [19:28] sergiusens, i think bill added the same MP from ted a few mins ago [19:28] ricmm: I see a lot of stuff like this: [19:28] 17:50:27.031 INFO testcase:554 - Killing process 2565 [19:28] 17:50:27.033 INFO testcase:557 - Appears process has already exited. [19:28] why would autopilot attempt to kill maliit? [19:28] ogra_, that's good [19:28] ricmm, it stops unity [19:29] ricmm: I don't know that it's trying to kill maliit, and I haven't looked at the tests [19:29] ogra_: right [19:29] asac, can we please upload that maliit fix :P [19:29] ogra_, ricmm: I see a unity8 crash with it too [19:29] then your unity tests will work [19:29] plars: what line? [19:29] ah this log has no lines of course [19:29] plars, can you hack the install ? [19:30] ricmm: this was run by hand on my local device [19:30] ogra_: if you can tell me what you want? [19:30] tell me the test then [19:30] plars, remove the second "stop on" line from the maliit upstart job [19:30] asac: publishing HUD \o/ [19:30] plars, and then try again [19:30] i bet it will work [19:30] ogra_: ok, just to make sure we are talking about the same thing, are you talking about the unity8 tests or the ubuntu-keyboard tests? [19:30] or both? [19:31] plars, i talk about maliit crashing if it gets stopped ... which i.e. is triggered by stopping unity [19:31] ricmm: the mp for the version of the keyboard tests I'm looking at is at https://code.launchpad.net/~veebers/ubuntu-keyboard/update_autopilot_emulators/+merge/190319 [19:31] ogra_: ok, so we should see it in either one [19:31] right [19:32] the job is broken, so it doesnt actually stop [19:32] but unity does [19:32] so you pull the floor out underneath [19:32] unity exiting will have the result of maliit exiting if the keyboard has been invoked once [19:32] as when Mir exits it asks all of its clients to clean up and die gracefully [19:33] ricmm, right, but maliit has a "stop on session-end" ... you end up in an incosistent state, upstart thinks it died and tries to respawn it [19:34] ogra_, I tested earlier with stop on stopped unity8 [19:34] right [19:34] with the "stop on stopping unity8" it will just properly shut down [19:34] so im not so sure about that [19:34] is autopilot bringing up unity8 with upstart? [19:34] well, i have no crash files here [19:35] which i do as soon as i re-add that line [19:35] ogra_, ricmm it's not most likely [19:35] right but thats because you also dont have any maliitrunning [19:35] right after a shutdown [19:35] unity8 will not invoke maliit if the keyboard is needed [19:35] but it will die every time that unity8 goes, if its alive [19:35] that means that tests that restart unity8 and need keyboard would needto somehow make sure that maliit is running [19:35] ogra_, ricmm why not have the start stop logic for maliit in the unity8 test setup? [19:35] I saw other apps test like that [19:35] exactly [19:35] ricmm, nontheless the upstart job is broken :) [19:36] ogra_: not saying it isnt [19:36] ogra_: where is that job located? [19:36] /usr/share/upstart/sessions/ [19:36] what im saying is that, even with the fix, ocne maliit dies once it wont come back [19:36] thanks [19:36] so when unity8 is killed, maliit will not return [19:36] unless manually invoked [19:36] ricmm, but it doesnt die, it gets propely stopped before unity [19:36] and it will get started again with unity [19:37] that'd be true if unity8 autopilot's invokations were happening via upstart [19:37] but its just running the process manually and then killing by hand [19:37] err [19:37] ogra_: So I'm removing the stop on session-end and leaving the stop on stopping unity8 in place, correct? [19:37] no? [19:37] plars, right [19:37] ricmm, that would be most evil ... [19:38] ricmm, and the log looks like it is at least stopped via upstart [19:38] ogra_: the first stop yes [19:38] ogra_: but the runs actually bring it up manually [19:38] ugh [19:38] now thats super evil [19:38] evilpilot [19:39] and i doubt we can solve this without moving autopilot to use upstart [19:39] asac, ``` [19:40] or adding a startup of maliit-server for tests that need keyboard [19:40] its all crap [19:41] yeah [19:41] Hi guy [19:41] s [19:42] * thomi sees a lot of 'autopilot' pings from this channel, and wonders what the beef is [19:44] thomi, you probably need to start unity8 autopilot tests with upstart for the keyboard stuff to function correctly [19:45] thomi, well, veebers has to :-) [19:45] sergiusens: OK, that should be reasonably easy to hack up. veebers should be online in 15 minutes or so [19:45] right :) [19:45] I cna help him though [19:45] cool [19:45] thomi, would be similar to the click stuff we have [19:45] thomi, but using a legacy APP_ID [19:46] sorry no [19:46] ignore that [19:46] sergiusens: oh? not just stop/start 'unity8'? [19:46] that's for legacy apps :-) [19:46] haha [19:46] thomi, yeah, start/stop unity8 [19:46] right [19:46] He'll have to hack it into the test suite, and we can remove it once we start using autopilot 1.4, where we can add this support properly [19:48] asac, ogra_: Ok, stgraber has tested on touch: http://paste.ubuntu.com/6219448/ [19:49] jdstrand and/or I can confirm his results in a few minutes [19:49] plars: are we retrying the yellow ones :)? [19:49] plars: maybe we can hit 100% with crashes :)? [19:50] asac: right now I'm trying something for ogra_ will go back to that in a moment [19:50] plars, well, if the tests actually not use upstart to start and stop the shell these tests are moot [19:50] ogra_: I had some issue with that change - the tests wouldn't even start, but without they behave (modulo the known failures) [19:50] asac: coming? [19:51] plars, asac, the issue is that the tests just start up unity and not use the upstart start/stop commmands so maliit never gets started along [19:52] didrocks: yep [19:58] asac, gerry's work around in unity-mir did solve the signon-ui problem, i've proposed a branch for unity-mir [19:58] asac, so it won't require changes in the signon stack to fix [19:59] kenvandine: did he add the exception I said? [19:59] yeah [19:59] sorry that I forgot [19:59] https://code.launchpad.net/~ken-vandine/unity-mir/lp1235337/+merge/190459 [19:59] ricmm, ^^ [20:00] kenvandine: did you try starting apps with that? [20:00] as in start signon-ui and then start another app right away [20:00] yup [20:00] is signon-ui actually trying to create a window? [20:00] even if it isnt displaying anything [20:00] no [20:00] only on the desktop [20:01] not on the phoen [20:01] phone [20:01] the fenceNext troubles me there if its not bringing up anything [20:01] lemme do some testing before happroving [20:01] plars: yeah would be good if we want mir team to look at the regressions still left [20:01] or if they are flaki [20:01] some flakiness is i think normal with those crashes [20:02] asac: well the unity8 ones are not going to get any better than what you see there, but the others are retrying that have 1 or 2 failed tests [20:02] ricmm, thx [20:03] plars: yeah the rest i mean [20:03] unity8 is still failing [20:03] asac, yes, see backlog [20:03] the tests will need fixing [20:04] (and the maliit upstart job needs fixing, else this wont work either) [20:05] can i get someone to add something to a row in the landing asks page? [20:10] asac: I can confirm stgraber's results [20:11] asac: tyhicks was able to as well [20:11] didrocks: thanks [20:11] didrocks: wait] [20:11] unity8 came up, dbus is running, can launch apps, etc [20:11] asac: yw ;) thanks to you [20:11] asac: yeah? [20:11] didrocks: so jdstrand wants to land dbus for desktop [20:11] didrocks: and wants to do so before the final freeze [20:11] I guess we need to assess the impact on all AP tests for phone [20:11] didrocks: the code is supe minimal [20:11] is that a risky change? [20:12] didrocks: right. from reading code its really zero :) [20:12] do you have a quick diff? [20:12] we change a stat() to an open() [20:12] didrocks, it has been tested by various people and is a smal fix [20:12] didrocks: jdstrand just said that they tested it? [20:12] jdstrand: what can stgraber confirm? [20:12] can you restate exactly what he did? [20:12] ok, as long as unity8 AP tests are running (maybe just try an app tests as well?), I'm fine with this [20:12] didrocks: http://people.canonical.com/~tyhicks/fixes/dbus/source/dbus_1.6.12-0ubuntu10.debdiff [20:12] asac: tyhicks said that stgraber confirmed dbus works. tyhicks and I independently confirmed that [20:12] didrocks: a few null op diffs and then a small one [20:12] * didrocks looks [20:12] didrocks: i checked that the mask file is there ... it has permissions: [20:12] here's what all three of us did: http://paste.ubuntu.com/6219415/ [20:13] * jdstrand did a bit more than that [20:13] (but did do that) [20:13] didrocks: 21:07 < asac> -r--r--r-- 1 root root 0 Jan 1 1970 /sys/kernel/security/apparmor/features/dbus/mask [20:13] didrocks: here is what tyhicks and jdstrand and stgrabver did to ensure its safe: http://paste.ubuntu.com/6219415/ [20:13] (havent read) [20:13] note that there are some typos in the commands, unfortunately, but it shows that dbus-daemon comes up successfully and detects that apparmor is available [20:14] tyhicks: you gusy seems to feel resistant of testing it on the phone :)? [20:14] its not clear if you tested it on the phone with our touch image [20:14] yeah, and the code change seems not really risky at all ;) [20:14] stgraber's paste doesn't have typos: http://paste.ubuntu.com/6219448/ [20:14] but [20:14] I would love just that the AP tests are run on phone for unity8 and one app [20:14] just to ensure we can still start/connect through AP [20:14] didrocks your call. i believe we could let it in with some confirm that our touch image is not busted [20:14] asac: grouper isn't enough? [20:14] asac: not resistant at all - we were just short on time [20:15] jdstrand: grouper isnt existing for us :) [20:15] tyhicks: right we are also short in time. if this busts we loose 18 hours [20:15] which is 20% of the time left :) [20:15] * jdstrand puts it on his avengers device [20:15] risk is low, but stake is still high [20:15] asac: notes was the first to run, just fininshed and it got worse by one test [20:15] plars: retry? [20:15] is it flaki? [20:15] note, grouper mir with unity in devel-proposed [20:15] worked [20:15] plars: given how maguro loks i would assume its super lfaki [20:16] asac: I think grouper for just starting and ensuring dbus connect is fine enough [20:16] asac: yeah, I'm talking about mako, and I've already kicked it off to retry [20:16] jdstrand: but please do run the AP tests [20:16] asac: maguro is much farther away [20:16] as autopilot is connecting through dbus to pilot the app + unity8 [20:17] didrocks: all of them? it passed all tests on desktop. we had far less barriers for the changes a couple days ago. we just change a stat to an open [20:17] didrocks: this bug is blocking autopilot tests on desktop [20:18] jdstrand: well, autopilot doesn't connect the same way on phone and desktop (for unity8) as from what I got, apparmor is not setup the same way [20:18] is testing dbus on the phone or grouper is such an issue? [20:18] we did test on grouper [20:18] (I think running both unity8 + let's say notes-app tests is taking 10 minutes in total) [20:18] testing on mako is an issue [20:18] running AP tests? [20:18] yeah, so running the AP tests on grouper is enough IMHO for this one [20:19] * ogra_ adds line 182 to landing asks and goes for dinner [20:19] does this sound so unreasonable as tests? [20:20] didrocks: are AP tests know to work on unity8? I just did a left to right swipe and the display is hosed [20:20] sorry, on grouper [20:20] jdstrand: I don't know on grouper, I would say a before/after would help [20:20] test* [20:21] didrocks: sorry, there is another friend asking for landing something that was not known before :) [20:21] tyhicks: I can't do it sorry. I have to go [20:21] see -touch [20:21] jdstrand: understood [20:21] tyhicks: I guess the desktop AP tests will continue to fail until I can test on mako [20:21] is this the way to run all of th AP tests: https://wiki.ubuntu.com/Touch/Testing/Autopilot [20:21] ? [20:22] tyhicks: you can try on grouper-- but I'm guessing it isn't going to work well [20:22] tyhicks: no, that's pretty outdated information [20:22] https://wiki.ubuntu.com/Touch/Testing I think is what you need [20:22] tyhicks: if you get it done in time, perhaps someone in this channel can sponsor for you [20:23] * jdstrand -> appt [20:23] ok [20:23] slangasek: what do you think about staging nm before the freeze in proposed but wait till we are able to test it tomorrow morning? [20:23] tyhicks: sorry I couldn't help more [20:23] cyphermox: didrocks: ^^ [20:23] thomi: can you confirm that https://wiki.ubuntu.com/Touch/Testing is what I need [20:23] slangasek: release team hat on i am asking :) [20:23] jdstrand: no problem [20:23] asac: I think that delays us being able to master images for the release [20:23] and am therefore against it [20:24] slangasek: so it would then have to wait for opportunistic landing for a new image? [20:24] tyhicks: yes, that looks *much* better :) [20:24] thanks [20:24] asac: effectively, yes [20:24] this is a fix specific for touch so if you don't think you have time to further test it and all, I just won't upload [20:24] slangasek: likelyhood of getting a respin by monday is > 80% :)? [20:25] didrocks: what happens if this sits in proposed? does this block others from entering? [20:25] asac: if there is no ABI break, it shouldn't [20:25] I think staging in proposed is the way to go: at worse, if we don't do any respin (hem hem), this will go in -updates [20:26] slangasek: so how likely is it that you might defer building master image later than tomorrow morning from current look? [20:26] slangasek: right. or maybe its a candidate for -updates? [20:26] didrocks: right. [20:26] asac: we would not defer it at all, we need to get candidate images out to start the testing [20:26] slangasek: right. just thought maybe you fight another blocker still [20:27] not asking if you defer it for this [20:27] didrocks: i agree. your call. if you know that we can still land other stuff then its good [20:27] asac: well, we might go in knowing we would need a respin after, but we would still produce images and start testing them [20:27] didrocks: cjwatson stated that we can even do things in -updates and pull it into the images [20:28] but in this case, no, I'm not aware of any late-landing fixes that we're expecting [20:28] jdstrand: line 217 in landing ask [20:28] didrocks: asac: slangasek: got a working gst/hybris for video thumbnail and video playback in the browser, can we still push that today? [20:28] right. but if you know you need a respin, you could process it right after the image was build :) [20:28] or tomorrow only? don't know the current state of things [20:28] hehe [20:28] slangasek: fyi....merged & ask sheet....https://code.launchpad.net/~vorlon/unity8/lp.1235649/+merge/190443 [20:28] didrocks: so yeah. lets do that. i think we could even wait with uploading until we have it fully tested [20:28] kgunn: cheers [20:28] asac: agreed [20:29] didrocks: ok, so noo upload. we will check and upload to proposed for opportunistic landing tomrrow after the first image [20:29] and then route it to -updates in case no new image is coming up [20:29] right [20:30] rsalveti: same, I guess, stage that to proposed, and we'll try to pick it up tomorrow [20:30] rsalveti: just ensure with the release team you can pin it [20:30] rsalveti: we want another mir-fix-only image and then process the rest [20:30] gst-plugins-is fine, just need a change in hybris and one android respin [20:31] not related with mir this one [20:31] didrocks: did we freeze the proposed transition as well? [20:32] rsalveti: I would love that, let me add a comment in the landing spreadsheet [20:32] rsalveti: isnt this exclusively for phone? [20:32] asac: it is [20:32] if so we can take the same route as we do with network-manager [20:32] e.g. we dont need to rush it imo [20:32] anyway. didrocks knows :) [20:33] don't need to rush, but getting that today would be fine unless you want me to block it [20:33] rsalveti: please, just ensure it will be staged in proposed, we'll unblock it tomorrow [20:33] rsalveti: I guess slangasek can hint britney for it [20:33] rsalveti: we have a good plan. that will get us a full recovered MIR image tomorrow. after that everyone will land. if proposed staging is good and doesnt block landing everything i agree with didrocks [20:33] that you cdan stage it [20:33] if the release team is aware to not take it :) [20:33] what am I hinting? [20:34] would need to stage libhybris, gst-plugins-bad1.0 and android [20:34] rsalveti: you handle the android rebuild as well? [20:34] didrocks: yes [20:34] fginther, according to your list music should be unblocked, was that a typo? if not https://code.launchpad.net/~sergiusens/music-app/manifest/+merge/190250 is stuck [20:35] rsalveti: thanks ;) [20:35] slangasek: do I need to ask the release team to block the proposed transition of those packages specifically? [20:35] or is everything frozen already? [20:36] rsalveti: things are not frozen in proposed-migration currently, only at archive accept [20:36] sergiusens, yes, that was a typo, it's disabled [20:36] ack thanks [20:36] rsalveti: updated req 148 FYI [20:36] slangasek: right, but can I request to freeze the proposed migration for these packages specifically? [20:37] as it seems people want stuff to be in proposed til tomorrow [20:37] rsalveti: yes [20:38] slangasek: do you prefer me doing it? [20:38] (the hint) [20:38] I'm already doing it ;) [20:38] excellent! [20:38] thanks slangasek, rsalveti! [20:38] slangasek: great, thanks [20:38] rsalveti: we're getting it in just after the first image tomorrow morning [20:38] (94 if all goes according to the plan) [20:39] didrocks, can we get a new settings on the image? ;-) [20:39] didrocks: great, thanks [20:39] seb128: tomorrow morning, same image than rsalveti :) [20:39] didrocks, thanks [20:39] well at least this one is universe [20:40] so the release team is not in the way [20:40] yeah, let's hope that tomorrow's Mir image will be as good as we hope [20:40] and we can try to catch up [20:41] didrocks, asac: we have fixes that depends on indicator updates, do we plan to build images with SRU updates if needed? [20:42] seb128: I'm afraid that we need to discuss that one seriously tomorrow now that the archive is frozen [20:42] seb128: before the release we might divert uplaods to -updates, but the current thinking is that we focus on T [20:42] or we can move stuff to -updates until then [20:42] heh [20:42] asac, which means releasing 1.0 with those bugs? [20:42] seb128: dont worry. things will work fine [20:42] sounds like the saner plan [20:43] seb128: I guess we'll have image building with -updates for touch [20:43] asac, well, I don't care much for myself, but we have e.g the "sound keys trigger weird notification" bug in that category [20:43] seb128: no, we can build images with things in -updates - if it comes to it - i was told [20:43] seb128: so we can land for a bit in the archive and then in -updates and release a great story next wednesday [20:43] thats my - potentially overly optimistic - hope :) [20:44] * didrocks tries to be overly optimistic as well for once :p [20:44] anyway, I think it's high time [20:44] * didrocks waves good evening :) [20:44] asac, so it means SRUs to get in -updates? [20:44] didrocks, night [20:44] didrocks: enjoy and and sleep like a rock :) [20:44] ttyt [20:44] night didrocks [20:44] asac: thanks will try! ;) [20:44] seb128: no. you just upload to saucy [20:44] good night seb128, cyphermox ;) [20:44] seb128: through CI ... and things will go into -updates in case we dont want to respin a desktop image [20:45] (and our touch image picks it up from there before the release) [20:45] asac, how are things going to go to -updates? that's for verified SRUs [20:45] seb128: its before the release, so its not a SRU [20:46] well, it kind of is. so yeah. you need to do proper testing on desktop [20:46] asac, what happens to those -updates version at release time then? do we wipe them out? or are desktop users getting them as SRUs? [20:46] seb128: yes. desktop useres get them on day one [20:46] asac: stgraber reports that the unity8 AP tests pass on mako with the updated dbus package [20:46] asac, I doubt slangasek & co are going to move uploads to saucy-updates without following the SRU rules [20:47] seb128: right. however, i doubt we want to land anything at this point that doesnt meet the SRU quality standards [20:47] seb128: i think you just need to take care that you dont change the desktop UI and that you have rock solid quality on your changes [20:47] which is beneficial at this point anyway [20:47] right [20:47] tyhicks: unity8 AP pass? [20:48] tyhicks: that would good news, because they officially fail :) [20:48] lol [20:48] I'll relay that to stgraber [20:49] yep. asked him [20:49] on -touch [20:50] asac: any objections for landing platform-api? [20:51] sil2100: is it in plan? did we validate that it doesnt regress our nice AP ratio here: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/92:20131010.2:20131010/4662/ ? [20:51] sil2100: if both are true, go for it [20:51] image 93 is what it needs to be scheduled for [20:53] tyhicks: can you just take a phone and test it against latest image please? [20:53] i know its tempting to just test it against something undefined, but if we do that its not even worth the time there as we still wont know [20:53] asac: I don't have the hardware [20:53] asac: no idea if we validated that, but I know didrocks poked me about getting that in, as it fixes the maliit-server issues (probably) [20:55] sil2100: right. we need to test [20:56] thats was the whole idea of poking i am 99.9% sure [20:56] tyhicks: ack [20:56] asac: should I run all the AP tests with it? [20:56] thats the problem with having busted tests. i would say unity8, but now we dont know [20:57] sil2100: run webbrowser :) [20:57] lol [20:57] sil2100: what is this trying to fix again>? [20:57] asac: hehe, but how does it look on a vanilla system? All pass? Smoketesting say so, but is that accurate? [20:57] sil2100: mallitt crashes? just run a few APs that have maliit crashes [20:57] sil2100: pick 2 here: http://reports.qa.ubuntu.com/smokeng/saucy/touch_mir/mako/92:20131010.2:20131010/4662/ [20:58] asac: https://launchpad.net/bugs/1233988 [20:58] Ubuntu bug 1233988 in platform-api (Ubuntu) "With Mir enabled: maliit-server crashed with SIGABRT in __gnu_cxx::__verbose_terminate_handler(), thrown from mir::client::DisplayConfiguration::copy_to_client()" [High,In progress] [20:58] sil2100: all pass means that you can get it to all pass [20:58] sil2100: as soon as one fails we cant say that its impossible to get to all pass [20:58] thats the dilemma and why we can only shoot for 100% [20:58] sil2100: right. so pick a few crashes we see here [20:58] sil2100: and confirm that they go away [20:59] sil2100: so messaging-app is pretty quick test [20:59] seems to produce crashes reliably [20:59] sil2100: share app as well [20:59] sil2100: and if you are strong run webbrowser app [21:39] asac: so, we've tested the code thoroughly, but you want us to run the unity8 tests on mako, but the unity8 tests are known to fail [21:39] asac: it doesn't matter now anyway, but this is pretty frustrating esp when we tried to do the right thing and got blocked [21:40] stgraber has ran the tests and reported on the results in #ubuntu-release [21:40] right, but wasn't he told he had to update to the latest, which is known to fail? [21:42] asac: apparently the tests passed [21:42] so we've met the criteria "Please run AP tests on phone for unity8 + one app to ensure that the tests still starts and communicate with the app", but the status in the landing plan is TODO, so I don't know if we should upload or not [21:43] asac: passed on surfaceflinger (see #ubuntu-release for details) [22:02] I might add that the risk assessment is not medium [22:03] * jdstrand changes [22:09] jdstrand: i thought we already said tis ok etc. [22:09] and yu have a landing slot [22:09] if you did that [22:09] if didrocks said something else then let me know [22:09] didrocks gave the landing slot, but there was a TODO [22:09] I don't know what that meant. I updated the page [22:09] (showing the testing involved) [22:10] I guess I'll upload-- it'll be blocked in unapproved and people can ping me about it there [22:10] * jdstrand will mention in #ubuntu-release [23:00] gst-plugins-bad integration is blocked by bug 1238312 [23:00] bug 1238312 in Mir "Segfault when closing apps that link against ubuntu_application_api_mirclient" [Critical,Confirmed] https://launchpad.net/bugs/1238312 [23:00] mir specific as well [23:02] slangasek: smells like toolchain related ^ [23:02] only with armhf, and when built natively [23:02] ! [23:04] rsalveti: could it be an issue with uninitialized memory that happens to wind up initialized differently by the compiler for cross vs. native? [23:05] that's what I was thinking as well [23:06] well, if it's uninitialized memory, then it's not actually a toolchain bug [23:06] because something should've initialized it :) [23:06] is libubuntu_application_api_mirclient in the mir source package? [23:06] nops [23:06] that's part of the platform-api [23:07] but links against mirclient [23:07] right, it seems affected by the toolchain, but it might indeed be an issue in the mir code itself [23:14] rsalveti: do you know where the exit handler is set from? [23:15] nops [23:16] you're not in #ubuntu-mir btw [23:17] I am in far too many channels already [23:17] yeah :-)