/srv/irclogs.ubuntu.com/2012/07/02/#ubuntu-unity.txt

didrockssil2100: hey08:26
didrockssil2100: what's the status about unity/dee/the stack release for quantal?08:27
sil2100didrocks: we were testing the nux + unity + lens, ETA for the stack is for tomorrow08:34
thumperdidrocks: should the unity devs be based on quantal now?08:35
didrockssil2100: ok, I think your coordinated this with upstream, right? and have a versionning with bugs that are closed/new features :)08:35
thumperor is it still ok to be precise?08:35
didrocksthumper: I guess they should, some code was committed in compiz not compatible with the newer boost for instance08:35
didrocksthumper: quantal is quite stable, didn't get any "no X, no sound" issue08:35
thumperdidrocks: orly? like what?08:36
didrocksthumper: I can find it back if needed, but can't happen anymore now. I added automerging to compiz08:36
thumper\o/08:36
thumpernot entirely necessary08:36
thumperI was just curious08:37
sil2100didrocks: I hope I do...09:52
=== yofel is now known as Guest57033
sil2100didrocks: for dee, will we be releasing a new tarball? Or should we stick with the old one? Since I don't see too much changes being made besides the overrides thing...10:19
didrockssil2100: well, I want the overrides thing, then, check with upstream about the better plan for them :)10:19
sil2100ACK10:19
sil2100mhr3: could you comment on that ^ ^10:20
mhr3let me check the changes10:23
sil2100mhr3: I just see the overrides thing and one debugging thing10:27
sil2100mhr3: nothing more10:27
mhr3sil2100, right, so you're basically going to sru current trunk :)10:28
sil2100mhr3: not only SRU - also a quantal release10:28
sil2100mhr3: so, we can stick with the current tarball?10:29
mhr3sil2100, up to you really, if you say you want new tarball i'll make it10:29
sil2100didrocks: in the case of not doing a new tarball release, should I also cherry-pick the 'post release bump' commits from trunk? i.e. the commit that changes the configure.ac version number to a development  version (in this case, 1.0.11)?10:31
didrockssil2100: if you are not doing a new tarball, please don't cherry-pick the bump10:32
sil2100didrocks: ok10:32
didrocksthis is confusing then that the package version doesn't match the library one10:32
sil2100mhr3: if you don't mind, I'd stick with the current tarball for now - since there's not too many changes made, so I personally would think it would be a bit pointless10:33
sil2100mhr3: so I'll just cherry-pick those10:33
mhr3sil2100, as you wish10:34
sil2100mhr3: btw. is the whole dee trunk SRUable? Or are there any commits on there that I should not cherry-pick?10:42
mhr3sil2100, you can take everything10:43
mhr3minus the bump as didrocks said10:43
sil2100mhr3: excellent, thanks!10:44
mhr3sil2100, are you rebuilding lenses with latest vala in quantal?10:56
sil2100mhr3: I think I will have to I guess - is there some work involved with it?10:56
mhr3they'll probably start crashing10:57
sil2100mhr3: and what about dee? Does anything need to be done for everything to work with the new vala? Or should I stick with the old one?10:57
mhr3not right away, we just saw music lens does after activating an album10:57
mhr3dee doesn't care about vala10:57
sil2100Good10:57
sil2100hm10:58
sil2100didrocks: can we stick with the old vala with the lens for now?10:58
sil2100Since I see you started the switch for libunity already10:58
didrocksmhr3: ah! see it's not stable10:58
didrocksseb128: FYI, you were telling the new vala world was great^10:58
didrocksmhr3: can we keep libunity with the new vala?10:59
didrocksas it's fixing the gir bug10:59
didrocksmhr3: sil2100: unity-lens-music is building with vala 0.1411:00
mhr3didrocks, i thought you moved everything to 0.18 for Q?11:01
didrocksmhr3: not everything :)11:01
mhr3didrocks, so what was moved?11:01
mhr3just libunity?11:01
didrocksmhr3: yep11:01
didrocksmhr3: the crash happens with the package?11:02
mhr3didrocks, it's ulm-specific11:02
didrocksdoesn't tell me if it's with the package :p11:02
mhr3didrocks, but i can't really tell if the switch in libunity will introduce more crashes11:02
mhr3i didn't see anything too bad here, but i've been just running some tests, not real system11:03
didrocksmhr3: but this ulm crash, it's not the package right?11:03
didrocksjust if you build locally with latest vala?11:03
mhr3didrocks, no, it'll happen if you rebuild current src wih 0.1811:03
didrocksok, you got me scared ;)11:03
didrocksmhr3: let's see and try libunity with vala 0.1811:04
didrocksand keeping the rest/upgrading to 0.1611:04
didrockssil2100: ^11:04
pstolowskimhr3, didrocks missed beginning of your discussion, but the sooner we switch the better, even if it crashes a lot...11:04
mhr3didrocks, agreed, that's the only way to see what happens :)11:04
didrockspstolowski: no11:04
didrockspstolowski: we won't push into ubuntu versions that crashes11:04
sil2100didrocks: ACK11:04
sil2100So 0.18 for libunity, rest - the old one?11:04
mhr3didrocks, oh, wait, 0.16 will crash as well11:05
mhr3only 14 is safe for lenses atm11:05
didrockssil2100: yeah, apart if when you do the package upgrade, you see a lot of crashes11:05
didrocksmhr3: ok, so keeping 0.14 for lenses11:05
didrockswhich is what they are on11:05
mhr3fine11:05
didrockssil2100: so, basically, you have nothing to do :)11:05
mhr3there's still the icon path 5 -> 611:05
didrockssil2100: just ensure if you do a local build (not in pbuilder), that you selected the right update-alternatives valac version first11:05
sil2100One moment, meeting11:05
sil2100mhr3: I fixed that locally, so I'll just re-do and push11:14
sil2100didrocks: when I'm done, I'd like to confirm some things ;)11:14
sil2100(done with the meeting)11:14
seb128didrocks, re, I was at lunch ... what's up with vala? ;-) please use 0.16 or 0.18 if you can11:20
didrocksseb128: we can't11:21
didrockscrashes with both11:21
didrocksfor lenses11:21
didrocksthat's just the ping was about :)11:21
seb128oh11:21
seb128didrocks, that's likely a bug in mhr3's code!11:22
didrocksseb128: agreeing ;)11:22
mhr3seb128, didrocks but it's because they keep changing vala behaviour :P11:22
* didrocks heard one week before that "vala was really stable now, nothing to fear, blablabla" from 2 people here ;)11:22
seb128lol11:23
mhr3didrocks, you got me now :P11:23
didrocks\o/11:23
=== MacSlow is now known as MacSlow|lunch
=== fenris_ is now known as Guest20501
=== _salem is now known as salem_
=== mhall119_ is now known as mhall119
=== MacSlow|lunch is now known as MacSlow
sil2100Ok everyone, I would like to release a new unity stack for quantal around tomorrow - any objections to that?13:25
gordmhr3: think we should wait till after release before merging in the preview api branch to unity core? ^^?13:27
mhr3gord, well, it's just api, and not hooked up, so it could go in imo13:30
sil2100It won't break anything?13:30
sil2100Since I'd like to ask for a trunk freeze for tomorrow13:30
gordshouldn't do13:32
mhr3anyhow, someone will have to approve the branch first :)13:33
gordneeds to wait for QA to re-review, yell at thomi tonight and maybe it'll get merged in before the morning13:40
sil2100Well, I do hope that tomorrow we'll have a freeze, since otherwise packaging work is a bit pointless13:40
mhr3i think qa is broken atm, cause i pushed and it didn't re-complain13:40
gordi mean thomi did a review already and wanted some changes right?13:41
gordnot the qa lab stuff13:41
gordthe qa lab stuff is just dice rolling13:41
mhr3ah, no, marco was looking at it13:41
mhr3Trevinho, so feel free to re-review now ;)13:42
gordhum, you are right. whos review am i thinking of..13:42
Klap-inseb128: i have a question about a bug fix, is there a chance that it can be included in the precise? https://bugs.launchpad.net/ubuntu/+source/compiz-plugins-main/+bug/90420513:46
ubot5Ubuntu bug 904205 in Compiz "Desktop wall: Bindings for next/previous don't wrap to the next row" [Wishlist,In progress]13:46
seb128Klap-in, hey, yes it seems to be fine for a SRU13:47
seb128Klap-in, but it needs to be reviewed,commit first13:47
sil2100Klap-in: could you submit a merge request? Probably best for both compiz trunk and compiz-plugins-main13:48
Klap-insil2100: ok, i did a merge request on the compiz-plugins-main13:49
Klap-inon the default, thus on the devel branch.13:49
sil2100Klap-in: ah, see it now, thanks13:50
=== Guest57033 is now known as yofel
sil2100Klap-in: we'll try having that reviewed13:50
Klap-inis a extra mergerequest on compiz usefull?13:50
sil2100Klap-in: well, it's good to have the same fix applied on both trunk and precise - if the bug appears on both, which is the case here13:51
sil2100lp:compiz now has all the plugins merged into its source tree, so it's essentially just cherry-picking it into the new tree and testing if it works on trunk13:52
Klap-insil2100: ok, then i do also a merge request on compiz13:57
=== zyga is now known as zyga-food
Trevinhomhr3: you was referring to the ap fix branch? or to the previews one?14:10
mhr3Trevinho, the latter14:11
Trevinhomhr3: ok, looking at the introspection bamf...14:12
Trevinhomhr3: I guess we had a bug about it to link...14:12
mhr3Trevinho, hmm, i'm still wondering if there should be two girs (one for gtk2)14:13
Trevinhomhr3: probably yes...14:14
mhr3do we care about gtk2? :P14:14
mhr3it's not like introspection worked too much there :)14:14
Trevinhomhr3: I mostly did on the code... there was some projects depending on libbamf (not 3 version)14:15
=== dandrader is now known as dandrader|afl
=== dandrader|afl is now known as dandrader|afk
=== zyga-food is now known as zyga
Trevinhomhr3: bamf_matcher_get_xids_for_application returns a GArray that should be free'd... (even if now it's not implemented, I've implemented in the gdbus branch [that it's done, but I'm always forgetting to request for merge...])14:26
mhr3Trevinho, that's why the annotations are in the .c files, so they can be easily updated when the method changes14:28
sil2100didrocks: regarding music lens - do you want to release a new tarball with that one commit you got merged in, or can I do it as a patch?14:34
didrockssil2100: would be nice to do a release for this one I guess14:34
sil2100didrocks: a new tarball, yes?14:36
sil2100ACK14:36
Trevinhomhr3: bamf_matcher_get_active_window could return null (if not connected to the daemon or other odd cases)14:36
sil2100didrocks: so maybe we'll SRU it as well?14:36
didrockssil2100: hum, no14:36
didrockssil2100: how do you want to SRU that?14:36
didrocksit's a new feature :)14:36
sil2100didrocks: ok - just saw bugs linked to it like 'unknown albums might bla bla' so I thought it's a bugfix too14:37
didrockssil2100: it is, but it adds translations14:37
didrockssil2100: as if you looked at merge request content would have tell you ;)14:38
didrockssil2100: you really need to start looking at those!14:38
sil2100didrocks: I know I know! But asking first never hurts! ;)14:38
sil2100Besides, I have yet to master vala14:39
sil2100As my vala-foo is weak14:39
didrockssil2100: well, you really don't want to :p reading is easy, writing is less IMHO :p14:40
sil2100didrocks: since we will be switching to 6.0 with this release, should I switch the paths from 5 to 6 in the new tarball already?14:40
sil2100For the assets14:40
sil2100Or should I leave it as 5 and just do a patch with switching to 6?14:41
mhr3Trevinho, actually i didn't use that properly, gir specifies nullable only for method parameters14:42
sil2100hm, crap, I'll have to submit merge requests to all the lenses :|14:42
mhr3Trevinho, pushed fix14:42
didrockssil2100: don't move it if it's not necessary to move it14:42
Trevinhomhr3: oh, yes... so we can avoid to specify that14:43
sil2100didrocks: so a patch?14:43
Trevinhoimplementers should always perform null checking...14:44
didrockssil2100: if upstream moved it, move them14:44
didrocksif not, don't :)14:44
sil2100didrocks: well, when releasing unity 6.0, we'll have to switch from 5 to 6 or the icons will be not visible - upstream lenses didn't move yet because unity 6.0 is not released yet14:45
sil2100But it will have to sooner or later14:45
sil2100Anyway, I'll release the tarball in the state the trunk is now14:45
didrockssil2100: good :)14:45
sil2100And then just proceed normally ;)14:46
mhr3didrocks, you evil little you, don't discourage sil2100 from writing patches for us :P14:46
sil2100mhr3: ;)14:46
sil2100mhr3: you want to roll-out the new tarball, or should I do it?14:46
sil2100mhr3: for the music lens?14:46
mhr3you want a tarball?14:47
mhr3did we actually have changes?14:47
didrocksmhr3: I am that evil!14:47
didrocksmhr3: you have awesome changes14:47
didrocksin the music lens14:47
mhr3didrocks, of course we do :)14:48
mhr3sorry didn't notice your stuff got merged already14:48
didrocksmhr3: I see how deeply sorry you are, that's fine :-)14:49
mhr3omg!14:49
mhr3a see a tab in the diff14:49
mhr3how could you?!14:49
sil2100!!!14:50
gordBURN THE TABBER14:50
sil2100RED, cannot release! Tab in the diff!14:50
Trevinhomhr3: in lp:~mhr3/unity/update-core-preview-api what about using std::unique_ptr for pimpl? (just to be more C++11 :))14:52
mhr3Trevinho, can't, it doesn't compile14:52
Trevinhomhr3: yes, you just need to keep an empty destructor14:53
Trevinhomhr3: keep the ~Class(() {}... and it will work14:53
mhr3Trevinho, for the Impl?14:53
mhr3Trevinho, or the Preview subclasses?14:53
Trevinhomhr3: no, for the wrapper class14:53
mhr3sil2100, so i suppose that would be ulm 6.0?14:55
sil2100mhr3: hm, not sure14:56
sil2100mhr3: ...probably?14:56
mhr3sil2100, and should i bump the icon path in the .lens file as well?14:57
sil2100mhr3: well, the icon path - I originally wanted to wait with that till unity 6.0 gets released - but if the tarball will have version 6.0, I think this would make the most sense14:57
sil2100mhr3: so I think yes15:01
mhr3crap15:11
mhr3i almost tagged bamf instead of ulm15:11
mhr3sil2100, anyway tarball up15:14
sil2100mhr3: \o/ Thanks!15:21
mhr3sil2100, btw when producing tarballs i push straight to trunk without the mergers, that's how we used to do it with didrocks15:22
mhr3not sure what/if that changes anything for you15:23
sil2100mhr3: not really, it's ok for me too15:30
mhr3Trevinho, pushed unique_ptrs15:35
Trevinhomhr3: good ;)15:36
mhr3and i continue to think c++ is stupid :)15:36
mhr3i would expect it to have proper error messages after being developed for 20+ years15:39
mhr3but i suppose they're too busy having to implement glr parser :P15:43
sil2100didrocks: new tarball imported (new upstream release) in packaging for ulm: lp:~sil2100/unity-lens-music/quantal_PRE15:54
didrockssil2100: please, prepare everything and I will review when everything is ready15:56
didrocksand that you have tested and so15:56
sil2100didrocks: ACK15:56
sil2100Well, I have almost everything else prepared too, just waiting for the possible trunk freeze tomorrow for unity and nux15:57
sil2100So that I won't have to re-do all the work15:57
sil2100mhr3: regarding libunity - you want to release a new tarball, or you think it's too early?16:17
=== zyga is now known as zyga-afk
mhr3sil2100, we released 5.90 already16:22
mhr3although i should do a new one16:22
mhr3sil2100, give me a moment16:22
sil2100mhr3: with this one we're switching to the new vala, so maybe a new release would be good16:23
sil2100Actually, hmm16:24
mhr3hmm at very least make check passes with latest valac :)16:26
mhr3didrocks, weren't you going to patch libunity to ship the gir or something?16:27
didrocksmhr3: we did it16:27
didrocksbut wanted to remove it16:28
didrocksas the latest valac fixed it16:28
didrocks(so removed in the package)16:28
mhr3didrocks, but wasn't the gir missing for -dev?16:28
mhr3s/for/from/ ?16:28
didrocksmhr3: it's in now16:28
mhr3but it should be there anyway, no?16:29
sil2100Ok, so no new tarball for libunity?16:29
mhr3sil2100, it's on the way16:29
didrockswe ship 5.12.0-0ubuntu2 in quantal16:29
didrockswith the gir fix16:29
mhr3didrocks, ok nevermind, i thought we're forgetting to install it, but that doesn't seem to be the case... i guess i misunderstood you16:30
didrocks5.90 was something I wanted that we do during the sprint16:30
didrocksmhr3: no, you did understand me very well, but in this case, it was more "blame the packager", but shhhh :)16:30
mhr3heh ok :)16:31
sil2100...me? ;)16:31
didrocksno, was me :p16:32
mhr3sil2100, released16:34
mhr3sil2100, btw i'm evil and doing post-release bumps, you shouldn't push those to distro ;)16:35
mterryHeyo folks.  Just wanted to raise bug 1020115 to attention.  It affects the latest update-manager16:35
ubot5Launchpad bug 1020115 in unity (Ubuntu) "Spawned commands can't use pkexec" [Undecided,New] https://launchpad.net/bugs/102011516:35
seb128seems like a mhr3's bog! ;-)16:36
mhr3seb128, you knew something about that though, didn't you? ;)16:36
seb128mhr3, me? no, you must confuse me with somebody else16:37
seb128;-)16:37
mhr3seb128, your jedi tricks won't work on me :P16:37
seb128:-(16:37
seb128mhr3, that would likely be a bug in the app lens right?16:40
sil2100Holy shit, bzr merge-upstream just crashed16:40
mterryseb128, it happens even from indicators16:41
seb128mterry, glib doing? g_spawn_...?16:41
mhr3seb128, well... we just use glib, so...16:41
mterryseb128, I wouldn't think it would set ppid to 1 automatically16:42
mhr3seb128, strangely synaptic uses it, but it's executed from a wrapper script16:43
mhr3(and works)16:43
seb128mhr3, mdeslaur added it to synaptic with that comment16:43
seb128"- debian/rules, debian/synaptic-pkexec: install wrapper to work around16:43
seb128      .desktop files not being able to call pkexec directly."16:43
sil2100Did anyone get an error like this before in bzr merge-upstream?16:44
mhr3doesn't really tell why :/16:44
sil2100NoSuchRevision: CHKInventoryRepository('file:///home/sil2100/Work/canonical/release/libunity/tmpVbvIP4/upstream/.bzr/repository/') has no revision didrocks@ubuntu.com-20120614074747-nnnfjs5hyplbqaj016:44
mhr3sounds like it's missing something16:45
sil2100Ah, one moment16:45
didrockssil2100: never got that, did you -r -1?16:45
seb128sil2100, try asking #bzr?16:45
didrocksor -r -216:45
sil2100Yes, but uno moment16:45
sil2100o16:45
didrockswith the post release bump :)16:45
mterrymhr3, specifically, ppid is 1.  I don't know why unity is causing that, but that's the root cause of the pkexec problem16:45
mhr3mterry, oh16:46
mhr3mterry, glib reaps the child processes by default16:46
seb128mterry, is that specific to unity or any g_spawn wrapper will do that? you said indicators have the same issue16:46
sil2100didrocks: probably the problem is in the fact that in the distro branch, there's an unreleased version present16:47
mterrymhr3, ah, so the g_spawn* family does set ppid to 1 automatically16:47
mhr3and when launching a desktop, there isn't really a way to *not* do that16:47
sil2100Aaaa16:47
sil2100No, waaait16:47
sil2100My mistake it seems!16:47
didrockssil2100: no, it's not the problem16:47
sil2100grrr16:47
didrocksah ok :)16:47
didrockssil2100: it seems that you are releasing before testing right?16:47
mterrymhr3, seb128: OK, so you think it's not a unity problem.  Do you think pkexec is being too strict, or should execl just be forbidden in this situation?16:48
didrockssil2100: meaning that potentially the tarballs are not working fully/tested?16:48
didrocksso low quality?16:48
didrocksnever trust mhr3's tarballs :)16:48
mhr3my tarballs have a "it's a trap" sticker16:48
sil2100didrocks: I'm not releasing anything, I'm preparing everything for testing16:48
sil2100didrocks: I prepare my distro branches so that I can build clean test packages from those16:49
didrockssil2100: well, mhr3 did the release and published the tarball, right?16:49
sil2100And if it's green, I then just push it to the ether16:49
mhr3mterry, can't say really, maybe :)16:49
didrocksin case the tarball is not good, you would redo all the import?16:49
mhr3mterry, ie perhaps it has a reason :)16:49
sil2100didrocks: I would just do it all over again - but since it's only stored locally on my machine, it's not a big deal16:50
sil2100didrocks: and if it's not broken, I don't have to do anything but just pushing ;)16:50
mhr3mterry, but as can be seen with synaptic, a simple wrapper script workarounds the issue16:50
seb128mterry, I'm not sure to understand enough of the rational to comment on that, the comment seems to indicate that's to avoid running the command if the processus exited16:50
seb128mterry, would be good to check with mdeslaur or pitti16:50
didrocksseb128: as long as the tarball is not published into the Internet and no confusion is around :)16:51
seb128sil2100, ^16:52
seb128sil2100, didrocks has tab completion issues ;-)16:52
sil2100It happens to me as well!16:52
didrocksgrrrr ;)16:54
didrocksseb128: stop speaking!16:54
didrockss<tab> complete the latest one :)16:54
sil2100didrocks: btw. the bzr merge-upstream problem I had - it seems to be caused by the previously unreleased version16:55
sil2100Since when I removed it, it worked fine16:55
didrocksok16:55
sil2100Do I need something specific to make it work, or should I just remove the unreleased one and add it as the new one?16:55
mhr3didrocks, that's what you get for asking me to do tarball and then not releasing it! :P16:56
* sil2100 is confused with all the maddness going on here!16:56
sil2100You guys are crazy!16:56
sil2100;p16:56
mhr3sil2100, don't worry, it's infectious16:56
sil2100:'(16:56
mhr3you'll be assimilated16:57
mhr3anyway, eod.. cyas16:58
didrocksmhr3: well, blame the sprint :)16:58
sil2100Ok guys, I'm finishing for today as well, time to let the PC cool off from all this afternoon heat-maddness16:59
sil2100See you tomorrow!16:59
didrockssee you sil210016:59
MCR1Do developers here have any comment on this: http://www.ubuntuvibes.com/2012/07/unity-revamped-adds-always-visible.html ?17:05
=== dandrader|afk is now known as dandrader
=== dandrader is now known as dandrader|afk
=== dandrader|afk is now known as dandrader
=== philipballew_ is now known as philipballew
=== mfisch` is now known as mfisch
=== mfisch is now known as Guest84007
=== dandrader is now known as dandrader|afk
TrevinhoMCR1: you should probably ask designers...20:44
=== dandrader|afk is now known as dandrader
thumpermorning20:59
MCR1Trevinho: Where can I find the designers ? ;)21:18
MCR1Is someone here responsible for maintaining the Unity/staging PPA Quantal ?21:19
MCR1Unity in the PPA still depends on compiz-core-abiversion-20120305 and libgnome-desktop-3-2, so it is not installable :(21:22
MCR1I wonder how it passed the unit-tests :-/21:23
TrevinhoMCR1: you can ask on unity-design ML, or to JohnLea21:31
seb128not sure about the daily ppa but the packagers are after their end of work day21:33
popeythey're all tucked up in bed getting their beauty sleep21:35
popey(or should be)21:35
MCR1Trevinho, seb128 thx 4 the infos.22:12
=== salem_ is now known as _salem

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!