[00:00] nope. maybe not [00:00] but i can't seem to trigger it by hand [00:03] Can laptop number two still reproduce? Let's find out! [00:13] RAOF: i wrote the (admittedly braindead) code in http://pastebin.ubuntu.com/706392/ to try and poke this into happening by hand, but it doesn't seem to [00:13] does that look right to you? [00:14] i wrote it based on the output from http://pastebin.ubuntu.com/706368/ [00:18] You might want to ungrab the server at the end, but that looks about right to me. [00:19] oh wait - that time it worked [00:19] well, "work" [00:19] it seems to only sometimes result in the broken state [00:19] Heh. [00:20] And if you pass 0 as the timestamp? [00:22] also breaks intermittently [00:22] in both cases it's much more intermittent than when gnome-desktop does it [00:25] Let's see if I can get a laptop to do that... [00:39] argh. it would be nice if compiz didn't crash every time i re-configured my desktop [00:40] It mostly manages to not crash here. [00:40] it crashes consistently every time i switch back to two displays. looks like it might actually be unity and not compiz [00:40] null boost::shared_ptr deref [00:44] Oh, it's totally going to be unity, yes. [00:44] broder: How many iterations do you tend to need before that engages in broken state for you? [00:45] not sure. i actually haven't been able to reproduce it recently [00:47] and we have a winner! so...apparently 10 or so with the code i pastebinned [00:47] Because my next step was going to be putting 0x4 into /sys/modules/drm/parameters/debug and seeing the differences between a correct and incorrect modeset. [00:48] i assume i need to do that before triggering the issue? [00:48] ooh [00:48] Yeah. [00:48] there's a bunch of "[drm:i915_get_vblank_timestamp], crtc 0 is disabled" in dmesg already [00:48] relevant? [00:49] I don't _think_ so, no. [00:50] Except, perhaps, that i915 probably should either grab vblanks from crtc 1 when it's the only one enabled, or that it shouldn't spew irrelevant debugging mesasages if it already does that :) [00:51] argh. ok, now that i've got it to work, my code always results in the broken state [00:51] let me...reboot and then see if i can get dmesg from not-broken [00:52] in the mean time, http://paste.ubuntu.com/706405/ should have the broken transition [00:52] it probably also has a bunch of other stuff [00:53] and i stripped out those "crtc 0 is disabled" complaints [00:53] the last transition should be the one that broke [00:59] RAOF: http://paste.ubuntu.com/706406/ should be a good transition, i think [00:59] RAOF: going to go grab food; bbiab [01:00] TA [01:01] Or, even, ta. [01:15] someone throw me a bug [01:15] like [01:16] I'm actually out of bugs for this SRU [01:16] and I have 14 times more as anyone else [01:16] smspillaz, congratz? [01:16] thanks but [01:16] I need to fix something today [01:17] smspillaz: broder was having compiz crash every time he re-enabled his second monitor [01:17] already fixed that [01:17] https://launchpad.net/unity/+milestone/4.26.0 <- anything that's not fix committed or in progress with a branch linked [01:19] You could fix framebuffer > GL_MAX_TEXTURE_SIZE in unity, right? That'd be sruable :) [01:20] it's a massive change [01:20] I'd have to rewrite core [01:20] Yes, I know. [01:20] If you're really hurting for things to do, I'm *sure* you could toolsmith / unittest. Everyone loves test suites! [01:21] :( [01:21] hehe [01:29] smspillaz: is there a PPA or something i can grab from? [01:30] RAOF: also, back now :), in case there's any other info that would be helpful [02:13] smspillaz: I think the correct solution for GL_MAX_TEXTURE_SIZE is xrandr 1.4 [02:13] or am i misunderstanding how outputs work with that? [02:17] *if* that happens, last i heard there were no plans to even get it in anymore [02:26] Sarvatt, no specific plans but I think it's just backburnered; I expect we'll see xrandr 1.4 eventually [02:31] It got raised again in IRC; aaronp clarified that he didn't object to the API, just to the time it was introduced last time. [03:06] broder: oneiric-proposed [03:07] Amaranth: eh? what about windows that have absurdly large sizes ? [03:07] smspillaz: you mean that cover two outputs? [03:07] (or single output sizes that happened to be larger than the maximum texture size) [03:07] (which happens on ARM) [03:07] even ARM GPUs support 4096 [03:08] right, although I read somewhere that there are some arm chips that have ridiculously huge scanout buffers, can't remember which ones though [03:09] its probably better to do the same trick that we have for textures in compiz, eg, split one image up into lots of little textures with appropriate offset matrices [03:09] and then maintain an appropriate scene translation inside of the fbo [03:35] grrr translations from X co-ordinate space to GL co-ordinate space [03:35] gets me every time [03:41] yay +1 for me, fixed the stupid artefact [03:52] RAOF: if you have nvidia/fglrx/radeon/intel or have multimonitor with different resolutions can you give https://code.launchpad.net/~smspillaz/unity/unity.big_fbo/+merge/78202 and see if there are any rendering issues ? [03:52] *give it a try === eeejay is now known as eeejay_is_afk === smspillaz is now known as smspillaz|out [03:53] Good morning [04:00] Morinng. [04:02] smspillaz|out: You'd like that tested with nvidia binary? [04:04] RAOF: anything really. [04:05] Ok. I'll whirlygig on an intel system, then. [06:36] Ok, that's enough of this monitor blinking every 10 seconds. [06:44] RAOF: ?? [06:45] Oh, just flipping HDMI1 from CRTC to CRTC every 10 seconds. This doesn't reproduce the problem where it sometimes goes CRAAAAAZAY. === tkamppeter_ is now known as tkamppeter [06:45] ok [06:46] RAOF: did you get anything useful from my kernel logs? [06:46] broder: No, they looked like perfectly ordinary modesetting. Setting various outputs to [NOCRTC] then associating a crtc with it. :/ [06:47] hmph, ok [06:47] http://paste.ubuntu.com/706517/ is what I'm currently modifying to flip between dual-head and single-head. [06:47] RAOF: can you tell me if https://bugs.launchpad.net/ubuntu/+source/unity/+bug/796513 is still an issue? I don't think it is anymore as far as I can tell [06:47] Launchpad bug 796513 in hundredpapercuts "Switching to application with multiple windows alternates the target windows" [Undecided,Confirmed] [06:48] My goal now is to write a test case that reliably reproduces, and send it upstream. [06:48] smspillaz|out: I'm pretty sure that got fixed early on in the new alt-tab? I shouted at DBO until it worked, because that behaviour was infuriating. [06:48] kk [06:49] yay I have no more bugs to my name [06:49] time to go to uni to see if I actually have this class today or not [06:49] :) [06:50] Oh, man. Now that I see that bug I know what bug I should have filed a bug for you to fix :) [06:50] But I won't tell you about it until you've been to uni and back :) [07:03] hi pitti. is it too late for bug 872389? :) [07:03] Launchpad bug 872389 in webfav "Please remove source and binaries from Oneiric" [Undecided,New] https://launchpad.net/bugs/872389 [07:05] ooh, can we please :D [07:06] chrisccoulson: did you ask ogra_? [07:07] chrisccoulson: no, it's not [07:07] micahg, i haven't, but i'm sure i've mentioned it before to him [07:07] pitti - cool :) [07:07] zero rdepends [07:08] * pitti kilsl [07:08] kills, too [07:08] pitti: epiphany-extensions-more also waiting to be removed as long as you're in the mood :) [07:08] is there a bug for it? [07:08] bug 870187 [07:08] Launchpad bug 870187 in epiphany-extensions-more "Please remove epiphany-extensions-more source and binaries from oneiric" [Wishlist,Confirmed] https://launchpad.net/bugs/870187 [07:10] done; I'll run through the other ubuntu-archive bugs to see what to mop up [07:11] pitti - thanks :) [07:11] pitti: thanks [07:11] with backports enabled by default, it's a lot easier to wield the removal axe as long as there are no rdepends [07:11] RAOF: which one ? === smspillaz|out is now known as smspillaz [07:13] RAOF: (didn't have a class, false alarm) [07:13] morning [07:15] broder: Does http://paste.ubuntu.com/706526/ reproduce for you? It switches dual-head?single external display?dual head, switching the CRTCs around. And it works peachily here. [07:16] smspillaz: The bug that I'm thinking of is where an app maps a window, it comes to the top but doesn't have focus, and compiz doesn't know that it's on top. You need to select it before you can switch away. [07:17] RAOF: oh I've fixed that [07:17] RAOF: use oneiric-proposed [07:17] OMG HUZZAH [07:17] * RAOF enableds oneiric-proposed forthwith. [07:17] ? [07:17] RAOF: annoyed the shit out of me too [07:17] That was just one of the more annoying sporadic annoyances :) [07:17] then I realized it was a one line typo [07:18] Ah, joy. [07:19] so uh [07:19] any other bugs ? [07:23] pitti: seeing your ubuntu-default-builder upload, we should enable choosing the greeter as well. (I added an option --greeter to lightdm-set-defaults to lightdm for derivatives). Even if the file isn't there, it will just DTRT [07:23] didrocks: ah, right; but that's something for precise [07:23] didrocks: .. and good morning [07:24] pitti: indeed, I'm adding that to a note :) [07:24] oh, seems I didn't say good morning there, indeed! guten morgen pitti :) [07:25] hi, I was wondering is it known that sometimes dash does not appear when "Super" key is pressed? It takes me to press super until the launcher appears, depress it and press again to bring dash [07:25] so, please feel free to report a bug for it, to avoid forgetting about it [07:26] pitti: doing so [07:26] rye: yeah, seeing that as well, never found a good reproducible testcase though [07:30] mvo: good morning [07:31] mvo: thanks for the s-c reupload; not sure whether you saw, but software-properties also needs a reupload to -proposed [07:35] pitti: I didn't, have a mail here "subject: [ubuntu/oneiric-updates] software-properties 0.81.11 (Accepted)" [07:36] mvo: right, and a followup to -devel@ that I immediately removed it again [07:36] and a followup in the bug report [07:37] I saw too late that it went to -updates [07:37] pitti: ok, I have not seen this one, sorry for that, I will reupload [07:37] mvo: danke! [07:39] pitti: interesting, the ftp accepts upload to -updates? [07:40] yes [07:40] it shouldn't probably [08:01] hey [08:01] salut seb128 [08:01] hi [08:01] lut, ca va bien ? [08:01] hi seb128 [08:01] hey smspillaz chrisccoulson [08:01] how are you? [08:02] i'm good thanks [08:02] how are you? [08:03] salut seb128 [08:03] hey pitti [08:03] chrisccoulson, I'm fine thanks [08:04] seb128: ça va bien, et toi? [08:04] nickel! ;-) [08:06] didrocks, just curious but did you point rick or jasoncwarner_ to the u1 lens scope email discussed on the list? [08:07] didrocks, (I will do it if you didn't yet) [08:07] seb128: not yet, waiting for my o1o if we have one [08:07] ok [08:07] I will let you handle it then [08:07] thanks [08:07] thanks :) [08:08] didrocks, just for the record (not sure what is your exact thinking on it) I think it's worth to have some thinking on the potential issue and address concerns before it becomes one [08:08] i.e I don't think we will change that not for Oneiric but we might want to communicate officially on it [08:09] seb128: indeed, agreed, that's why I wait for my 1o1 to raise it and see what should be done before a more formal discussion at UDS [08:09] thanks [08:09] chrisccoulson, did you manage to catch federico yesterday? [08:09] seb128, no, but i did send him an e-mail in the end [08:09] chrisccoulson, ok, good as well ;-) [08:10] chrisccoulson, there is a spinning bottle on this channel, whoever it stops on has to fix gnome-desktop gnome-rr next cycle :p [08:10] lol [08:11] i should leave this channel and hang out in #ubuntu-devel instead then ;) [08:12] ;-) [08:12] I should do a cross channel bottle spinning :p [08:12] heh :) [08:12] joke aside a good part of the g-c-c bugs are with the xrandr panel [08:13] that doesn't surprise me ;) [08:13] Yeah; it's a difficult part. [08:13] I've been serial closing GNOME2 bugs that don't apply to g-c-c in GNOME3 yesterday [08:13] and I noticed like a good chunck of the still open bugs are xrandr ones [08:13] what doesn't help is that federico doesn't really work on it nowadays and nobody else picked it up [08:13] isn't that because it's pretty much the only thing that you can configure in g-c-c now? ;) [08:13] combined with ping-pongs "that's an xorg bug, not it's not" [08:14] chrisccoulson, in a troll mood today? ;-) [08:14] :) [08:14] but yeah, I closed over an hundred bugs with "that option doesn't exist in GNOME3" [08:14] so I confirm, g-c-c dropped quite some options on the way :p [08:15] bryceh and I are thinking about extracting a librandr-utils out of the upstream xrandr tool and gnome-rr (and, concievably, whatever KDE does here too) so that there can be a shared set of randr algorithms that is maintained in Xorg where the actual driver developers will run into it. [08:18] RAOF, that would be great [08:23] pitti: I uploaded new software-properties now, both .12 and .13, sorry for that, only .13 is relevent, the fix was split into two different branches so I missed the second one initially [08:24] mvo: so want me to reject http://launchpadlibrarian.net/82586937/software-properties_0.81.12_source.changes ? [08:24] pitti: please do [08:25] mvo: .13 should then include both changelogs (build with -v), or you reupload that as .12 [08:25] pitti: I think re-uploading as .12 is the best choice then [08:26] pitti: the changelog for .13 should be good though as its essentially the same fix, just done in two different branches (the debian/ bit was in the lp:ubuntu/ namespace, the generic one was in trunk) [08:27] mvo: ah, sounds fine then; as long as .13 refers to the same bugs as .12, I'm happy [08:27] pitti: yeah, it does :) [08:28] uhm...wow? I'm running proposed. Has compiz/unity spread just taken a nose dive in terms of usability? I can't move windows around my spread without it flipping out? [08:28] didrocks njpatel ^^ [08:29] jasoncwarner_, you mean that they are jumping around? [08:29] jasoncwarner_, also, you mean expo not spread, right? [08:29] smspillaz, you wanted bugs, can you figure out what's wrong? ^ [08:29] njpatel: yeah, among other things. like, right now, I can't even "find" my firefox window. it is open, but alt+tab won't switch to it and luancher won't bring i tup [08:30] jasoncwarner_, did compiz crash recently? [08:30] jasoncwarner_, or did you restart it manually [08:30] ? [08:30] njpatel: not since I rebooted [08:30] njpatel: it crashed last night and then I rebooted [08:30] I'm going to take a video to show you. [08:31] thanks [08:31] jasoncwarner_: it will be fixed next week [08:31] jasoncwarner_: already uploaded the fix for that to oneiric [08:31] jasoncwarner_: are you using -proposed? [08:32] smspillaz didrocks yeah, running proposed [08:32] jasoncwarner_: had to make a tradeoff between having jumpy windows and fixing a stacking bug temporarily [08:32] so, it's the jumping windows bug [08:32] but I've fixed the jumpy windows in trunk now [08:32] smspillaz: we should maybe reput the fix and reset the counter then [08:32] this feels like more than jumping windows, like I can't find firefox windows sometimes [08:32] right, the snap plugin will go a bit nuts [08:33] yeah, like I can't actually get to my firefox window now and I have no idea how to get it back. [08:33] didrocks: so you want a patch for it then ? [08:33] smspillaz: yeah, let's reset the counter for it with this additinal patch [08:33] smspillaz: the firefox issue seems different though, isn't it? ^ [08:33] it's a "bug" in the snap plugin [08:33] so you'll need to do a cpm update too [08:33] :/ [08:34] ok, resetting the counter for both [08:34] pitti: FYI ^ [08:34] smspillaz: cpm update? [08:34] jasoncwarner_: compiz-plugins-main [08:34] chrisccoulson, what did you mention to me ? [08:34] smspillaz: ok then, simple question, how do I get my firefox window back ;) [08:35] jasoncwarner_: *shrug* [08:35] didrocks: if you do a followup upload, please use -v to include the previous SRU changelog [08:35] jasoncwarner_: metacity moves stuff back on screen on --replace [08:35] so you can metacity --replace & compiz --replace & [08:35] pitti: sure, thanks! [08:36] smspillaz: can you point me to the 2 patches? [08:36] didrocks: I need to make them first [08:36] or can you just do it from the merge proposal ? [08:36] smspillaz: point me to the merge proposal [08:37] sure [08:38] https://code.launchpad.net/~compiz-team/compiz-core/compiz-core.fix_864330/+merge/78935/+preview-diff/+files/preview.diff [08:38] https://code.launchpad.net/~compiz-team/compiz-snap-plugin/compiz-snap-plugin.fix_872161/+merge/78927/+preview-diff/+files/preview.diff [08:39] smspillaz: thanks [08:39] np [08:40] jasoncwarner_, njpatel, smspillaz: but did the libreoffice issue from yesterday went anywhere? do we have a bug report tracking it? [08:40] smspillaz: can you add test cases on bug #864330 and bug #872161 please? [08:40] Launchpad bug 864330 in unity "The larger windows, moving towed." [High,Fix committed] https://launchpad.net/bugs/864330 [08:40] Launchpad bug 872161 in unity "When grabbing a window to demaximize it after dragging it up, sometimes the position will not be where you expect it to be" [Medium,Fix committed] https://launchpad.net/bugs/872161 [08:40] smspillaz: I'm targetting the right component, subscribe the team and such meanwhile [08:40] seb128: I never created a bug. I asked Sweetshark to work with njpatel and others. Sweetshark njpatel is there a bug? [08:40] seb128, not sure, jason was looking into it, was waiting for a bug report from jasoncwarner_/sweetshark [08:41] waiting = to milestone [08:41] not to work on the fix, of course [08:41] didrocks: ok [08:43] didrocks: done [08:46] njpatel, jasoncwarner_: bug #844119 is somewhat similar but not exactly the same [08:46] Launchpad bug 844119 in bamf "Hitting Alt-Tab does not display Libre Office in list of switchable open applications" [Undecided,Confirmed] https://launchpad.net/bugs/844119 [08:46] or it's part of the issue [08:47] so, um, anyone like to hazard a guess as to why I can't log in this morning? lightdm accepts my password and tries to start a session, but apparently fails and drops me right back to the greeter [08:47] cjwatson: could you please put your ~/.xsession-errors somewhere? and perhaps /var/log/lightdm/lightdm.log? [08:47] smspillaz: hum, the first patch doesn't apply partially, there are a lot of context change, can you have a look at it? I don't want to mess with this [08:47] smspillaz: looking at cpm meanwhile [08:47] bug #758613 as well [08:47] Launchpad bug 758613 in unity-2d "In Firefox and LibreOffice sometimes the global menu does not appear." [Medium,Confirmed] https://launchpad.net/bugs/758613 [08:48] pitti: .xsession-errors has nothing written to it since late last night [08:48] cjwatson, stalled .Xauthority from a previous session or with wrong user ownership [08:48] that's my bet ;-) [08:48] didrocks: sure [08:48] http://paste.ubuntu.com/706574/ [08:48] oh yes, .Xauthority is root:root, how did that happen [08:48] cjwatson, if you have a .Xauthority rm it and try again? [08:48] cjwatson, ok, I've seen a few bugs about that [08:49] cjwatson, bug #871667 [08:49] Launchpad bug 871667 in lightdm "Stalled .Xauthority can block login" [Undecided,New] https://launchpad.net/bugs/871667 [08:49] cjwatson, what do I win? (please don't tell "the right to fix the bug :p" ;-) [08:49] WARNING: Error reading existing Xauthority: Error opening file: Permission denied [08:49] seb128: chances are high that you do [08:49] pitti, :-( [08:50] cjwatson: so it's likely that bug indeed; you can kill ~/.Xauthority (check its permissions before) and try again? [08:50] right, that works now, thanks a lot [08:50] pitti, " oh yes, .Xauthority is root:root, how did that happen" [08:51] stgraber had the same problem yesterday [08:51] cjwatson, yw, you can probably subscribe to that bug [08:51] yep [08:51] I will point it to robert_ancell [08:51] not sure how that happens but it's bitten some users so there is a bug somewhere [08:52] I got it sometimes on my 10v test box but not recently, I though it was maybe fixed but seems not [08:54] seb128: I had that the other day too! I suspected it was me messing around with sudo and testing gui upgrades or somesuch [08:54] didrocks: hmm, did we ever get my fix to fix window shading ? [08:54] that context change is about that [08:55] maybe I might have missed it when I was giving you patches [08:55] didrocks: anyways, do you want the window shading fix + the fix for this, or just this fix ? [08:55] (the window shading fix fixes windows disappearing forever once they are shaded ;-)) [08:58] smspillaz: I remember we had this kind of fix, but seems not, so yeah, please, the second one as well, with a bug # + step to reproduce [08:58] cjwatson, pitti, mvo: robert_ancell wontfixed bug #855773 saying that lightdm drops privileges for security reasons so can't do a lot to remove the stalled file owned by root, it might not be lightdm that writes the file... [08:58] Launchpad bug 855773 in lightdm "Failing to write ~/.Xauthority" [Medium,Won't fix] https://launchpad.net/bugs/855773 [08:58] smspillaz: testing cpm meanwhile [08:59] seb128: lightdm coudl still remove it [08:59] cjwatson, pitti, mvo: well anyway I updated the bug settings, assigned to robert_ancell to get his opinion and commented on it [08:59] pitti, well not sure why robert_ancell wrote "I'm going to close this. I think lightdm can't do anything sane in this case" on that bug [08:59] pitti, will need to check with him [08:59] followed up to that bub [08:59] bug [08:59] pitti, danke [09:00] you don't need write permissions on the file in order to remove it -- that requires write permissions on the dir [09:00] unless the dir is sticky (like /tmp/) [09:00] seb128: it could move it away fr a start [09:01] right [09:03] didrocks: ok. I'm just going to test the build to make sure everything workse [09:03] smspillaz: great! thanks [09:03] (the delta is getting huge now) [09:06] it was already huge :) [09:07] mvo, btw I think you didn't read what I wrote yesterday, you left at the same time, I've assigned you a few software-properties bugs that might be worth looking at, feel free to unsubscribe you when you do, I just used it as a way to have the emails reach out if you filter assignees out of the launchpad spamming ;-) [09:08] seb128: yeah, I noticed this morning [09:08] mvo, ;-) [09:09] mvo, *whip cracking noises* :p [09:09] * seb128 hugs mvo [09:11] manager material that guy ;) [09:11] * ogra_ tries to imagine seb128 with a tie .... and fails miserably [09:11] mvo, "that guy", that's how you call me now? ;-) [09:11] his Yedi master rickspencer3 had strong whip cracking fu! [09:11] pitti, true story! ;-) [09:14] hmmm, might be time to restart my session already [09:14] didrocks: it's not perfect (still get warnings about unhandled configures) but you shouldn't see any bugs [09:14] my laptop feels like it's filled with glue [09:15] smspillaz: ok, let's have a try when you land it over to me :) [09:15] didrocks: the only bugs will come out of cases where a configure notify event was not matched and remains in the queue and that's not happening anymore [09:15] chrisccoulson, time to get a new laptop then? ;-) [09:15] didrocks: sure, I'll give you the patches now. Can I just mail them to you ? [09:15] oh you just did :p [09:15] lol [09:15] chrisccoulson, do you use unity-3d again btw? [09:16] smspillaz: sure, did you adapt the first one for jumping windows? [09:16] yes, I've adapted them both [09:16] smspillaz: if not, I just need the other one [09:16] seb128, yes, i've been using unity-3d again this week [09:16] smspillaz: ok, then, yes please, email [09:16] chrisccoulson, was it anything important you referred to above (wrt having talked to me about it) ? [09:16] smspillaz: there is a bug for the second one, isn't it? [09:16] ? [09:16] smspillaz: with a test case? (people on the french forum finds your test case quite cryptic to understand btw ;)) [09:17] didrocks: (so ... kind of fixed window shading and kind of fixed geometry updates ... the price we pay for having a huge diff) [09:17] didrocks: yeah, hang on [09:17] https://bugs.launchpad.net/compiz-core/+bug/864478 [09:17] Launchpad bug 864478 in unity "Window shading is broken" [High,Fix committed] [09:17] I still see rendering artefacts but at least the window isn't completely broken now [09:18] smspillaz: just add a test case to trigger it effectively [09:18] didrocks: I'm wondering if it might be more appropriate to have a tarball for the next sru, we've got like 18 distro patches now and the diff is starting to cause changes in the way the code functions [09:19] smspillaz: no tarball for compiz SRU, last time there was one, some distro-patch were reverted by it… [09:19] that was like 6 months ago :) [09:19] no patches that I've given you aren't in trunk now [09:19] ogra_, nothing important [09:19] smspillaz: no, that was… a month ago :) [09:19] we were discussing removing webfav from the archive, but i vaguely remember mentioning it before [09:19] i could be mistaken though [09:20] didrocks: I just don't want to get into a situation where people are encountering bugs on the SRU that I can't reprdouce using trunk because the distro patches have applied in a slightly different way than what's in trunk [09:20] didrocks: which is actually starting to become a problem [09:20] smspillaz: I retargeted the bug and add the team, just missing the test case [09:20] didrocks: perhaps I should just go over all the distro patches again and make sure that there's some sanity ... check the diff between trunk and what we've got etc [09:20] smspillaz: think about SRU team member who will have a 5000 diff lines content… [09:20] chrisccoulson, i cant remember that, but we use ubuntu-desktop now, so whatever is good on x86 (with the exception of mono apps) is also good on arm [09:21] ogra_, excellent. we removed it already anyway ;) [09:21] smspillaz: we will need to kill your "distro patch in trunk", indeed, but that's for precise I guess [09:21] heh [09:21] didrocks: well, the only reason it's like that is because a distro patch is required for unity to run (which is no good) [09:21] rickspencer3, "?" to you too ;-) [09:22] and also because I've modified some plugins in certain ways that definitely wouldn't make sense upstream [09:22] and the upstream "maintainer" for the grid plugin apparantly hates me and won't accept my changes anymore [09:22] which is why I like this model better :) [09:23] smspillaz: we should let distro-patches, as distro-patches [09:23] didrocks: the problem with that is it makes it impossible to develop unity on compiz trunk [09:23] not merge in some kind of half upstream-trunk [09:24] smspillaz: we will speak about it, but everytime there is a change done, I spotted some revert on our distro-patch [09:24] and I don't want that anymore [09:24] and you've also got things like unity-window-decorator etc which are impossible to maintain as patches [09:24] rodrigo_, there? I need somebody not under unity to test something ;-) [09:25] didrocks: sure, well, maybe we should see how much distro stuff we can kill and just go back to a "regular" trunk [09:25] smspillaz: that's what I proposed you last cycle :p [09:25] but at the moment, it's sort of like, last cycle we ended up forking compiz because of distro requirements and then this cycle we did it again [09:25] like the "add ccp" on --replace without any value [09:26] didrocks: there's a better solution to that. Merge compizconfig into core (whcih is what I am doing next cycle) [09:26] then there's also the mass reverts of stuff that we did which makes the diff unmaintainable [09:27] anyways, the point that I'm trying to get at is that both methods suck, but I'd like to avoid a situation where we've got distro patches which actually fundamentally change the way things work so that people get bugs on them and then I have to maintain our patches and trunk [09:27] smspillaz: yeah, but that's a discussion we had 8 month ago [09:28] smspillaz: so better to see solutions right *now* :) [09:28] of course [09:28] well maybe next cycle we should drop all of the patches and disto stuff and return to a clean trunk again and then rebuild the distro stuff from there [09:29] being really careful to minimize the diff as much as possible :) [09:30] if we can avoid having a ~compiz-team/compiz-*/precise and avoid having very many distro patches at all that would be a massive win [09:30] and I'm willing to implement distro stuff "properly" upstream in a way that makes sense upstream in order to avoid that [09:31] smspillaz: that's what I did for 10.10, took me two weeks [09:31] that was garbaged because you removed it to integrate in trunk [09:31] and that's when I asked you to review them :) [09:31] that was so I knew what was going on :) [09:32] the problem as I see it is that the more "distro" stuff you have, the harder it becomes to maintain, so lets try and avoid it period next cycle [09:32] smspillaz: I agree, but some of things doesn't integrate upstream for no good reason [09:33] but anyway, waiting for your patch :) [09:33] I recognize that my approach to have a "distro branch" didn't quite work the way we wanted it to, but I also know from hard experience that having lots of distro patches made my life really difficult in 11.04 [09:33] so what I'm saying is - lets avoid both and do this the Right Way [tm] and get it all upstream, even if it requires more reworking [09:33] seb128, I am off today (national holiday), but what do you want me to test? [09:33] didrocks: yes, patches [09:34] rodrigo_, slacker! :-P [09:34] rodrigo_, oh, don't bother then, that can wait tomorrow (and I tested in unity-2d meanwhile), enjoy your day off ;-) [09:36] didrocks: the biggest problem that occurred this cycle as I see it was that a distro patch got required for unity to run, which pretty much meant that I *had* do to the oneiric branches. So it's going to be a really imporant priority to make it so that unity can run off of trunk compiz again [09:36] (and no, that patch is not acceptable upstream) [09:37] didrocks: (patches sent) [09:37] chrisccoulson, :) [09:37] seb128, I am checking mail right now, so I can do it now if you want [09:37] didrocks: the good news is that most of the distro stuff we have in now we should just be able to drop [09:38] rodrigo_, the keyboard capplet, keybinding, can you assign those by using the key as described at the bottom or do you need to double click the lines to put them in "edit" mode? [09:38] smspillaz: not the settings change and we need the gtk_init fix as well [09:39] rodrigo_, i.e if you select a line and do shift-Z does it assign the keybinding or do nothing until you click to have "new keybinding..." written on the line, then you can press the keys [09:39] didrocks: like, the settings defaults can be implemented in settings upgrade files, the --replace ccp thing will go away once we have compizconfig in core, the gtk_init thing can die once I do the require stuff in unity to make it die, and the settings transition can be made into a more generic module I think [09:39] and hopefully, just hopefully, we can upstream the metacity patches and kill unity-window-decorator [09:39] seb128, yeah,you have to click on the keybinding column to have the 'new keybinding' show up [09:39] rodrigo_, ok, thanks, I will open a bug [09:40] rodrigo_, it's not what the label at the bottom tell you to do, that's confusing ;-) [09:40] didrocks: the number one thing I want to die is gtk_init inside of compiz. it's just wrong on so many levels :) [09:40] seb128, yeah, right [09:40] seb128, assign the bug to me, I'll have a look tomorrow [09:40] rodrigo_, I will just open, let a chance to bastien to fix it for you :p [09:40] if he doesn't you can still fix it tomorrow ;-) [09:41] smspillaz: the settings default can't be implemented in settings upgrade. Think about new installs… [09:41] seb128, ok, good plan :) [09:41] didrocks: they get run as soon as compiz runs [09:41] smspillaz: seems totally wrong for a boot speed process [09:41] didrocks: only happens once [09:42] smspillaz: still, the first boot time count [09:42] smspillaz: we can do that in a gsettings override and forget about other backend [09:42] so this can be fixed this way [09:43] sure [09:43] smspillaz: anyway, it won't be fixed today, can you rather add a testcase as I asked you twice already, please? :) [09:43] sure [09:44] didrocks: done [09:44] smspillaz: thanks [09:54] smspillaz: ok, your change looks good, will upload soonish in -proposed [09:54] didrocks: cheers [09:54] smspillaz: you didn't add steps to reproduce on bug #860646, can you fix this? [09:54] Launchpad bug 860646 in compiz-plugins-main "snap movements can cause infinite loops in window movements" [Undecided,Fix committed] https://launchpad.net/bugs/860646 [09:55] volonteers to confirm bug fixes need clear instructions :) [09:55] hmm, thought I just updated that one [09:56] apparantly not [09:57] done [09:58] thanks! [10:29] pitti: compiz and cpm in -proposed FYI [11:34] * mpt wonders why Apport doesn't come up for Compiz crashes any more [11:34] mpt, maybe it's disabled because we're close to release [11:34] i think you can enable it again manually [11:37] yes, disabled since sunday [11:54] mpt, edit /etc/default/apport or run the command indicated in there [11:55] seb128, oh, it wasn't my computer, someone else's with the beta installed ... but thanks :-) [11:57] mpt, yw ;-) [11:57] mpt, still fwd the message if you can, we want to know about segfaults still happening [11:59] i don't seem to be able to escape adverts for chrome [11:59] even http://whatsmyuseragent.com/ has a chrome advert right in the middle of the page [12:09] mpt, because compiz has stopped crashing and you're just imagining it now [12:11] he's blinking :-) [12:12] actually, since i started using 3d again this week, compiz hasn't crashed yet :) [12:12] * chrisccoulson keeps fingers crossed === m_conley_away is now known as m_conley === eeejay_is_afk is now known as eeejay [15:03] wow...proposed compiz/unity (whichever one it is) really feels buggy! [15:04] didrocks and smspillaz, there seem to be a bunch of things with it...what are we doing with proposed compiz? [15:04] jasoncwarner_: the new compiz in proposed for a few horus? [15:04] hours* [15:04] tremolux, I get a Wunderlist banner, but clicking it says "not found" [15:04] didrocks: not sure, then...updated this morning...checking now [15:05] tremolux, though that "not found" page has 3 ratings...? [15:05] jasoncwarner_: can you apt-cache policy compiz? [15:05] jasoncwarner_: was discussed briefly in -devel; it's being tagged verification-failed [15:05] jasoncwarner_: hum, no the other one isn't pushed yet [15:05] jasoncwarner_, how buggy? [15:05] jasoncwarner_: what issues do you get apart from the jumping windows? I have it for 2 days and don't get any other issue [15:05] compiz: [15:05] Installed: 1:0.9.6+bzr20110929-0ubuntu4 [15:05] Candidate: 1:0.9.6+bzr20110929-0ubuntu4 [15:05] Version table: [15:06] jasoncwarner_: yeah, so previous version (the other one didn't hit -proposed yet) [15:06] jasoncwarner_: what do you experience, apart from the jumpy windows? [15:06] didrocks: window controls missing from super+s view? [15:06] jasoncwarner_: people on the french forum are testing it since it's in proposed as well, and they are almost putting some pressure so that it's in finale, what I'm telling no way ;) and just discovered the jumpy windows as well as a regression [15:06] jasoncwarner_: this is already in previous version [15:06] mterry: hmm, yeah..I'm not working on that but I'll check with the folks who are [15:08] jasoncwarner_: do you experience any other issue? [15:09] jasoncwarner_: the jumpy window fix are waiting for a SRU member to ack it in -proposed, you should get it tomorrow at least === zyga is now known as zyga-afk [15:42] hi, during upgrade of my laptop to oneiric nm stopped being able to connect to wifi and upgrade failed. Is this known to anybody? [15:42] ugh, wrong channel, sorry [15:55] good night everyone [16:00] good night pitti [16:10] good night pitti [16:20] 'night pitti [16:24] rye: failed because you couldn't finish downloading the files? [16:36] chrisccoulson, if you get some time for desktop issues could you maybe have a look to bug #857168 [16:36] Launchpad bug 857168 in gnome-control-center "totem doesn't inhibit the screen blanking as it should" [Low,Confirmed] https://launchpad.net/bugs/857168 [16:36] ? [16:36] seb128, sure, can do [16:36] that's broken again? [16:36] chrisccoulson, no hurry, it's only one bug so far so I'm not sure if that's really broken but it seems worth investigating, and I think you have a bit of experience with how inhibit works [16:37] chrisccoulson, well I didn't confirm but I don't watch movies often on my computer nowadays, I tend to use the TV :p [16:41] cyphermox, bug #805311, vuntz added a patch to bugzilla which made be worth sru-ing [16:41] Launchpad bug 805311 in gnome-bluetooth "gnome-control-center crashed with SIGSEGV in g_type_instance_get_private() (device_services_changed)" [High,Triaged] https://launchpad.net/bugs/805311 [16:42] Whoever came up with the idea about pop-up confirmation dialog when you hit power button will get a round of beers on me at the next UDS. [16:42] Or whatever the drinking preference of theirs... :) [16:45] Hey, I get a weird behavior of gedit on a fresh installation of latest 11.10 builds, launching gedit from a terminal does nothing and the command just hangs [16:46] is it known ? [16:50] seb128: cool, I was still messing around with that but tbh without much luck :/ [16:50] cyphermox, well vuntz had luck it seems ;-) [16:50] I can sense a facepalm moment in my near future [16:50] jibel, no [16:50] jibel, can you strace it? [16:50] jibel, where do it stop? [16:51] BigWhale, ;-) [16:52] seb128: ah, I see. well, not as life-changing as I had expected, but that fix would work :) [16:52] preparing the sru [16:52] seb128, unfortunately you'll have to wait for the next European UDS. I forgot to apply for sponsorship and blew my chances to be in Orlando. :/ [16:53] BigWhale, no worry that will work as well ;-) [16:53] Orlando is a bit too far to go there on my own expenses. Hungary was within the driving distance. :> [16:54] have a good night everyone [16:54] seb128, http://people.canonical.com/~j-lallement/gedit_strace.jpg [16:54] seb128, reproduced on 2 installs. now trying on live session === zyga-afk is now known as zyga [16:56] jibel, if you run it a second time from somewhere does it open or not? [16:59] seb128, no, it is reproducible and hangs each time. 2 different installations [16:59] it worked with previous images [16:59] jibel, like if you run "gedit &" and then "gedit" [16:59] it still hang? [16:59] what about dbus-launch gedit? [17:00] seems like a dbus screwup, I wonder if that has to do with the issue cjwatson pointed today they had to fix [17:00] seb128, running it twice with 1rst instance in background work. 2nd instance comes up [17:01] ok [17:02] jibel, seems similar to bug #827414 [17:02] Launchpad bug 827414 in unity-lens-files "gedit fails to start on first try, but does on the second" [High,Fix released] https://launchpad.net/bugs/827414 [17:02] or bug #819447 [17:02] Launchpad bug 819447 in unity "[Oneiric] Unity launcher and dash opens gedit only on second attempt" [Undecided,Invalid] https://launchpad.net/bugs/819447 [17:02] jibel, env | grep DBUS [17:06] seb128, I'm reinstalling on a physical machine with good network connection. It will be easier than a networkless german VM to gather information. one moment. [17:06] jibel, ok [17:13] seb128, why is MC version 1:5.9.1-0ubuntu2 ? [17:13] what does the 1: mean ? [17:15] xclaesse, it's the debian way to say that a lower version is actually highe [17:15] higher [17:15] xclaesse, it's likely that somebody in debian used it because they uploaded a new serie in unstable by error and wanted to roll back to the previous one [17:16] xclaesse, since "updates" go by increasing versions the only way to do that is to add this number [17:18] right that's like that in debian too, ok :) [17:18] thanks [17:31] seb128, reproduced on hardware, it affects only unity-2d === allison_ is now known as wendar [17:43] jibel, ok, and it started today? [17:43] jibel, what gives the env | grep DBUS? [17:45] seb128, hold on, keymap doesn't map keyboard layout :( weirder and weirder, I can reproduce on livesession BTW [17:47] seb128, http://paste.ubuntu.com/706873/ [17:48] jibel, does unsetting DBUS_STARTER_ADDRESS workaround it? [17:48] seb128, yes [17:49] ok, some similar issue to the lens bug I pointed [17:49] could you boot an iso from yesterday and see if that was set? [17:49] or from this morning [17:49] or if that's a side effect of the respin they did today [17:49] I think it only impact visibly on gedit btw, but still, it's an issue [17:51] we can't work out how it's a side-effect of our respin [17:51] since it happens after you've generated a machine-id and then rebooted [17:53] is DBUS_STARTER_ADDRESS supposed to be unset, or is it supposed to have a different value? [17:54] it's apparently set for basically everything from gnome-session down [17:58] cjwatson, it's unset for me on my oneiric (not updated today) [17:58] $ strings /proc/$(pidof compiz)/environ | grep DBUS_STARTER [17:58] $ [17:58] $ strings /proc/$(pidof nautilus)/environ | grep DBUS_STARTER [17:58] $ [17:58] etc [17:58] let me check, I don't know what DBUS_STARTER_ADDRESS is for [17:59] mterry, ^ do you have an idea? it's you who pointed the issue for the unity lens when that was set [17:59] seb128, yeah, it's set when something is auto-started via dbus [17:59] seb128, only DBUS_SESSION_BUS_ADDRESS was set on previous build [17:59] ok [18:00] mterry, do you know what could lead to have it set for the session process since today? [18:01] is it possible that the lightdm upstart job is racing the dbus job? [18:01] seb128, something calling a gnome-session dbus api before it is started manually [18:01] yeah, possibly a race [18:01] except that doesn't make sense since it happens after a reboot too! [18:01] after the machine-id has been created [18:01] what on earth else is persistent? [18:03] hum, good question... [18:04] not sure why that started today, jibel says the previous build didn't have the issue [18:04] now he gets it every time [18:05] indeed; I agree it's compelling to suggest that it's related to the respin, I just can't see *how* [18:05] yeah, me neither :-( [18:05] going slowly insane [18:15] Hi, I was recommended to go here with an issue I have in Oneiric: Unity doesn't seem to obey my compiz config settings. Is some unity startup script or something overriding these settings? It seems to work fine if I restart compiz once well logged into my Unity session. [18:15] seb128: stgraber has reproduced it ten times in a row with yesterday's image [18:15] on a machine that boots really really quickly [18:16] so at worst, all this respin has done is widen a race slightly somehow [18:24] evening [18:24] what's broken? (sorry, not read the whole backscroll yet) [18:29] gedit fails to launch the first time due to DBUS_STARTER_ADDRESS being set for reasons unknown [18:30] frequency apparently increased after recent respin but we can reproduce it before that too [18:31] cjwatson, is that only in the 2d session? [18:32] somebody here reproduced it from startx too [18:32] I don't believe we can reproduce it from unity 3d at present [18:32] It should be reproducible from almost anything that isn't unity-3d, I suspect. [18:35] chrisccoulson, jibel says he gets it under 2d only [18:35] but they he said he gets it on the livecd as well [18:35] cjwatson, ok [18:35] i'll switch back to 2d in a moment :) [18:35] cjwatson, well it only breaks gedit so it's not the end of the world, still annoying... [18:37] and even that only the first time [18:37] cjwatson, oh, only the first time? I didn't understand that from what jibel said when he said he was getting it on the installed system [18:38] chrisccoulson, what is broken with 3d? [18:38] the first time, it detects that it's being auto-started and backgrounds itself to start as a service [18:38] the second time, it connects to the previously started service [18:38] (so I'm told) [18:38] seb128, nothing (other than the nauseating blurry icons in the dash). i was going to switch to 2d to see if i get this problem with gedit [18:39] oh ok [18:39] cyphermox, re: failed because could not download files - well, it failed to download flashplugin [18:40] cyphermox, by the time it came to flashplugin download it dropped the network connection and does not want to reconnect to wifi using the applet [18:41] cyphermox, so it affects people who have flashplugin installed && connected on wifi && upgrade from natty [18:41] oh, of course. i can switch sessions whilst my laptop is updating, thanks to screen :) [18:41] brb [18:43] bryceh, hey, btw gnome-desktop is gnome-desktop3 nowadays ;-) [18:43] bryceh, gnome-desktop is the old gnome2 version [18:43] ok, got the problem with gedit straight away [18:44] at least it's reproducible :) [18:44] seb128, ah ok, thanks for the reminder [18:44] bryceh, yw [19:09] seb128, ok, the gedit issue is because the fix for bug 827414 doesn't do anything [19:09] Launchpad bug 827414 in unity-lens-files "gedit fails to start on first try, but does on the second" [High,Fix released] https://launchpad.net/bugs/827414 [19:09] chrisccoulson, ? [19:10] that fix removes DBUS_STARTER_ADDRESS from the places daemons, but it's actually unity-2d-places which spawns gedit when you click on it in the dash [19:10] chrisccoulson, it fixed the bug in 3d [19:10] seb128, i guess in 3d that it's unity which spawns gedit (which isn't dbus activated) [19:10] but in 2d, it's actually unity-2d-places [19:10] ok [19:10] which is dbus activated [19:11] so unity-2d issue ;-) [19:11] yes [19:11] great [19:11] so i don't think it's related to any changes cjwatson made to today [19:12] bug 873027 [19:12] Launchpad bug 873027 in unity-lens-applications "DBUS_STARTER_ADDRESS and DBUS_STARTER_BUS_TYPE aren't always unset from environment making gedit and possibly others fail to start" [Undecided,New] https://launchpad.net/bugs/873027 [19:12] that's basically where we are at after looking on a few systems here [19:13] stgraber, the issue isn't with unity-lens-applications (which does clear those variables from it's environment already) [19:13] the issue is with the actual dash in 2d, which is what is really spawning your applications [19:13] IIUC ;) [19:13] if you kill unity-2d-places and start it manually from a console, then everything will magically work correctly :) [19:14] so i'm pretty sure that's where the issue is [19:14] and that explains why this isn't a problem in the 3d session [19:15] ok; is the startx case entirely different then? [19:15] not sure, i've not tried that [19:15] ah, never mind, that's still unity-2d [19:15] it was just ruling out lightdm; I misunderstood what stgraber was doing [19:19] ok, just rebuilding unity-2d now [19:22] is it expected that the indicator menues auto-close when I don't hold the mouse button while selecting an entry? (unity-2d) [19:23] geser - multiple screens? [19:23] chrisccoulson: yep, 2 screens [19:23] geser, bug 869196 [19:23] Launchpad bug 869196 in unity-2d "unity panel menus don't stay open when clicked on second monitor with different geometry" [Undecided,In progress] https://launchpad.net/bugs/869196 [19:25] exactly my problem, thanks [19:26] how long does unity-2d normally take to build? [19:26] hopefully not as long as firefox ;) [19:27] "took 12 minutes, 32.9 seconds" (the i386 build of unity-2d) [19:28] yeah, it took around 3 minutes or so here [19:28] brb, session restart === zyga is now known as zyga-brb [19:30] seb128, stgraber, ok, that works :) [19:31] cool [19:32] chrisccoulson: Awesome. Any cleanup needed on that, or is it suitable to push it to -proposed for us? [19:32] infinity, it will be fine for proposed [19:32] chrisccoulson: Then fire away. ;) [19:33] i'll bet money that i can't upload it ;) [19:33] I bet I can. [19:33] But it's fine if you wait for a DXy person to review and upload too. [19:34] oh [19:34] i can :) [19:34] I'd just like to know this has a fix on the way so I can stop caring. :P [19:34] Archive Upload Rights for ubuntu-desktop: archive 'primary', package set 'ubuntu-desktop' in oneiric [19:34] nice :) [19:40] does unity-2d have a single branch for upstream and packaging? [19:49] has someone an idea how I've broken my unity-2d dash that it doesn't list any application or find anything with the search bar? === chrisccoulson_ is now known as chrisccoulson === m_conley is now known as m_conley_away [21:16] Amaranth: hi! you have work item in https://launchpad.net/ubuntu/+spec/security-o-compiz-screenlocking: "[amaranth] implement gnome-shell style screen locking in compiz plugin". whatever happened to this? is it something that is still valid for precise? [21:16] jdstrand: hehe, was wishful thinking that I'd manage to do it [21:16] I think smspillaz is planning on doing it eventually [21:17] ok, I'll mark it postponed for now then [21:17] Amaranth: thanks [21:17] smspillaz is only allowed to work on fixing stacking bugs isn't he? [21:17] ;) [21:18] that is actually a stacking bug of sorts [21:18] the reason why it came up is that the screensaver would be behind other things [21:18] s/would be/would sometimes be/ [21:19] (you can see that in the whiteboard) [22:55] jml, question from upstream on the glitch behavior [22:55] jml, https://bugs.freedesktop.org/show_bug.cgi?id=41733 [22:55] Freedesktop bug 41733 in Driver/intel "Screen jitters every so often, especially when laptop under load" [Normal,New]