[01:33] G'evening [01:33] Is anyone around by any chance? I can't get unity to play nice with the binary ATI drivers [02:26] F^4: what's wrong ? [02:28] smspillaz, The menu bar and dock are missing [02:30] F^4: can you click on them as though they were there ? [02:30] I didn't even try tbh. I just installed KDE and I've been using that. [02:30] Let me switch users and see. === hggdh_ is now known as hggdh [06:42] good morning [06:42] good morning didrocks and welcome back! [06:44] thanks Mirv! happy new year ;) [06:46] Mirv: Happy new year. Sorry to ruin it but I found: https://bugs.launchpad.net/bugs/1095001 [06:46] Launchpad bug 1095001 in compiz (Ubuntu) "[regression] compiz spends 31% of its CPU time in regexec()" [High,Fix committed] [06:47] indeed [06:48] duflu: yeah, I noticed your branch [06:48] Just so you know :( [06:48] the good news is that the previous SRU:s are all in, so we can do small ones [06:48] Mirv: did you see that unity is FTBFS on the staging ppa (I did quite a lot during my holidays btw)? [06:49] didrocks: not yet. I tried to stay away from Ubuntu during holidays but just had to eg. verify SRU:s etc :) [06:50] Mirv: that's a good resolution! :) you are just back today? ;) [06:50] didrocks: I did came back on Wednesday already [06:50] ah ok :) [06:52] it seems satoris' "Fix build settings" did the contrary [06:53] how it got into staging should be the interesting part [06:54] Mirv: yeah [06:54] Mirv: can you ask mmrazik once he's around? [06:54] sure [06:55] Mirv: also, let's wait for Satoris and sil2100 once they are around to eventually revert their change [06:56] yes [06:56] Mirv: hum, no, it was already FTBFS for some days: https://launchpad.net/~unity-team/+archive/staging/+packages?field.name_filter=unity&field.status_filter=superseded&field.series_filter=raring [06:56] rev 3002 on 2012-12-19 [06:57] that's the day I started to saw it as well, weird that nobody noticed :/ [06:59] but in staging rev 3008 built fine? https://launchpad.net/~unity-team/+archive/staging/+builds?build_text=unity&build_state=all [07:01] and in daily two bzr3008:s, one of which built and the other use_pch enabled and failed [07:01] Mirv: yeah, I'm looking at that… I don't understand why 2 builds by rev [07:02] even more for 3002 [07:02] actually five for 3008 as well, I wonder if they are doing some by-hand use_pch testing [07:02] maybe the situation is in control :) [07:02] Mirv: uploader is ps-jenkins though [07:03] we'll find out [07:03] Mirv: please keep me posted :) [07:03] thanks! [07:03] yep [09:57] Mirv: 2 things, thanks for the fw first :) (compiz was one of the first thing I checked before reading my emails TBH ;)) [09:57] Mirv: so, you are tracking with duflu the mesa 9 issue on precise with x-updates? [09:59] Mirv: I'll ask the second right now, so you did check with mmrazik about the multiple uploads in the ppa with the same version? (like 5 uploads in the staging ppa) [10:00] didrocks: unity-team staging ppa/ [10:00] ? [10:00] mmrazik: yep [10:01] mmrazik: look at https://launchpad.net/~unity-team/+archive/staging/+packages?field.name_filter=unity&field.status_filter=superseded&field.series_filter=raring [10:01] mmrazik: you can see that there is multiple uploads (like 5) with the same bzr [10:01] do you know about it? [10:01] didrocks: but the version is different, isn't it? [10:02] didrocks: its probably because something libunity depends on was uploaded [10:02] so a new dput was triggered to make sure libunity gets rebuild [10:02] (on a first glance; otherwise fginther will need to have a look) [10:05] mmrazik: ah, and what the version in the end is about? [10:05] seems pretty random, but there is maybe a logic :) [10:05] didrocks: the last number (after raring) is I believe a jenkins job # or something like that. [10:05] ah ok [10:05] so its indeed kind of random but unique and monotomic [10:05] makes sense, thanks :) [10:10] didrocks: do you think it should be tracked, the mesa 9 (only)? so, not yet, since it's not what the .2 release is about [10:10] duflu is of course aware of the issue, though [10:11] Mirv: yeah, even if it won't realistically be in .2. It should be tracked I guess [10:11] if we can confirm that the LTS-Q stack work, we can then try to narrow down why only mesa upgrade isn't enough [10:11] yep :) [10:11] ok, consider it being tracked [10:11] thanks Mirv :) [10:20] sil2100: hey! checking up now that I'm back from holidays: did you get any progress during the week you were there and I wasn't about the indicator autopilot tests? [10:20] IIRC, we wanted the number to get down to 0 :) [10:34] sil2100: sorry, didn't receive a reponse if any, system lock :/ [10:40] didrocks, hey there, nice to see you back. Enjoyed your Christmas time? [10:40] didrocks: ok, so - not completely sure what's the current number of failures is, since there has been no new unity daily package since a while ;p But locally I only had 4 failures, all of which were real regressions [10:41] davidcalle: hey! yeah, was great, even if not as restful as expected (mindwise, hard to disconnect ;)). Yourself? [10:41] From what I see the regression causing those 4 failures is not yet fixed though, since it's a low priority bug... [10:42] sil2100: did you open a bug for them? Maybe we can bump the list [10:42] them* [10:42] didrocks: https://bugs.launchpad.net/unity/+bug/1089482 [10:42] Launchpad bug 1089482 in Unity "Unity HUD uses wrong icon when switching from the Dash to HUD" [Undecided,New] [10:42] didrocks: I also had to re-visit the 'empty-menu-entries' issue with larsu, but after a while we finally sorted it out [10:42] didrocks, same for me ;) [10:43] Since in one moment I thought it was not fixed again, but then in the end it was fixed, but gedit was simply broken ;p [10:44] So even with larsu's fix, I had to hack-up the autopilot test to work properly with applications like gedit, which add seperators to the main menu bar (huh) [10:44] sil2100: the 4 failures are because of that? [10:44] sil2100: oh ok ;) [10:44] No, the 4 failures are because of #1089482 [10:44] :) [10:44] sil2100: I think it's worth fixing it then, right? [10:45] I think so - some days before the holidays I pinged Trevinho about this bug, but I think he was busy with other things [10:45] I could also ultimately look into it myself [10:45] sil2100: oh, that would be cool! and then, from your side, it's the only one that needs to be fixed and then, we live in a wonderful world, isn't it? ;) [10:46] didrocks: let's hope so! The next daily unity package will reveal if all is as wonderful as my test-box says ;p [10:46] sil2100: heh, indeed :) [10:47] sil2100: let's wait for unity to have the merge we are waiting on and then, I'll try a build in a ppa [10:47] sil2100: and then again, we can run the tests :) [10:47] (with the ppa content) [10:49] Would be good, since I even added some additional safety waits to some of the risky Autopilot tests, since some failures were due to e.g. a window not opening fast enough [10:49] excellent! :) [10:49] thanks again for working on that, it's a hard work that was needed :) [10:51] * sil2100 just noticed one old MR not merged in due to some jenkins failure [10:52] I wonder if it'll still be mergable [10:53] try it! :-) [11:35] hey andyrock, how are you? [11:35] didrocks, not bad you? [11:35] I'm good thanks! [11:35] andyrock: looking at https://code.launchpad.net/~3v1n0/unity/remove-bamf-indicators/+merge/140519 [11:35] andyrock: so unity won't build/work with older bamf? [11:36] or is it the contrary rather? [11:36] those functions are not used at all in the current code, isn't it? [11:36] in the current code they are not used [11:36] in 6.0 trunk yes [11:36] excellent, just wanted to confirm! Thanks andyrock :) [11:37] so current code will not work with older bamf [11:37] so [11:37] we need to bump the build-dep in debian/control, right? [11:37] and old unity will not work with new bamf ;) [11:37] to ensure we build with newer bamf? [11:37] didrocks, yup i think there was a discussion about that between Trevinho and mterry [11:38] I don't see the build-dep bump MR though? [11:39] didrocks, weird, do you want me do it? [11:39] and Trevinho isn't around or my ping went into void :) [11:39] well, let's wait for Trevinho maybe [11:39] Trevinho usually start to work in the afternoon [11:39] *starts [11:39] ok, let's see, this can wait now :) [11:39] thanks andyrock :) [11:39] ok let me know [11:39] np [11:41] andyrock: oh, while you are around: https://code.launchpad.net/~andyrock/unity/update-dash-string/+merge/139897 [11:41] andyrock: you are not testing the gsettings key if remote search is disabled [11:43] didrocks, you are right. I'll ask John [11:44] andyrock: thanks! :) [11:48] i dont think it's desired to check the key, it's a hint, scopes can ignore it [11:49] and btw hello didrocks, good holidays? :) [11:49] mhr3: yeah, but as told previously, not really succeeded in disconnecting mentally from work. Yourself? :) [11:50] it's a fairy tale that it's possible to disconnect mentally :P [11:50] heh ;) [11:50] but yea, enjoyed time family and friends... got some rest, killed some aliens... good times :) [11:52] Aliens?! [11:52] I rather avoid the world to collapse down replaying broken sword 1 to 4 ;) [12:14] mhr3, hi.... Can I ask you a thing about unity-lens-video, please? [12:46] alo21, sure [13:31] mhr3, did you used zeitgeist to use a signle instance when someone click on a film to see it? [13:33] use* [13:45] alo21, what do you mean single instance? [13:59] mhr3, if I open a new film from the dash, and leave totem opens, and then open a new film (from the dash), I can see the film from totem, which was open [14:01] you mean instead of opening a new totem window? [14:01] zg has nothing to do with that, totem is implemented that way (using GtkApplication) [14:03] mhr3, oh... OK... thank you for your time [14:05] yw === dandrader is now known as dandrader|afk [14:19] didrocks, heyo! sorry for not noticing the PPA failures myself [14:20] mterry: hey! :) No worry, I think I didn't tell you to monitor it and it's transiant until we have the daily build ppa [14:20] so now until we have it, I think you will look at it :) [14:20] mterry: how were your holidays btw? === davidcalle_ is now known as davidcalle [14:21] didrocks, good [14:21] didrocks, wait, how is the daily build ppa different than the staging ppa? [14:21] mterry: btw, I have some small questions if you have time right now [14:21] didrocks, shoot [14:21] https://code.launchpad.net/~3v1n0/unity/remove-bamf-indicators/+merge/140519 <- andyrock told me you spoke about it with Trevinho [14:22] it doesn't need to force the build-dep on latest bamf, right? [14:23] didrocks, mmrazik, so here's some background on the unity staging uploads. We noticed that satoris's PCH commit was causing an FTBFS in launchpad, so I hacked the source upload to disable pch builds to keep the ppa building while satoris debugged the issue... [14:24] fginther: I would prefer in that case that we don't go into this road :/ [14:24] fginther: rather reverting the branch [14:24] and have that in a separate one [14:24] fortunately, I disable daily-build during the holidays [14:24] autolanding on the pch commit passed through jenkins because it does a coverage build which somehow avoids the build failure [14:25] also this kind of hack on a merger side will create a backlog [14:25] fginther: it was still failing on launchpad though? [14:25] see https://launchpad.net/~unity-team/+archive/staging/+packages?field.name_filter=unity&field.status_filter=superseded&field.series_filter=raring === dandrader|afk is now known as dandrader [14:26] every revision failed (even if there are multiple uploads per revision) [14:26] didrocks, correct. launchpad builds continued to fail for future autolands. The rebuild job was disabling use_pch [14:27] (all of the *raring0 uploads are from an autolanding job, the others are from a rebuild due to a changed dependency or manually intervention) [14:28] didrocks, I'll be more cautious in the future and request these commits be reverted when there are launchpad build failures [14:29] fginther: thanks :) [14:29] mterry: forgotten me? :) [14:30] didrocks, oh goodness [14:30] didrocks, you didn't highlight me! :) [14:30] didn't I? [14:30] ah yeah :) [14:30] it was 10 seconds after last sentence! :) [14:30] come on ;-) [14:31] didrocks, right, doesn't need to bump build dep, since it is merely not using old functionality, rather than using new functionality [14:32] didrocks, ain't no one got time to hang around 10s [14:34] :) [14:34] mterry: I think you forgot to rereview https://code.launchpad.net/~3v1n0/bamf/gir-package/+merge/140742 btw [14:34] didrocks, oh ah [14:34] mterry: also, FYI, sil2100 is working on fixing https://bugs.launchpad.net/unity/+bug/1089482 which will fix the latest tests blocking daily landing [14:34] Launchpad bug 1089482 in Unity "Unity HUD uses wrong icon when switching from the Dash to HUD" [Undecided,New] [14:35] and that's it from me for you :) [14:35] didrocks, I talked on IRC about that. We wanted some guidance on gir package naming. Do you know the policy on that? [14:35] That's right! (once I finish lunch) [14:35] mterry: the debian one? I think it's an email [14:35] one sec [14:36] sil2100: lunch? this is sooooo 2012 [14:36] sil2100: with the mobile target, no more lunch! [14:37] I couldn't find the email. I think it's in their bug tracker somewhere... [14:37] mterry: http://lists.debian.org/debian-gtk-gnome/2009/09/msg00011.html [14:39] :'( [14:39] ;) [14:39] didrocks, hrm, OK. That explains how to name the package based on what upstream gives. But I guess I really wanted some guidance on what upstream should do. But clearly the gir typelib does *not* include SONAME (though seems to me like it should?) [14:41] mterry: yeah, the typelib should have it IMHO [14:44] didrocks, that's what's bugging me. These versions they use for the typelib don't make sense. They have Webkit-1.0.typelib, when they should have Webkit-1.0-0.typelib. :( Ah well, guess I can't fight convention here [14:45] mterry: I agree, for me, it should just be the SONAME, it's what is of interest, with a path containing the gir version [15:56] didrocks, btw, you intimated that the staging ppa and the daily build ppa are different? [15:56] mterry: yeah, the staging is per commit, the daily build is per day [15:56] mterry: and we are going to kill the per commit once we have the daily landing on [15:56] ah [16:15] didrocks, why do we still build unity-2d? [16:15] mterry: on precise for SRUs, did you see anything else? [16:16] didrocks, the staging does for quantal [16:16] didrocks, or maybe it's just cruft from when we did build it [16:16] mterry: still? or was it at the beginning of quantal? we didn't remove unity-2d right away, so we can have some versions [16:17] didrocks, OK, deleted the quantal version from the PPA [16:17] thanks :) === dandrader is now known as dandrader|lunch === sil2100_ is now known as sil2100 === dandrader|lunch is now known as dandrader === dandrader is now known as dandrader|afk === dandrader|afk is now known as dandrader [21:31] I have a problem with unity/lightdm. When I try to run unity from lightdm, I always return to lightdm. I only can run unity with "sudo stop lightdm && startx && unity". [21:32] all works fine with the guest session or a new user account. [21:32] maybe I have to remove/reset some (dconf) settings?