[00:28] desrt: we don't grab all possible keycombos [00:29] desrt: we only grab keycodes that we have actions bound do [00:30] unfortunately, XGrabKey on a modifier key has some nasty side effects. namely that the grab activates when you press the modifier key and doesn't deactivate until its released [00:30] "we don't grab all possible keycombos yet" << fixed it for you ;) [00:31] desrt: so the reality is that the grab will remain when you press another key in addition to an already grabbed key [00:31] smspillaz: i have a workaround for that [00:31] that being said in compiz we use some trickery to ensure that the grab is released when you do that [00:31] when you get the first unexpected key you XAllowEvents ReplayKeyboard and drop the grab [00:31] yep [00:32] then the app that would have seen it in the first place gets to see it [00:32] yep [00:32] with the original timestamp, even [00:32] it ends up just looking like a processing delay [00:32] there is a race condition though [00:32] if you use Sync grabs there shouldn't be... [00:32] where the passive grab activates [00:32] the server is supposed to freeze the queue [00:32] hmm maybe that's been fixed [00:32] I'm not really the one who works on it [00:33] i have some sample code that i'm going to try to hammer into something meaningful over the next day or two [00:33] reviewing ted's dbusmenu work just now though [00:34] desrt: ok, we're using GrabModeSync for keyboard mode [00:34] in XGrabKey [00:34] that makes sense [00:34] should be fine, then [00:34] the x server freezes the keyboard event queue until you tell it what to do with XAllowEvents [00:34] you only see the very first keystroke [00:34] indeed [00:35] I hate grabs [00:35] (ie: the one that caused the establishment of the passive grab) [00:35] the X grabbing infrastructure is actually pretty sensible if you understand it [00:35] its sensible but like [00:35] my biggest gripe with it is that it chooses to solve a problem that's entirely too complicated [00:35] just the entire concept [00:35] and some different engineering/design choices could have resulted in a *much* easier to understand system [00:36] indeed [00:36] desrt: oh I found a funny problem with gnome-keyring the other day [00:36] speaking of grabs [00:36] desrt: it was a bug that happened in compiz and [00:36] *gasp* [00:36] gnome-shell but not [00:36] *gaps* [00:36] mutter [00:37] because gnome-keyring would establish a grab [00:37] on its input window [00:37] time to port unity-2d to mutter, clearly [00:37] errm [00:37] ah, well [00:37] s/mutter/metacity/ [00:37] well [00:37] no sorry [00:37] you should know that that's not a bug that happens in gnome-shell either [00:37] it would establish teh grab [00:37] it happens in canonical's patched version of gnome-shell [00:37] when the window became unfocused [00:37] desrt: only happens because [00:38] gnome-shell and compiz support _NET_WM_STATE_FOCUSED [00:38] and there was this bitshift bug [00:38] XD [00:38] drove me a bit nuts [00:38] spent ages thinking [00:38] "there has to be something we're doing wrong" [00:38] then it broke in g-s and then I instantly thought "nope, application must be broken" [00:40] its always hard to tell these things unfortunately, at the boundary of the application and wm, who is actually doing weird things [01:02] desrt: are we stuck with gnome-control-center not being able to write gsettings keyboard shortcuts for precise? [01:17] jbicha: unless you want to run the upstream ppa [01:19] ok, just checking [01:22] desrt: now I have to figure out whether to make the gnome-shell keyboard shortcuts match unity's new ones from bug 969235 [01:22] Launchpad bug 969235 in metacity "Keyboard shortcuts - Update some window management shortcuts to use "Ctrl + Super" instead of just "Super" " [Undecided,Fix released] https://launchpad.net/bugs/969235 [01:23] it's that whole "do people want upstream or do they want better integrated with Ubuntu" question again [01:24] smspillaz: hey, do you have a moment? [01:24] how long is a moment ? [01:25] smspillaz: could you take a look at bug 970932? I wonder if it's compiz related issue [01:25] Launchpad bug 970932 in unity "unity treats windows as belonging to multiple workspaces at the same time" [Undecided,New] https://launchpad.net/bugs/970932 [01:25] kklimonda: you need to get it added to the priority list [01:25] talk to either didrocks or seb128 [01:26] sorry, I get told off if I work on stuff that's not on the priority list [01:26] (http://wiki.ubuntu.com/Unity/PrecisePriorities [01:27] wow, this list is big for that close to the release :( [01:27] which is why getting new stuff on to it requires a process :) [01:32] damn beaurecracy, seb128 definitely told me to ping you. I'll ask him tomorrow to get it on the list ;) === _thumper_ is now known as thumper [04:36] Good Moring. [04:36] moring [04:36] whatever. [04:49] BigWhale: Good mooring :) [04:50] RAOF, I'm just pretending to be awake. :) [05:23] pitti: Just to make sure I understand you completely re at-spi2-atk and pyatspi and transitional packages, could you please tell me whether what I've done in these diffs is correct to solve the problem? Note other changes from debian are included as well: http://people.canonical.com/~themuso/at-spi2-atk.diff and http://people.canonical.com/~themuso/pyatspi.diff. [05:51] Good morning [05:51] TheMuso: hey [05:51] TheMuso: not quite -- you need to make the C/R/P versioned, not drop them completely [05:52] TheMuso: i. e. (<< your_new_version_that_introduces_the_transitionals) [05:54] pitti: ah right, makes sense, thanks. [05:56] shouldn't that be B/R/P since it's versioned? [05:58] Not sure about now for the transitional stuff, but these newer packages contain the same filenames and paths as the older at-spi stack packages. [05:58] Hense why they currently have conflicts. [05:58] TheMuso: that should be breaks for file moves [05:58] There are no file moves. [05:58] 2 packages have the same paths/files. [05:58] TheMuso: oh, ok [06:04] TheMuso: "newer packages" == the at-spi2 stuff, you mean? [06:04] pitti: yes [06:04] TheMuso: becasue the transitional at-spi and python-pyatspi packages should be empty (except changelog, etc.) [06:05] pitti: and I was referring to libatk-adaptor/python-pyatspi2 having the same filenames as the at-spi and python-pyatspi packages that we currently have in the archive. [06:07] TheMuso: right, hence we need the B/R/P [06:07] right [06:11] good morning [06:11] bonjour didrocks [06:11] guten morgen pitti, how are you? [06:19] pitti: I thought you couldn't have versioned provides... [06:19] TheMuso: right, just versioned B/R I mean; sorry [06:19] Right, just checking. [06:19] didrocks: quite fine, thanks! [06:19] didrocks: how about yourself? [06:34] pitti: seems everything's fine. Answering emails and the world didn't fall apart. Also, the QA labs is back on shape :) [06:35] ooh, is it? [06:35] indeed, nice! [06:35] pitti: was blocking all merge requests yesterday, wasn't fun ;) [06:44] pitti: Ok, if you wouldn't mind having a look at those above linked diffs once again just to make sure I've not msised anything, that would be appreciated. If all looks ok, I'll upload. [06:45] TheMuso: not (<< ${binary:Version}) please [06:45] TheMuso: just whichever version you are about to upload [06:45] TheMuso: it does not contain a changelog, so I can't say [06:46] TheMuso: that (fixed, static) version is from which on at-spi will be transitional; it does not change any more [06:50] ah ok now I'm with you. [06:54] pitti: Ok please check again. I don't want to have to upload and have it not work. :) [06:55] TheMuso: looks fine now; you should just add (LP: #966845) to the changleog [06:55] TheMuso: thanks! [06:57] that was the plan, I just want to get the important bits right. :) [06:57] s/was/is/ [07:00] Ok uploaded, and hopefully fixed. [07:00] didrocks: do you think it would be ok to move the privacy settings from "system" to the "personal" category in control-center? [07:01] didrocks: I think that would be appropriate, and fix the UI overflow introduced by the new landscape applet [07:01] didrocks: I didn't find an existing bug, will create one, but wanted to get your thoughts first [07:01] pitti: would make sense I guess. Let's do it then! [07:01] pitti: do you want me to change this? [07:02] didrocks: hang on, need an UIFE bug and some signoff first [07:02] you have an overlow on de? [07:02] I don't have any in fr [07:04] didrocks: bug 973130, has a screenshot [07:04] Launchpad bug 973130 in gnome-control-center "[UIFE] Adding landscape installer caused UI overflow" [Undecided,New] https://launchpad.net/bugs/973130 [07:04] didrocks: odd, what is missing for you in the "system" category? [07:05] pitti: ah, it's because of landscape. It seems I don't have installed [07:06] I guess we need to wait for jbicha to come online [07:06] right [07:21] pitti and didrocks on the landscape thing, I upgraded a test system today and I didn't get landscape in the system settings. I have it on my two main systems, but not sure why I didn't get it on the test system [07:25] landscape-client-ui-install is seeded and a recommends of ubuntu-desktop [07:26] I didn't get it on upgrade as well [07:29] strange; ordinarily updates should get new recommends [07:29] ah, I got the new package on 2012-03-23 [07:29] with an apt-get upgrade [07:29] (well new "ubuntu-desktop") [07:29] that's why the new recommends wasn't installed, as it wasn't in a dist-upgrade [07:30] I tend to apt-get upgrade first [07:30] then, apt-get dist-upgrade and look the output [07:30] thinking of it, it's an issue with all the new recommends, like, in this case, you get the package recommending a new one on upgrade and not dist-upgrade [07:31] hmm...so that means we wouldn't get it if we did just an upgrade for that particular one? [07:34] sudo apt-get install --fix-policy [07:34] jasoncwarner_, didrocks: that will install all the recommends you skipped during "upgrade" [07:35] pitti: already did that and confirms it did it :) [07:48] hey [07:48] it's a seb! [07:48] aisleriot FTBFS fix uploaded [07:53] salut seb128 :) [07:53] hey pitti didrocks, how are you? [07:53] seb128: quite fine, thanks! [07:53] seb128: I'm fine, thanks, you? [07:53] seb128: would you be able to create a desktop team status page tomorrow evening, and send it to u-release@? [07:57] didrocks, I'm good thanks (busy reading @tech discussions) [07:57] pitti, yes, I can do that, I don't work friday by tomorrow night is fine ;-) [08:12] seb128: smspillaz told me that my bug has to go to the http://wiki.ubuntu.com/Unity/PrecisePriorities before he can take a look at it ;) [08:12] kklimonda, ok, fat chance then [08:12] kklimonda, that's really a corner case issue for multiple workspaces users and you are the first to complain, I doubt it lands on any priority list ;-) [08:15] seb128: huh, am I really the only person to use super+left with multiple workspaces? because that's how I hit the bug :( [08:15] (well, that's one of the ways) [08:16] pitti, bug #973130 I reassigned to the activity-log-manager, nothing we can do from the control center side [08:16] Launchpad bug 973130 in activity-log-manager "[UIFE] Adding landscape installer caused UI overflow" [Medium,New] https://launchpad.net/bugs/973130 [08:17] kklimonda, can you describe your bug? using super-left doesn't make the left pinned dialog show on the next ws or in its alt tab here [08:17] seb128: yeah, I think I was describing two bugs at the time [08:17] seb128: I've went to rewrite the report yesterday evening [08:18] kklimonda, what's the number again? [08:18] https://launchpad.net/bugs/970932 [08:18] Launchpad bug 970932 in unity "unity moves windows between workspaces when you try switching to them" [Undecided,New] [08:20] kklimonda, I can't confirm, I did that: [08:20] - open gedit on ws [08:20] - open nautilus on ws2 [08:20] - go to ws1, focus gedit, super-left [08:20] click on nautilus in the change, it moves me to ws2 [08:20] click on something running ws1 [08:20] it brings me back to ws1 [08:21] gedit didn't move nor anything else [08:21] seb128: do you use super+left to move gedit to the left part of the screen? [08:22] kklimonda, " - go to ws1, focus gedit, super-left" [08:22] ah [08:22] sorry, still drank my first coffee yet ;) [08:23] kklimonda, you will need to come with as exact sequence to reproduce the bug on a guest session [08:23] your description is not good enough, or it's specific to something on your config [08:31] seb128: is your launcher set to hide? I've written steps I took to reproduce it in the guest session in the bug 970932 [08:31] Launchpad bug 970932 in unity "unity moves windows between workspaces when you try switching to them" [Undecided,New] https://launchpad.net/bugs/970932 [08:31] kklimonda, no, it's always on (which is the default) [08:35] kklimonda, no, I can't reproduce in a guest session using your steps but I use unity staging ppa so maybe it's already fixed in trunk [08:37] ok, I'll update to the staging and see if it's fixed for me :) [08:43] micahg: I cannot exclude the possibility, that they might also work with breaks. But if that is changed, it would need to be changed at debian to keep the delta small. And debian is less than convinced of giving this a try. [08:43] micahg: And conflicts->breaks would be of theortical gain only as we already have conflicts to 3.4 for essential libreoffice packages inplace, so the limitations on the upgrade order are already inplace. [08:44] micahg: Also there is a good chance that conflicts: are actually needed and breaks: isnt enough. The whole layout of the libreoffice binaries (libraries paths etc.) got simplified a lot between 3.4 and 3.5 (moving libs around) and there are quite a few maintainer scripts to cope with that. So lowering the requirement to Breaks: might introduce problems in cornercases -- something I dont want to risk this short before the release. [08:53] damn, still happening on a guest account.. I'll test it on a netbook later, maybe it has something to do with nvidia driver (which is a very long shot) :/ [09:28] didrocks: do you know about a bug report that opening dash and typing does not type into teh dash, but the nautilus desktop readahead search? [09:29] pitti: not that I noticed. Is that the first time you open the dash? [09:29] or everytime? [09:30] didrocks: hm, could be [09:30] I don't use it that often [09:30] pitti: as the first time you open it, there is the lenses loading, that can impact some race on focusing maybe? [09:53] didrocks: re (sorry, had 1-on-1) [09:53] no worry :) [09:54] didrocks: so, it's plausible that it only happens the first time [09:54] didrocks: anyway, seems you aren't aware of it, so I'll just file it if I can reproduce it more reliably [09:54] pitti: I restarted my session and didn't get it, but if it's racy, can be related to computer speed [09:54] yes please :) [10:27] pitti / didrocks clean boot on my machine. pressed "Super" then typed "hello". The "h" is on my desktop in a nautilus thing in the bottom right, the "ello" is in the dash [10:27] gord: ^ [10:27] machine was idle, some time after boot up, core i7, fast ssd, so not a 'slow' machine [10:27] gord loves this kind of focus dealing, he did it for the HUD and dash already :) [10:28] haha [10:28] tbh I was pretty quick at typing 'hello', but it is certainly possible to trigger it [10:29] popey: also please file a bug and add that to the bug list we are waiting for (upstream tests regressions) [10:29] mostly does it the first time, less so if you "super"+wait a fraction of a second+ type [10:29] popey: can you cook that list with sil2100's results, please? [10:29] I can't today, on holiday [10:29] about to go afk all day [10:29] popey: ok, I'll ask sil2100 directly then :) [10:29] please do! [10:49] grmbl. update-manager ignored my ppa during the upgrade. why? [10:54] didrocks: we got an ack for bug 973130 [10:54] Launchpad bug 973130 in activity-log-manager "[UIFE] Adding landscape installer caused UI overflow" [Medium,New] https://launchpad.net/bugs/973130 [10:54] didrocks: do you know what to change, and want to/can handle it? [10:54] pitti: can do, sure :) [10:54] will report it upstream then [10:55] didrocks: merci beaucoup [10:55] pitti: mais de rien ;) [10:55] you were quick to assign it (just got spamm by the email) ;) [10:55] so long, need to leave [10:55] ttyl pitti! [10:56] working offline in the train this afternoon, and holiday tomorrow [10:56] see you all next Tuesday, enjoy your Easter holidays! [10:56] pitti: oh, enjoy your long week-end then ;) [11:00] pitti: what was the way to simulate a do release upgrade with a ppa enabled all the time? The AllowThirdParty=yes doesnt seem to cut it [11:00] Sweetshark: 12:57:33 <-- | pitti (~pitti@213.9.93.70) a quitté (Quit: ZNC - http://znc.sourceforge.net) [11:00] Sweetshark: too late ;) [11:02] *grmbl* [11:02] * Sweetshark weakly shakes angry old man fist. [11:03] chrisccoulson, you should conflicts on non compatible locale binaries or something, it's not the first time I upgrade tb and get it in english because I didn't pick the locale-fr binary in update-manager ;-) [11:03] who wouldn't want to use english? ;) [11:04] actually, i should correct myself [11:04] it defaults to "american" rather than "english" ;) [11:04] ;-) [11:04] we should fix that [11:04] default to french with an hard depends you mean? [11:04] yeah, I agree! [11:04] heh [11:04] * didrocks prepares a French version of Unity for chrisccoulson :) [11:04] thanks! [11:05] chrisccoulson, well, ignore me, I guess I'm the only one too lazy to download 150mb updates a day and who picks up partial upgrades in update-manager ;-) [11:06] seb128: stop cherry-picking the love, takes all of the delivery :) [11:06] didrocks, I would if Sweetshark stopped pushing 90mb of libreoffice binaries to download only to fix a control typo [11:06] lol [11:07] i should do the same for firefox and thunderbird too! [11:07] ;) === MacSlow is now known as MacSlow|lunch [11:07] seb128: I would love as well, it's not a download speed issue here, but install one with high level i/o :) [11:07] chrisccoulson, you do but yours are "only" 23mb ;-) [11:08] wououh, french tb again! ;-) [11:09] chrisccoulson, on the good new side still no tb hand on sent since 11.0.1 [11:09] seb128: hey, that wasnt only fo the typo! [11:09] that's good :) [11:10] seb128: although it was more than 90MB, so the contribution of the typo might be roughly 90MB [11:10] Sweetshark, lol [11:10] seb128: but I hear, you want to review my next upload. [11:10] where is mvo? we really need delta downloads [11:10] * Sweetshark makes notes. [11:11] waouh, ajmitch is still active ;-) [11:12] didrocks, did you get other people to complain about the unity staging ppa being seen as an untrusted source by update-manager? [11:12] seb128: hm what? [11:12] it's 3 times I try to desactivate, reactivate the ppa or refresh the index [11:12] ajmitch, nothing, just good to see you doing uploads ;-) [11:12] heh [11:12] seb128: no. Well, it's really an untrusted source in fact :) [11:12] * ajmitch has done a few for precise :) [11:13] seb128: I think people using the ppa are more using the CLI [11:13] seb128: you do have the ppa keys? [11:13] ajmitch, yeah, I just noticed looking to my changes box, I skipped all the haskell stuff :p [11:13] didrocks, I guess so, I've the ppa configured for some months, it usually works fine ... I guess it's rather a mvo issue [11:14] all the haskell stuff was boring syncs anyway :) [11:15] seb128: either that or a ppa one [11:15] like launchpad not signing well the Release file [11:17] seb128: I was only half kidding, reviewing the upload is greatly appreciated [11:17] Sweetshark, what upload? you mean looking to your diff to see if I spot errors? I don't think I know enough about lo for that ;-) [11:20] ajmitch crossed back over from the other side last Halloween [11:21] Laney, hey, how are you? thanks for the "default app selection" UDS topic email ;-) [11:22] * ajmitch was so looking forward to another round of banshee vs rhythmbox at UDS, too :) [11:23] seb128: you still might spot typos like "libreofficeu-core" that I miss because I autocomplete after libreo... [11:23] hehe [11:24] seb128: good thanks. no worries - i promised to raise it at the time :-) [11:24] isn't it about time for another Firefox vs. Chromium debate or have people just got bored? [11:25] jbicha_, hey ;-) [11:25] seb128: good morning [11:25] jbicha_, how are you? [11:25] jbicha_, have you seen the plans for the chrome UI? ;) [11:25] ie, drop gtk [11:25] chrisccoulson, oh, they copy firefox then :p [11:25] seb128, no, they are switching to their new aura framework [11:26] http://www.chromium.org/developers/design-documents/aura-desktop-window-manager [11:26] isn't firefox having its own rendering stuff as well anyway? [11:26] like aren't they moving away from cairo? [11:26] seb128: https://launchpadlibrarian.net/99774571/libreoffice_1%3A3.5.1-1ubuntu2~ppa1_1%3A3.5.1-1ubuntu5~ppa1.diff.gz <- thats the debdiff [11:26] seb128, the only thing they don't use cairo for is the 2d canvas atm [11:27] although, i think the plan is to stop using cairo for all content [11:28] seb128: ah, forget that, the debdiff is against ubuntu1, not ubuntu4. :/ [11:28] chrisccoulson, "The goal is to produce a new desktop window manager and shell environment with modern capabilities" <- wth, isn't chromium a web browser? [11:28] chrisccoulson, is that chrome os stuff? [11:29] Why's Chrome got to be a GTK3 hater?! ;) [11:29] jbicha_, who isn't :p [11:30] I think I'm slightly a GTK3 hater as well when they say stuff "nobody use GTK out of GNOME so we don't care about keeping compatible behaviours" === vuntz_ is now known as vuntz [11:40] seb128: it's the base for their future OS [11:41] didrocks, is that different from Chrome OS? [11:41] seb128: no, from my readings the past few days it's the new architecture for Chrome OS [11:41] seb128: it's just called "Chromium OS" [11:41] (not sure if they will have one based on FLOSS and another with extra content as with the browser) [11:42] didrocks, ok [11:42] "interesting" === ZarroBoogs is now known as Pici === MacSlow|lunch is now known as MacSlow [12:17] chrisccoulson: whoops: http://www.linuxuser.co.uk/opinion/5-problems-with-ubuntu-12-04-part-1-unity-dash-usability-issues/ [12:18] chrisccoulson: is it too late to add "email" somewhere in the thunderbird desktop file or what not? :P [12:19] do people really type "email" rather than "mail"? [12:19] i can't believe that nobody has reported that, even though it's been the case for a while [12:19] it was the same in oneiric too [12:20] chrisccoulson: you type "mail" for "email"? [12:21] heck, I probably would have typed "e-mail" [12:21] well, i don't typue anything. i already know what to look for. but i'm just surprised that nobody has ever mentioned this as an issue before [12:22] chrisccoulson: there used to be a big "email" icon in the default dash...that went away not too long ago [12:22] chrisccoulson: not many people knew about keywords until it was posted on the planet [12:25] chrisccoulson, firefox not using gettext is a pain though, it means it's not easy to translate the .desktop or keywords :-( [12:25] well firefox or tb [12:26] did the dash in 11.10 even support keywords? I know I added some to gnome-games but that was more intended for Software Center searching [12:26] jbicha_, it did [12:26] jbicha_, X-GNOME-Keywords by then [12:26] but only gnome-control-center was using some === ogra is now known as Guest67850 === ogra_ is now known as ogry === ogry is now known as ogra_ [12:54] seb128: 3.5.1-1ubuntu5 uploaded to chinstrap. please review and sponsor. [13:00] Sweetshark, did I sign for that? ;-) [13:02] seb128: oh, apropos sign ... [13:03] Sweetshark, you will ask me to write on your wikipage next? ;-) [13:03] seb128: you disclosed my evil plan! [13:04] seb128: now they are signed too. [13:04] Sweetshark, what are signed? [13:04] seb128: the *.changes file and the *.dsc ;) [13:05] Sweetshark, oh, I didn't ask you to sign the chinstrap, I was wondering if I said I would review,upload lo, but I guess since pitti is not around I've to do it ;-) [13:05] Sweetshark, I will do that in a minute [13:05] happy wednesday === desrt_ is now known as desrt [13:05] seb128: unless didrocks gets cocky like last time ;) [13:05] desrt, happy thursday [13:06] seriously? awesome!! [13:06] happy friday? :) [13:06] happy april 1st? [13:06] desrt, well it's good friday (is that how you call it?) this week so tomorrow is friday, i.e end of work week :p [13:06] ya. i guess that makes sense. [13:06] as for didrocks... i don't know what he's on about [13:07] I start hating vcses or didrocks dunno [13:07] * didrocks looks seb128 in an angry way :) [13:07] each time I want to push a change to gnome-control-center I get a [13:07] bzr: ERROR: These branches have diverged. See "bzr help diverged-branches" for more information. [13:07] i.e people keep commiting to g-c-c while I'm working on it :p [13:07] seb128: bzr rebase FTW! [13:08] seb128: but I got it, I'll stop working then ;) [13:08] and will write "as required by seb128…" :p [13:08] didrocks: happy Friday then! [13:08] jbicha: \o/ [13:08] graaaa, I screwed up my work totally [13:09] it was jbicha not didrocks who commiting to gcc under my feet [13:09] if it's thursday, i wonder if that means we get to not have the team meeting [13:09] jbicha: good job! [13:10] seb128: tell me once you rebased, so that I can push before you do for more fun ;) [13:10] didrocks, I will rebase on friday [13:10] :-p [13:10] how about we reassign all g-c-c bugs to seb128? [13:11] how about we reassign all bugs to seb128? [13:11] seb128: well, remember? I'm here on Friday AND Monday, so you will have to work on week-ends to avoid that! :p [13:11] chrisccoulson: +1 [13:11] heh [13:11] * seb128 hides [13:11] chrisccoulson: I have a handy script for that if needed ;) [13:11] nice :) [13:13] Brasero is broken? Gives me the 2,76MB free for a CDR ... [13:14] dupondje, could well be, it's a buggy software without a maintainer upstream for over a year [13:15] good morning [13:38] nice, https://areweslimyet.com/ ! [13:55] desrt, https://launchpadlibrarian.net/99845092/Stacktrace.txt does it make any sense to you? or is that again schemas install issues? [13:58] sure wish i knew what the ?? was :) [13:58] what process is this? [13:59] desrt, hud-service [13:59] desrt, bug #972868 sorry [13:59] Launchpad bug 972868 in indicator-appmenu "hud-service crashed with signal 5 in g_settings_get_value()" [Medium,Confirmed] https://launchpad.net/bugs/972868 [13:59] desrt, https://launchpadlibrarian.net/99926744/Stacktrace.txt is similar [13:59] so i changed the name of a key in the schema [14:00] or not [14:00] will we get a retrace on this? [14:00] desrt, oh, so maybe old process still running [14:00] seb128: shouldn't matter [14:00] the compiled schemas db is opened when the process starts [14:00] desrt, we got a retrace [14:00] desrt, g_settings_schema_key_init (key=0x7fff7a05a930, schema=0x1a6bd00, name=0x40d220 "end-drop-penalty") [14:00] end-drop-penalty [14:00] yup [14:00] so the only way that could happen is if the new hud-service gets started in the middle of an upgrade [14:00] ricotz: ping? [14:00] ie: before the new schema database gets compiled [14:01] desrt, sorry I gave the wrong url first [14:01] desrt, https://launchpadlibrarian.net/99859391/Stacktrace.txt [14:01] desrt, that's the retraced version [14:01] seb128: so this would happen 100% reliably if you did a hud search in the middle of an upgrade between those two versions [14:01] ok [14:01] and would otherwise be impossible [14:01] desrt, did I tell you how much I hate gsettings for such issues :p [14:02] seb128: or apt/dpkg? [14:02] when will we have our atomic upgrades? [14:02] desrt, well gsettings for not being robust to any real world issues [14:02] but yeah, better upgrades would be nice [14:02] seb128: like using your system in the middle of an upgrade? [14:02] desrt, anyway, thanks [14:02] if you do that, you're gonna have a bad time :p [14:03] desrt, right [14:03] desrt, the other one seems a real bug, I assigned it to you [14:03] bug #973203 [14:03] Launchpad bug 973203 in indicator-appmenu "hud-service crashed with SIGSEGV" [Medium,New] https://launchpad.net/bugs/973203 [14:03] uh oh [14:03] #0 0x000000000040b7e0 in hud_token_list_distance (haystack=0x7f6a2c0013a0, needle=0x1584520, matches=0x0) at /build/buildd/indicator-appmenu-0.3.96/./src/hudtoken.c:319 [14:04] uh oh [14:04] cost = hud_token_distance (haystack->tokens[0], needle->tokens[0]); [14:05] SegvReason: reading NULL VMA [14:05] that's a NULL dereference, right? [14:05] desrt, I think it is yes [14:06] so i think that could be caused by an application that has a menu item with no label... [14:06] like, a label of "" [14:06] oh. shit. [14:07] i think that might be caused by typing a single space character into the hud, actually [14:07] you'd end up with zero needles that way... [14:07] desrt, indeed, typing space segfault the hud-service :p [14:07] confirmed here ;) [14:07] NICE [14:07] desrt, archive testing for the win ;-) [14:08] okay. that's a pretty obvious fix :) [14:08] * desrt will have something in a few minutes [14:15] https://code.launchpad.net/~desrt/indicator-appmenu/no-search-tokens/+merge/100801 [14:16] seb128: looks like a did a bit too good a job of stripping spaces :) [14:16] anyway -- testing appreciated, of course [14:16] desrt, please make sure that get into today's tarball ;-) [14:17] desrt, I will test to confirm it fixes the bug for me [14:17] should I assume, neon 0.29.6-2 to be synced to precise still? https://launchpad.net/ubuntu/+source/neon27 [14:17] because I assume that would break libreoffice http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=667043 [14:17] Debian bug 667043 in libneon27-gnutls "libneon27-gnutls: Breaks bzr-svn (via subvertpy): undefined symbol ne_ssl_context_get_flag" [Serious,Open] [14:17] Sweetshark, pong [14:19] ricotz: sooo, just wanted to ponder if you would join the Hackfest in Hmaburg, AFAIK there is still some travel budget available ... [14:22] Sweetshark, hi, this sounds really nice, but i think i cant :\ -- i am getting busier lately [14:27] hamburg hackfest? sounds delicious! [14:27] ricotz: oh, bummer. [14:27] Sweetshark, yeah, sorry [14:28] desrt: heh [14:28] ricotz: np, just would have been nice ;) [14:31] Sweetshark, libreoffice uploaded, the debdiff seemed fine to me (but doh, slow to download lo and debdiff versions :p) [14:34] seb128: well, you are in training now, and 3.5.2 final will be out tommorrow if no natural desaster strikes our mirrors, so guess what I will ask of you tommorrow ;) [14:34] ;-) [14:36] desrt, your fix works for me confirmed [14:39] seb128: i gotta say... thanks for managing my interaction with launchpad :) [14:39] desrt, lol, yw ;-) [15:07] mterry, hey [15:08] seb128, hello! [15:08] mterry, good work on the unity-greeter 3 confirmations that the bug is fixed on the bug and another one who had the greeter segfaulting in loop confirmed the update bring it back to a working state [15:08] seb128, yar glad to see, especially as reporters on that bug were starting to suggest outlandish fix scenarios :) [15:09] ;-) [15:29] * chrisccoulson legs feel like jelly [15:31] chrisccoulson, doing exercice? [15:31] seb128, yeah, finished now though [15:31] chrisccoulson, what do you do? running? [15:31] seb128, biking. i think my knees would break if i ran [15:47] kenvandine, that gwibber merge this morning... is that a bug in vala? [15:48] i think it is a gee bug [15:48] which i am working around [15:48] i want to debug that a bit more [15:49] could be a vala bug, but looking at the generated C i think it is right [15:49] jbicha_: please think about unreffing the gsettings settings object when adding one :) [15:49] mterry, i also want to look at gee 0.7.2 and see if it is affected [15:49] might be fixed already [15:53] kenvandine, well, change seems harmless, assuming it fixes the issue. Might warrant a comment in code, because otherwise that coding pattern is odd and liable to be removed accidentally in future. [15:54] mterry, ok, i'll add a comment along with the bug i file against libgee [15:54] i plan to do that today [15:54] this was a bandaid :) [15:54] kenvandine, sure. marked the merge approved [15:54] it does seem to work around it [15:54] it never tries to use the iterator if it is empty [15:55] kenvandine, I've never used libgee. How do you find it? [15:55] right, time to make apport attach crash ID's to firefox bugs now [15:55] i'm fed up with asking people for them :) [15:55] chrisccoulson, heh [15:55] mterry, what do you mean? find libgee? [15:55] kenvandine, how do you like it? [15:56] love it [15:56] much better than the built in support for lists and stuff [15:56] I was always wary of adding another random library depends. But maybe if it's awesome... [15:56] it is a must imo [15:56] and we already have stuff using it [15:57] unity used to, when it was vala [15:57] libunity does [15:57] kenvandine, sure, in Ubuntu it's not a random library dep. But as an upstream it is a bit. I wonder if any GNOME apps use it [15:58] ah, true [15:58] well fedora and opensuse never complained about it as a dep for gwibber [15:58] * mterry waits for glib to get containers [15:59] * mterry goes afk for a bit === lool- is now known as lool [16:16] hmmmm, https://crash-stats.mozilla.com/report/index/33a6191e-99ab-4acc-a889-0660c2120404 :( [16:16] i hope that's the same issue i fixed yesterday [16:25] seb128, I need to debug something in nautilus, so I started checking out lp:ubuntu/nautilus [16:26] it says the branch is out of date [16:26] I don't know if that's something you have control over [16:26] I realize now that debcheckout is what I really want to use, so it grabs the desktop team's repo [16:26] but I thought you would be interested to know about lp:ubuntu/nautilus [16:28] cnd, thanks, you can open a bug against "udd" if you want on launchpad, those keep happening which is one of the reasons we don't use lp:ubuntu for desktop packaging [16:28] ok [16:29] cnd, I'm sure there are several other ones, I'm not interested enough to chase them down to have them break again next week though [16:29] ok [16:29] cnd, what issue do you work on? (just curious) [16:29] seb128, when I use a touchscreen and touch on the desktop, nautilus is grabbing the core pointer and not ungrabbing [16:29] which locks up the desktop [16:30] cnd, ok [16:47] Sweetshark: the problem is versioned conflicts makes dpkg work harded and as libreoffice is a core package, that might complicate upgrades, if it all works, great, but it's something to keep in mind [16:47] s/harded/harder/ === om26er_ is now known as om26er [17:00] * didrocks waves good evening [17:08] desrt, if I wanted to log messages from somewhere within nautilus, how should I do it? [17:08] I tried adding g_critical()s [17:08] and hoping stuff would appear in ~/.xsession-errors [17:11] nm, I'll ask in #gtk [17:20] micahg: well, the conflicts are only against outdated oneiric packages for which a newer version is known to exist in precise (from the same source package that is), so it is pretty safe. and yeah, I did full dist-upgrade test runs with the new upload in a ppa. [17:20] Sweetshark: release upgrades are usually the most difficult [17:27] Sweetshark: Wasn't the Conflicts added to (attempt to) make sync_extensions work? If so, I don't understand why those calls were removed from the preinst. [17:28] Sweetshark: dist-upgrade isn't a release upgrade [17:28] * Laney braves the downpour [19:01] seb128, so, i need a "Keywords" field in the desktop file, right? [19:02] chrisccoulson, oh yes you do ;-) [19:02] seb128, and before precise, it used to be "X-GNOME-Keywords"? [19:02] chrisccoulson, I wouldn't bother before precise, it was added as X-GNOME-Keywords in oneiric and nothing used it (unity supported it though) [19:03] chrisccoulson, http://standards.freedesktop.org/desktop-entry-spec/desktop-entry-spec-1.1.html if you need a reference [19:03] Keywords "A list of strings which may be used in addition to other metadata to describe this entry. This can be useful e.g. to facilitate searching through entries. The values are not meant for display, and should not be redundant with the values of Name or GenericName. " [19:04] thanks [19:04] chrisccoulson, while you are at it, can you update the unity list stuff to be spec compliant? [19:06] seb128, oh, whats wrong with it? [19:07] chrisccoulson, spec is http://standards.freedesktop.org/desktop-entry-spec/latest/ar01s08.html it got approved this cycle as official as well [19:07] chrisccoulson, [19:07] X-Ayatana-Desktop-Shortcuts=Compose;Contacts [19:07] [Compose Shortcut Group] [19:07] Name=Compose New Message [19:07] TargetEnvironment=Messaging Menu;Unity [19:07] Actions=Compose; [19:07] [Desktop Action Compose] [19:07] Name=Compose New Message [19:07] OnlyShowIn=Unity; [19:07] ups [19:08] chrisccoulson, http://pastebin.ubuntu.com/915015/ [19:08] chrisccoulson, basically [19:08] chrisccoulson, the pastebin is easier to read ;-) [19:08] seb128, ok, thanks [19:09] chrisccoulson, thank you for fixing those ;-) [19:26] kenvandine, hey [19:28] hey seb128 [19:29] kenvandine, ocrete was mentioning that our pidgin farsight,farstream patch might need to be updated [19:29] kenvandine, can you look at it? [19:29] just talked to him :) [19:29] kenvandine, yeah, he went for the query it seems and I was waiting on a channel discussion ;-) [19:29] kenvandine, you are on it then, all good [19:29] kenvandine, ocrete: thanks [19:30] i don't like touching pidgin :) [19:31] kenvandine, you touch it you maintain it! [19:32] i gotta stop touching packages :) [19:32] kenvandine, no no, we like it this way ;-) [19:33] mterry, that gwibber merge you reviewed turns out doesn't really work around the problem and it has been fixed in the 0.7 series [19:33] i need yet another work around :) [19:33] kenvandine, :) [19:33] i don't want to backport that fix to 0.6.x [19:33] kenvandine, ok [19:34] I can't remember who it is that develops lightdm [19:34] can someone remind me? [19:35] cnd, robert_ancell [19:35] kenvandine, ahh, thanks [19:35] and probably the one with the most knowledge about it that is actually awake right now is mterry [19:35] mterry, i assume you have touched lightdm in your unity-greeter adventures :) [19:35] cnd, robert_ancell is .au based, should be only in a couple of hours [19:36] kenvandine, you should make you fix a lightdm bug, I got told you are getting short of sources to maintain :p [19:36] kenvandine, sure [19:36] we should* [19:36] * mterry agrees with seb128 [19:36] * kenvandine hides [19:36] cnd, did you have a question? I might be able to help [19:37] mterry, we think we see some memory corruption in the X server [19:37] :( [19:37] it would be handy if we could start lightdm with a pre-started X server running under valgrind [19:37] or something like that [19:38] k, let me see how that might be done [19:38] * mterry looks at code [19:38] cnd, can't you just rename X to X.real and create a shell wrapper X calling valgrind X.real? [19:38] cnd, that's what I do usually for system services [19:38] seb128, hmm... yeah, that might work [19:38] I'll give it a go [19:41] cnd, you should also be able to set the seat property xserver-command in lightdm.conf [19:41] mterry, ahh, that sounds like exactly what I want to do [19:48] seb128, updating the desktop file for the fd.o spec whilst ensuring that it still works on older releases is a pain ;) [19:49] chrisccoulson, oh, right, you always have that fun job with firefox [19:50] i managed it though :) [19:50] chrisccoulson, well you made me read the automake documentation to fix your totem bug so it's just fair :p [19:50] heh [19:53] do i need to add "e-mail" as a keyword, in addition to "email", or is unity clever about that? [19:53] seeing as mdeslaur would search for the former, just to be awkward ;) [19:53] chrisccoulson, unity is not clever [19:53] well for that [19:54] :( [19:54] the hud is, but the dash isn't [19:54] chrisccoulson: also add "interwebs" and "googles" [19:54] chrisccoulson, oh, and add "free porn" to firefox.desktop [19:54] hehe [19:54] mdeslaur, i was thinking of adding "google" to firefox actually [19:54] do you think i should? [19:54] hahaha [19:54] chrisccoulson, would google be unhappy about that? [19:55] i'm not sure. i can't see why not [19:55] i like the "free porn" idea though [19:55] chrisccoulson, it's like "should we add "word" to libreoffice writer" [19:55] but not quite :p [19:55] because I think microsoft might not like this one [19:55] i'll add "Chrome" to the firefox one [19:55] well word is a word, "excel" is an interesting one [19:55] :-) [19:55] meh, just add "word processor, word, processor" :) [19:55] hehe [19:56] "libreoffice, is made to, excel" [19:56] ;-) [20:07] mdeslaur: the Dash, Software Center, etc. should be smart enough to show you word and/or processor without Keywords needing to comma separate the terms [20:08] jbicha: oh, hehe, that was just a tongue-in-cheek example...I'm not aware of how that works [20:09] jbicha: so just having word processor in the description would be enough for "word" to come up? [20:09] I think translators should definitely keep some English words in there too, for instance typing Help into the Dash in localized Ubuntu installs should show yelp but probably doesn't [20:09] mdeslaur: I don't deal with Unity more than I have to, but I think it does :) [20:11] cool [23:29] If I have an app which would like to schedule a task to run regularly, as a user, how should my app set that up? Editing the user's crontab programmatically seems a bit clunky, but running a daemon all the time to schedule the tasks itself seems like overkill [23:49] aquarius: /etc/cron.* entries? [23:50] micahg, that's running as root, not as the user, which means my app would need to be suid root to do it :( [23:51] aquarius: ah, is it just one non-root user or any real user on the system could use it? [23:51] micahg, any user on the system might run the app, and that app should then run something regularly for that user