[04:02] chrisccoulson: tbird growth> ah, -Os vs. -O3, I figure? [04:02] dobey: apport architecture> yes, that would be me [04:03] jasoncwarner_: I fully trust bryceh and RAOF about the driver decisions :) but that seems sensible indeed [04:03] Good morning everyone [04:17] how do I run dh_install --list-missing if I'm using bzr bd with ~ubuntu-desktop branches? [04:21] just a moment, rebooting [04:31] bryceh: still here? [04:32] jbicha: wb [04:32] jbicha: you can do bzr bd-do, which puts you in a full source tree [04:32] that's the usual mode of development in debian/ only branches [04:32] jbicha: you can hack there, build the package, run dh_install, and what not [04:32] jbicha: if you exit that subshell with 0, it'll copy debian/ back into the original bzr tree, so you can commit the changes [04:32] if you exit 1, it won't do that [04:33] bryceh: just followed up to https://code.launchpad.net/~bryce/apport/omit_empty/+merge/68616 [04:33] ok but immediately after doing bzr bd-do, I can't run dh_install [04:34] no, you have to debuild -b first [04:35] pitti: cool, thanks [04:36] * jbicha adds pitti as a search provider to his webbrowser [04:37] lol [05:11] good morning [05:18] Good morning didrocks! [05:18] hey RAOF! [05:20] How be you? Is compiz crazily slow for your-nvidia-using self also? [05:21] I'm having trouble running sudoku 3.1.3: http://paste.ubuntu.com/648808/ [05:21] I believe it should be running gtk3 for one [05:21] RAOF: compiz isn't only subtily slow there, not like crazy. However, my nvidia driver isn't running with latest kernel [05:22] RAOF: you can use gtk-w-d if you experience slowliness issue, that should fix it [05:22] RAOF: in the decor plugin in ccsm, replace compiz-decorator by gtk-window-decorator [05:23] Ah. I'm not experiencing a slowdown, I was just prodded by DBO this morning and wondered how widespread it is :) [05:23] Clearly the answer is “well known” :) [05:23] didrocks, did you see my email? [05:23] to the dx team list [05:24] DBO: not at that inbox right now :) [05:24] looking [05:24] also [05:24] for the next week [05:24] all compiz related complaints come to me [05:24] Sam is AFK [05:24] and if you see him [05:24] tell him to go away [05:25] DBO: got it [05:26] I have been using compiz 0.9.5 + uwd + 8px shadows for about 8 hours no [05:26] no slowdown [05:27] DBO: ok, let's try to make that upload, looking for property to change [05:30] pitti, ok sounds good; pushed updated branch [05:30] bryceh: ah, great [05:34] * didrocks reboots on a kernel supporting the nvidia bin driver to test the params change [05:42] DBO: it seems a little bit snappier, indeed, but can't really quantify as I didn't get the utterly slowdown there [05:42] still, will push the new theme, thanks! [05:42] didrocks, good, that will buy me enough time to find the root problem [05:42] DBO: indeed :) [05:42] what do you think of alt-tab? :) [05:43] I am pretty happy with it myself... [05:43] DBO: Where do I get that alt-tab? [05:43] DBO: it's gorgious there, especially with the latest fix :) [05:43] RAOF, press ctrl+tab [05:43] RAOF: when there will be an official unity release [05:43] oh he don't run trunk eh? [05:43] DBO: it's not in Oneiric right now, waiting for a release [05:43] DBO: No dice. [05:43] DBO: no, he's sane :-) [05:43] RAOF, run trunk [05:44] its how we role in DX [05:44] Heh. [05:44] I don't always test my code [05:44] but when I do, it's in production [05:44] DBO: TBH, I think that dx takes this joke a little too much seriously ;-) [05:44] * didrocks runs away [05:44] didrocks, did you see! [05:44] alt-tab comes with tests! [05:44] TESTS! [05:45] DBO WROTE TESTS! [05:45] DBO: yeah, I was thinking of a wrong copy and paste first :-) [05:45] DBO: and I took again a cup of coffee when I saw that [05:45] but it was real! [05:45] \o/ [05:45] wewt! [05:46] DBO: so, we won't have a good upgrade story before the gsettings backend is here [05:46] DBO: right now, conflict against static switcher [05:46] + changing for alt + tab [05:46] will do it [05:46] did you see the gsettings branch? [05:46] yeah, but I would hope someone in dx will do the review [05:47] that won't help with upgrade though [05:47] we will still have the same issue, unfortunatly, until compiz can handle deps on startup itself [05:47] DBO: a classy way if it was the case to handle upgrade would to create artifical deps + conflicts removal to add a new plugin to the list [05:48] (new theme uploaded btw) [05:48] didrocks, sexy sexy [05:49] RAOF: btw, any idea why my nvidia driver doesn't work with -5? I guess it's a dmks issue, maybe it hasn't run? [05:49] (I have to reboot with -3) [05:50] didrocks: I've got no particular idea. Could well be that dkms hasn't run for that kernel, although that would be odd. [05:51] RAOF: not that important for now TBH, will ping again if it's still the case at alpha3 with a new kernel [05:58] DBO, RAOF: GOOD MORNING [05:58] desrt, GOOD MORNING SIR [05:58] didrocks: and hello to you too :) [05:58] hey desrt, already awake? :-) [05:58] desrt: AND A FINE MORNING TO YOUR GENTLEMANLY SELF. [05:58] DBO: sorry. my shift key got stuck after typing your nicks =) [05:59] didrocks: rain drops keep falling on my head [05:59] desrt, CAPS LOCK IS CRUISE CONTROL FOR TALKING WITH CLASS [05:59] Real gentlement boom their salutations! [05:59] DBO: i appreciate your efforts towards not capitalising my nick [05:59] desrt, TAB COMPLETE IGNORES MY CAPS LOCK [05:59] desrt: don't afraid with some water falls in my appartment! :) [06:00] * didrocks knows that desrt will soon cross the 5 meters separating us, his power plug is in the main room :-) [06:01] desrt and didrocks are sharing a room? [06:01] these office romances always lead to disaster [06:02] DBO: much worse when there is also a third sharing [06:02] DBO: ...and she leaves the apartment during the day [06:02] DBO: that's even worse, he's at home and making everyone happy with some gtk/glib patch! :) [06:03] didrocks: from my sample size of 3 day, i have scientifically determined that it always rains in lyon [06:03] i now have an explanation for your productivity [06:03] desrt: ahah, because I can never go outside? :) [06:04] seems to be the case [06:04] desrt, I believe your finding conflict with a well known fact that rain makes people sleepy [06:04] desrt: it warned you this week had a shitty weather, and you told me it's the sane in Paris :) [06:04] desrt, I contest therefor that didrocks is always asleep [06:04] DBO: there is coffee and tea for that :) [06:04] didrocks: when i said that paris was also shitty, i wasn't talking about the weather... :) [06:05] same* [06:05] * desrt hopes steve isn't around [06:05] seems not :-) [06:05] DBO: deeper? [06:05] desrt, Im always happy to go deeper [06:14] Bah. Is someone fixing it so that when I accidentally click on the volume slider of indicator-sound it doesn't (a) crash unity-panel-service and (b) stop alt-tab working for 30 seconds? [06:16] RAOF: oh? I know there is a bug on the volume slider indicator for not working, but it just freezes everything for 3-5s here [06:16] not crashing u-p-s [06:17] Its IDO/GTK3 related afaik. [06:18] didrocks: Well, it does that too. It's rather curious, though. Compiz updates nicely - applications paint, holding down causes the numbers to fade in on the launcher - but doesn't respond in any other way. [06:18] RAOF: I have no paint at all there, but then, when it's painting back, it's working :) [06:19] weird, let's see if fixing the first issue will help for yours [06:19] I'd have thought the current crashy state of indicators was a priority... :) [06:19] Ah, software :) [06:20] pitti, any idea what causes this warning? http://paste.ubuntu.com/648829/ [06:20] pitti, I thought it was the X apport hook but am seeing it for other apport hooks too [06:23] wait, no [06:23] hrm. [06:28] bryceh: interesting, haven't seen that yet [06:28] didrocks: does the windows+number shortcuts still work for you? they broke for me with the new compiz [06:29] bryceh: ah, can reproduce [06:29] pitti, it seems to happen prior to entering add_info() [06:29] pitti: no, I reported the bug already [06:30] bryceh: also happens with another dialog [06:30] bryceh: thanks, looking [06:30] pitti, if I do ubuntu-bug compiz, it shows up when compiz' first dialog pops up, prior to the x hook firing off [06:30] however, I tried totem, firefox, linux, gedit and those all work with no warnings printed [06:31] also happens with apport-bug foo [06:33] strange, though; GtkMessageDialog _does_ want a GtkMessageType [06:36] bryceh: ah, got it [06:36] bryceh: fixed in trunk [06:36] pitti, great [06:37] * bryceh posts xdiagnose 0.9 with fixed up apport hooks for X and compiz [06:52] Speaking of apport, I just remembered I wanted to extend the alsa hook a little. [06:53] * TheMuso gets a branch ready, since what he wnats to do is rather trivial. [06:54] mvo: good morning :) [06:54] TheMuso: feel free to just commit it [06:55] pitti: To apport trunk? I didn't even know I had access to that. [06:55] TheMuso: lp:~ubuntu-dev/apport/apport-symptoms [06:55] pitti: Its in apport proper afaik. [06:55] TheMuso: oh, I thought you meant the audio symptom [06:55] TheMuso: you mean the one in hookutils? [06:55] pitti: No, the hook that dumps card info etc. [06:55] right, that needs a branch [06:55] yeshookutils [06:55] pitti: Right, I did so and made a merge proposal. [06:56] I have an 1 h appointment now, will look at it later [06:56] whoops ran lp-propose-merge incorrectly, fixing. [06:56] No hurry. [06:56] Its just something I have wanted to do for a while, to help with audio debugging. [06:57] good morning didrocks [07:06] are we considering switching to gksu-polkit? [07:06] hey mvo [07:06] jbicha: we don't, but we plan to switch to pkexec [07:06] jbicha: we have a PolicyKit patch in the pipeline for these [07:07] mvo: so, I'll start from the-aestetics branch for oneconf and try to bind it to the gtk3 branch, is it ok? (seems that the "all softwares" as I told you yesterday, doesn't work there with the categories) [07:08] didrocks: yeah, that sounds good, its a lot of work-in-progress in this branch, keep that in mind, but if you add the features as isolated as possilbe it should be ok [07:08] hey pitti [07:08] mvo: nice! I'll just need to see why the categories aren't working as I'll base the oneconfpane on it [07:10] Morning all! [07:11] didrocks: what exactly do you see ? i.e. any error message or what do you click? [07:11] mvo: no, I have the busy cursor and nothing appear (the other panes are ok though) [07:13] mvo: the pane itself stay grey, even no spinning draw [07:13] mvo: I guess it's because of ImportError: No module named mock [07:13] didrocks: oh, indeed [07:14] mvo: I'm installing python-mock, what's is it for? :) [07:14] appart from mocking me ;) [07:15] I'll need to test it for my oneconf backend, looks interesting :) [07:16] ah now a self.appname = unicode(appname) unicode issue (probably because of è in logithèque) [07:18] didrocks: its just there (the mock) for testing while the new banner feature is waiting for deployment [07:18] hey seb128! you are here early [07:18] hey mvo [07:18] "early", well trying to go back to "normal hours" ;-) [07:19] salut seb128 [07:19] lut didrocks [07:19] mvo, didrocks, how are you today? [07:20] pitti: libreoffice_3.3.3-1ubuntu2_source.changes is on chinstrap waiting for upload to natty-proposed (fixing lp#813148) [07:20] seb128: I'm ok, thanks, hoping that I'll have some time to code :-) [07:20] and you? [07:21] didrocks, I'm fine thanks [07:21] didrocks, "time to code", ahah good one, are you new around? ;-) [07:21] seb128: yeah, this is my first day! :-) [07:21] didrocks, that never works, there is always somebody who needs something :p [07:23] bonjour seb128 [07:23] hey pitti [07:23] hey Sweetshark [07:23] how are you? [07:23] getting time-pressured to have a reasonably clean "desk" before holidays :) [07:24] TheMuso: merged, thanks! [07:26] Sweetshark: oh argh -- mind to rebuild it again, with -v1:3.3.2-1ubuntu5 ? this will include the previous -proposed changelog [07:26] (required SRU policy) [07:26] Sweetshark: does that take very long? a source package build, I mean? [07:27] mvo: waow! fixing the unicode issue and then, it's very slick! nice work :) === tkamppeter_ is now known as tkamppeter [07:27] it could also be hand-crafted into the source.changes, but that's a bit fiddly [07:27] mvo: can you just commit that to not hate french people please: http://paste.ubuntu.com/648866/ ? :) [07:28] wow, it doesn't default to utf-8? [07:29] pitti: ascii strict [07:29] * pitti usually uses str.encode() which does [07:29] so è isn't in it :-) [07:29] didrocks: sure, will do [07:29] thanks :) [07:29] pitti: not compared to other LO operations. ;) [07:30] oh, I lied -- .encode() only defaults to utf-8 in python 3 apparenlty [07:32] * Sweetshark .oO( *puff* ) [07:32] pitti: should I also do a -sa upload as 3.3.2-1ubuntu5 had none of the source used in 3.3.3? [07:33] Sweetshark: no, -sd is fine [07:33] Sweetshark: the new origs are already in the archive [07:33] Sweetshark: -v only includes the previous changelog, nothing else [07:33] Sweetshark: it's mostly for getting a correct display in update-manager (all changes you'll get in -updates) plus correct bug ref parsing in the SRU reports [07:34] pitti: heh, I explicitly did a -v1:3.3.3-1ubuntu1 last time :( [07:35] that's the default (only include the topmost changelog) [07:35] pitti: libreoffice_3.3.3-1ubuntu2_source.changes is on chinstrap (again) waiting for upload to natty-proposed (fixing lp#813148) [07:35] uh, that was too much now [07:36] goes all the way back to 1:3.3.2-1ubuntu6~ppa1 [07:36] but I'll just remove the extra bits from the sources.changes [07:36] oh, that actually might be right [07:37] looks ok to me ... [07:37] uploaded, thanks! [07:38] hum [07:40] pitti, btw I pinged the yorba guys about shotwell on gtk3, they are investigating it and will reply within a week to tell if they will do it this cycle or not [07:40] robert_ancell, hey ;-) [07:40] hi [07:40] robert_ancell, how are you? [07:40] busy [07:41] pitti: does bug 813148 now need additional SRU infos, or should we continue to use 709778? [07:41] Launchpad bug 813148 in libreoffice "libreoffice in natty-proposed has icon theme regression" [Medium,Fix committed] https://launchpad.net/bugs/813148 [07:41] robert_ancell, I can imagine ;-) I just wanted to tell you that I like the unity greeter :p [07:43] Sweetshark: have an appointment now, bbl [07:43] vuntz, hey, do you plan to review that wnck patch? if not I will just upload a version with ubuntu_ namespace to the functions to oneiric if that's ok with you [07:44] vuntz, we can go back to the upstream version the day you decide on the naming then ;-) [07:45] robert_ancell, is there anything that we can do to help on lightdm? is testing trunk, or valgrinding it a bit, just to see etc useful or should we wait for stabilization a bit before doing that? [07:46] seb128, I'm just trying to get the 0.9.0 release out, then I'll be in a better position to say "it should be rock stable, go for it" [07:46] seb128, if you want to hunt for leaks, valgrind, or add unit tests that would all be helpful [07:46] robert_ancell, ok, great, I will let you work then ;-) [07:46] agh, lost my train of tought [07:46] robert_ancell, sorry :-( [07:46] :) [07:47] robert_ancell, I will let you work, good hacking ;-) [07:49] seb128: I'm having problems getting my gnome-sudoku to run [07:49] jbicha, what does it do? [07:49] seb128: http://paste.ubuntu.com/648808/ [07:50] hum [07:50] robert_ancell, pitti: ^ do you have any clue about that? [07:50] I thought it would be using gtk3 [07:50] seems gir ish [07:54] jbicha, weird [07:54] import gi [07:54] gi.require_version("Gtk", "3.0") [07:54] jbicha, do you have the gtk3 gir installed? [07:55] seb128: yes, gir1.2-gtk-3.0 and I went ahead and uninstalled the gtk2 one [07:56] jbicha, no difference? [07:57] seb128: right, but Rawhide's sudoku 3.1.3 works [07:57] hum [07:58] jbicha, well let's see if pitti has an idea when he's back, but let's not block the update on that, we can fix it later [07:58] especially if the issue is in the gir stack rather than gnome-games [08:00] seb128: building without --disable-gnuchess uses a bundled gnuchess to provide AI [08:03] jbicha, ok, usually if you change a flag just put a rational on why in the changelog [08:03] so it's easier for others to understand the motivation [08:04] morning [08:05] hey rodrigo_ [08:05] how are you? [08:09] seb128, I'm fine, and you? === jibel__ is now known as jibel [08:21] rodrigo_, i'm fine thanks [08:21] rodrigo_, what do you work on at the moment? [08:21] jbicha, hum, I didn't realize your gnuchess configure option would add a depends, would you mind reverting that for that update? [08:21] seb128, some g-c-c bugs [08:22] jbicha, gnuchess is in universe and we better handle that change in a following update rather than block the update and dh_python2 switch to a new mir [08:22] seb128: oh good point about the depends [08:22] rodrigo_, ok, don't bother with the g-s-d snapshot thing, that can wait monday new tarball [08:23] seb128, right, but I'll do the release and package tomorrow, as Monday is a holiday here, so will do both tomorrow [08:23] or maybe during the weekend [08:23] ok [08:24] need to run some quick errands, brb [08:25] seb128: pushed [08:30] jbicha, thanks! [08:37] Sweetshark: that's fine; I'll close this once it gets accepted into -proposed [08:41] seb128: uh, never saw http://paste.ubuntu.com/648808/ [08:42] seb128, jbicha: no, that's the static pygtk binding [08:46] seb128: let me work on this a bit more before uploading then [08:52] jbicha, ok [08:59] ugh, lp times out [08:59] is it in maintainance today? [09:06] ok, I can't figure it out, the only obvious reference to pygtk in the source is http://git.gnome.org/browse/gnome-games/tree/configure.in#n485 [09:07] jbicha, do you have a leftover .pyc on disk from the previous version that didn't get cleaned on update? [09:08] jbicha, like in the old python dir [09:08] since you switched to dh_python2 it changed dirs, we got bugs before where there was leftover pyc in previous packaging system dirs [09:10] jbicha, strace it and see what pyc it loads maybe [09:32] seb128: ok, can't figure it out [09:32] jbicha, did you strace it? [09:32] jbicha, can you pastebin the strace? [09:32] yes [09:33] you can also try the gnome-games from https://launchpad.net/~jbicha/+archive/dev it uses pysupport though [09:34] jbicha, try to strace gnome-sudoku 2>&1 | grep pyc I guess [09:35] http://paste.ubuntu.com/648951/ [09:40] jbicha, random idea, does uninstalling python-bugbuddy makes a difference? [09:57] seb128, pitti: it's python-launchpad-integration that was breaking it [09:57] oh, makes sense [09:57] you should use the gir for it [09:58] or it will bring the old gtk2 [09:58] jbicha, good catch ;-) [09:58] python-bugbuddy wasn't installed by the way but eventually digging through strace I found what was [09:59] ok [10:00] jbicha, btw is vinagre still broken? do you know if it is broken on other distros as well? [10:01] seb128: it's not broken on other distros, but I don't understand dso linking and all that [10:01] to know how to patch it for Ubuntu [10:01] is there a bug open about the issue with what you figured? [10:06] seb128: https://bugzilla.gnome.org/show_bug.cgi?id=652531 [10:06] Gnome bug 652531 in general "vinagre 3.1.2 - gcc warnings known to cause issues (raised to errors by Open Build Service)" [Major,Resolved: wontfix] [10:06] jbicha, thanks [10:07] we should have a bug in launchpad about it as well to track it for oneiric [10:08] jbicha, oh, that's the build issue [10:08] there's also https://bugzilla.gnome.org/show_bug.cgi?id=653558 [10:08] Gnome bug 653558 in general "Unable to connect to RDP or SSH" [Normal,Unconfirmed] [10:09] right, I was rather talking about that issue [10:10] my last vinagre upload didn't build on amd64 so that's what I thought you were referring to [10:10] jbicha: ah, that explains it :) (sorry, was on another call) [10:10] right, I know what the issue is for this one, it's similar to the one I fixed in the previous upload [10:10] they need to declare their functions before using them [10:11] which upstream claim will be fixed when they properly land the vala refactoring he's working on [10:12] jbicha, do you update gnome-games to fix the launchpadlib thing or should I or somebody have a look to it? [10:13] seb128: I just disabled lpi for sudoku & pushed [10:14] jbicha, ok, I will review what you have and probably fix the lpi thing before uploading, it should be a similar call but from the gir rather than from the static binding [10:15] jbicha, thanks for the update! [10:16] smspillaz, didrocks: is https://blueprints.launchpad.net/ubuntu/+spec/desktop-dx-o-compiz (i. e. mostly the gsettings migration) actually going to happen for oneiric, or should we move to p? [10:16] seb128: oh, I forgot that gi works that way, but you could probably fix it faster than I could [10:16] pitti: it has to happen if we want the new unity plugins to be activated [10:16] pitti: smspillaz did some work there [10:17] jbicha, if you want to fix it feel free but I'm fine doing it before uploading otherwise [10:17] and there is a pending branch waiting for review by dx [10:17] jbicha, just tell me if I should do it or not so we don't duplicate work ;-) [10:17] didrocks: then this should be milestoned for a3 or beta-1 at latest (which is already past FF) [10:17] seb128: you go ahead :-) [10:17] didrocks: do you think that's realistic for two weeks? [10:17] jbicha, will do, thanks [10:17] pitti: indeed, but it seems that dx isn't using the WI tracker [10:17] pitti: if it's reviewed as I reasked this morning, can be [10:18] didrocks: set to a3 for now, thanks [10:18] thanks :) [10:21] TheMuso: https://blueprints.launchpad.net/ubuntu/+spec/desktop-dx-o-unity-a11y looks strange -- there's three real WIs for you at the top, and a bunch of "almost formatted like work items" at the bottom [10:23] pitti: it's done, blocked on review [10:24] smspillaz: cool, thanks [10:33] smspillaz: gord had some questions he asked you on IRC IIRC [10:33] didrocks: they're not in my scrollback :( [10:34] smspillaz: I guess you can directly see with him :) [10:35] didrocks: he needs to comment on the MR [10:35] guys i'm right here [10:35] * gord waves [10:35] smspillaz, where do you get CompizGSettings from? [10:35] i can't build it [10:35] come on, it's not like you are not on the same team and you need me for binding the communication :/ [10:36] didrocks, i asked like three times ;) smspillaz is just not easy to grab a hold of this week [10:36] because I'm not meant to be here ;-) [10:37] tkamppeter, https://code.launchpad.net/~rodrigo-moya/ubuntu/oneiric/system-config-printer/fix-787694/+merge/68656 [10:37] gord: lp:compiz-compizconfig-gsettings [10:38] smspillaz, erm, thats what this is a branch of isn't it? it tries to load COmpizGSettings and cmake complains "include could not find load file" [10:40] rodrigo_, thanks, I will test today. [10:43] gord: oh, sorry [10:43] gord: there's a merge proposal to that [10:43] hang on a minute [10:46] https://code.launchpad.net/~compiz-team/compiz-compizconfig-gsettings/compiz-compizconfig-gsettings.port_to_gsettings/+merge/68026 [10:46] gord: ^ [10:47] hello everyone [10:47] smspillaz, right, thats the branch i am trying to build [10:47] smspillaz, it doesn't build because of the missing load file [10:47] gord: ah, dammit [10:48] gord: I think you need the schema generation stuff [10:48] so I'll get that for you [10:48] (sorry about that, it's my screwup) [10:48] pitti - yeah, thunderbird got hit by the broken optimizations too [10:50] hey chrisccoulson, how are you? [10:51] gord: https://code.launchpad.net/~compiz-team/compiz-core/compiz-core.gen_gsettings_schemas [10:52] rodrigo_, I have tried the patch, but the resulting .desktop file is not valid. It gives this error: [10:52] /home/till/ubuntu/system-config-printer/system-config-printer-1.3.3+20110712/debian/tmp//usr/share/applications/system-config-printer.desktop: error: value "Unity;" for key "OnlyShowIn" in group "Desktop Entry" contains an unregistered value "Unity"; values extending the format should start with "X-" [10:52] Error on file "system-config-printer.desktop": Failed to validate the created desktop file [10:57] rodrigo_, it was a "desktop-file-install" command which caused this error. [10:57] tkamppeter, that's a bug in desktop-file-install [10:57] tkamppeter, Unity was added to the supported values in the specification but the tools didn't get updated yet [10:58] seb128, do you have a quick workaround so that I can test rodrigo_'s work? [10:58] tkamppeter, hmm, are you running oneiric? [10:58] rodrigo_, yes. [10:58] ah, what seb128 said [10:59] tkamppeter, don't use desktop-file-install, just cp the .desktop as a workaround? [10:59] rodrigo_, you are doing your Oneiric development work on Natty? [10:59] tkamppeter, no [10:59] tkamppeter, but you should open a bug against desktop-file-utils [11:00] tkamppeter, we didn't nothing it before because other softwares simply cp their .desktop, they don't use desktop-file-install [11:05] seb128, I see now, the upstream Makefile uses it. Seems that I have to patch that away. [11:17] hi seb128, i'm good thanks. how are you? [11:17] i only just noticed your message there ;) [11:20] I'm fine thanks [11:20] just finished to eat and making some coffee to start the afternoon ;-) [11:23] heh :) [11:23] i need some coffee [11:26] seb128, rodrigo_, in my "Off button" menu there are no "System Settings" any more, how do I recover them? [11:26] tkamppeter, run gnome-control-center from whatever you usually use to run commands [11:27] it will be back in the ui later in the cycle maybe in the launcher [11:27] tkamppeter, yeah, run gnome-control-center on a terminal [11:28] rodrigo_, now I have done the following: [11:31] rodrigo_, I have applied the debdiff of your changes on s-c-p, but due to above-mentioned desktop-file-install problem I have commented out the OnlyShowIn=Unity; line by editing the patch. [11:31] rodrigo_, this way I could build s-c-p and install it. [11:31] rodrigo_, then I logged out and logged in again [11:32] rodrigo_, and started g-c-c from the command line (see problem above). [11:32] Now in g-c-c there is no printing icon, neither from s-c-p nor from the GNOME printing tool. [11:33] not even the gnome one? [11:33] rodrigo_, yes, not even the gnome one. [11:35] tkamppeter, right, the gnome one has OnlyShowIn=GNOME; and you removed the OnlyShowIn, so none of them shows up, right [11:35] can you edit /usr/share/applications/system-config-printer-gnome.desktop and add the OnlyShowIn=Unity; and try again? [11:39] don't forget to run sudo update-desktop-database after editing a .desktop [11:40] I did so now. After simply closing and restarting g-c-c I get again now printer icon at all, even not from the GNOME tool. [11:41] tkamppeter, on unity I see the gnome one, I thought it was due to some missing thing from mterry's work to teach g-c-c about OnlyShowIn, so I'll ask him later [11:42] rodrigo_, I have tried to log out and log in again and I have done sudo update-desktop-database now. All this does not make the s-c-p icon appear in the g-c-c/ === xclaesse is now known as Zdra-n900 [11:47] tkamppeter, commit the patch, it's correct, it's just non trivial to test due to the issues you are running into [11:48] tkamppeter, yeah, I'll talk with mterry and see what's missing for this to work [11:49] rodrigo_, tkamppeter: likely a call the gmenu-update-... [11:49] which is a trigger in the package [11:49] seb128, bug 814034 [11:49] Launchpad bug 814034 in desktop-file-utils "desktop-file-install rejects .desktop files with "OnlyShowIn=Unity;"" [High,New] https://launchpad.net/bugs/814034 [11:49] tkamppeter, thanks [11:49] urg [11:49] desrt, glib iz borked [11:50] seb128, I cannot commit the patch yet as it will FTBFS due to bug 814034. [11:50] Launchpad bug 814034 in desktop-file-utils "desktop-file-install rejects .desktop files with "OnlyShowIn=Unity;"" [High,New] https://launchpad.net/bugs/814034 [11:50] tkamppeter, ok, will get that fixed [11:50] seb128: what did i do? :( [11:50] desrt, dunno if it's you, 2.29.10 to 12 segfault applications on start [11:50] seb128: i should warn you. you are causing dconf-dump to be pushed down my todo list :) [11:51] seb128: backtrace appreciated :) [11:51] desrt, trying to get one, it's not easy since I can't open a web browser or anything [11:51] let me downgrade it [11:51] seb128: pastebinit! [11:51] it's doing an invalid free in the unico theme engine [11:51] oh. [11:51] known issue. it's a gtk bug, actually. [11:51] the workaround is committed upstream [11:51] hum [11:51] (to gtk) [11:52] great [11:52] "great" rather [11:52] desrt, would have been nice to let a notice of that somewhere [11:52] so what should I do, backport the gtk patch and make glib breaks gtk << backport? [11:53] seb128: http://git.gnome.org/browse/gtk+/commit/?id=7741f5a09a841c4dc93727b990defc303510ed2c [11:53] desrt, isn't that an abi break in some way? ;-) [11:53] i don't consider it to be [11:54] gtk was invalidly depending on undocument behaviour [11:54] is other code out of gtk likely to have issues? [11:54] i doubt it [11:54] and i've heard no reports [11:54] even see the GTK code in question has /* FIXME: hack hack hack, replacing pspec->name to include namespace */ [11:54] they knew they were causing trouble :) [11:54] desrt, do they both need to be updated together? [11:54] desrt, i.e will the gtk commit create issue on glib 2.29.10? [11:54] i don't know. [11:55] oh. no. definitely not. [11:55] actually, yes. it will. [11:55] sorry [11:55] :-( [11:55] that's a sucking situation [11:55] ya. that's kinda lame now that i think about it. [11:55] I think I will just revert the glib "fix" [11:55] it seems an abi break to me [11:55] that's going to cause other problems with new code [11:56] it's only an ABI break in so far as other code is doing invalid hacks to modify the memory owned by glib itself [11:56] desrt, having a glib update making every gtk application segfault on start if you don't update gtk is going to be a real issue [11:56] seb128: so do a lockstep upgrade to the new gtk? [11:57] you will have an hard time assure everybody upgrade both glib and gtk together [11:57] seb128: i'm not very happy with the updated gtk hack, to be honest [11:57] desrt, well, that will fix it for us [11:57] desrt, there is still going be lot of people "out there" building a new glib and going to be hit hard [11:57] getting a "nothing start on my system with the new glib" [11:57] yes. that's true. [11:58] i'm a bit sad that a note wasn't included in the NEWS at least [11:58] i think we didn't realise it would be a problem until it was already too late [11:58] that really sucks, you should put big warning "it's going to break your system if you don't have a GTK which has no tarball yet" [11:58] desrt, can't you just back out the change? [11:58] i'll see about getting a gtk release [11:59] hum, well we can deal with that [11:59] but I still see lot of people "out there" who will do a glib update because they need it [11:59] not especially gtk hackers [11:59] just glib users [11:59] that will get bitten hard [12:00] nobody ever stated that glib updates should be locked with gtk ones [12:00] and nobody ever stated that gtk will never have any bugs [12:00] well having a bug is an issue [12:00] stopping any software to start is a bit over "having a bug" [12:00] this code is buggy. even after the update, it is still buggy. [12:01] (just slightly less) [12:02] desrt, will it break gtk2 as well? [12:02] i hope not. let me check. [12:03] you're on gtk-2-24? [12:03] desrt, well, I'm not deciding for glib but I think it's a mistake to put a version out which will break system of people not tracking unstable gtk [12:03] desrt, yes [12:03] desrt, I can see lot of cases where people need gio or glib and track glib updates without gtk ones [12:03] here I am [12:03] seb128: you make a good point, from a pragmatic standpoint [12:03] seb128: i think the problem is that we didn't know of the problem until after the release was already done [12:04] desrt, it's maybe not technically an abi break but seems equivalent on the consequence [12:04] Cimi: unico is crashing with recent glib changes [12:04] desrt, well it's not too late to back out, it's only an unstable serie [12:04] desrt: nice :D [12:04] desrt: fix, error? [12:04] Cimi: and we're wondering if it is due to you copy/pasting code from gtk, or if it's just via normal gtk use [12:04] Cimi: http://git.gnome.org/browse/gtk+/commit/?id=7741f5a09a841c4dc93727b990defc303510ed2c [12:05] Cimi: this is the code [12:05] seb128: i can confirm that this evil code is not in gtk2 [12:05] ok, that's something [12:05] it means you will only hit people on gtk3 which update glib and not gtk [12:05] desrt: is this code in unico? I doubt [12:05] Cimi: okay. good. [12:06] so it's not a unico problem. strictly gtk. [12:06] which is likely still a non trivial number of users imho [12:06] seb128: yes. i agree. [12:06] desrt, you create a chicken egg issue as well [12:06] desrt, let's say you are on glib 2.28 gtk 3.0 [12:06] desrt: I didn't copy/pasted code from gtk+, just gtkroundedbox.c (yesterday), is not in oneiric [12:06] you need to update to glib 2.30 to build gtk 3.2 [12:06] desrt, but when you will update your glib you will break your system [12:06] how do you build gtk3.2? [12:07] you need to debootstrap out of the real system? [12:07] it's ridiculous... [12:07] desrt: bug trace^ [12:07] ? [12:07] seb128: you apt-get dist-upgrade and it happens at the same time :) === MacSlow is now known as MacSlow|lunch [12:07] desrt, as said it's not a real issue for us but I'm thinking from a glib community perspective [12:08] desrt, though I'm pretty sure some people do use glib trunk builds on natty and will get hit as well [12:08] seb128: i think those that follow the unstable cycle will encounter problems... [12:08] it's always a bumpy ride [12:08] desrt, no, my example was a stable to stable situation [12:08] right. of course. [12:08] 2.28, 3.0 combo install, GNOME 3.2 is out [12:08] what do you do? [12:08] i'm not sure we can avoid it, other than to backport the fix [12:08] you start by upgrading glib so you can build gtk [12:08] which lead to have no working box [12:08] to the stable series [12:09] desrt, do you really need that glib "fix" [12:09] (after adjusting it to work with both glibs -- which should be possible) [12:09] seb128: i don't *need* it, but it sure is nice... [12:09] desrt, you have an issue that you can't fix without breaking abi in some way, it was never a stopper, let it buggy the way it was? [12:11] seb128: here's my proposal: new point release on the 3.0 stable branch that addresses the issue [12:11] it will be picked up and packaged/shipped in the usual way [12:11] and by the time the new stable release of glib is out, most should have it [12:11] that should solve the issue for 99% of people who would encounter it [12:12] desrt, what issue not having that glib commit create? [12:12] hello, which change are you talking about? [12:12] seb128: having the pspec name interned lets us to value-comparison on the gparamspec name from notify functions [12:12] instead of string comparison [12:13] it also saves a bit of memory [12:13] it's an enhancement, strictly speaking [12:13] but it's one that i'm already using [12:13] ok, i guess it is http://git.gnome.org/browse/glib/commit/?id=706b2751166bb8590e41800d0b3c3056ba529daa [12:13] ricotz, glib's gparamspec name property change [12:14] seb128, thanks [12:14] ricotz: it's not that one. [12:14] ricotz: that's the one i did today to prevent situations like this from happening in the future [12:14] the one that really introduced the problem is slightly older [12:15] http://git.gnome.org/browse/glib/commit/?id=d6c30e1766c975dd79e6f252d73c6c0581b64b01 [12:15] here [12:15] desrt, alright [12:15] desrt, well, your call, I would value the stability over the small optimization [12:15] desrt, but I can understand if you go the other way around [12:15] it's just going to hit users whatever you do and I think the small win is not worth [12:15] seb128: i'll fix the problem in all affected branches of gtk and arrange to get some new releases [12:16] desrt, the fixed gtk versions will work on both glib < current et on current right? [12:16] like on 2.28 and trunk [12:16] yes. i will ensur ethat. [12:16] thanks [12:16] not very good to release a gtk 3.0.x that only works with unstable glib :) [12:16] well I still think it's a mistake but I've made my point [12:17] we just have to agree on disagreeing there [12:17] seb128: i think after the new gtk releases this will mostly be a bad memory [12:17] what you will do will mitigate the effects but not void them [12:17] that's true [12:17] but the fact remains that this is a gtk bug [12:17] so the fix should be with gtk [12:17] desrt, you are sure it's not breaking qt or anything out there right? [12:18] desrt, like not sure how they do their gtk theming integration [12:18] seb128: only if they are using glib in an invalid way [12:18] seb128: garnacho even realised when he wrote this code that what he was doing was wrong. he even added a comment. [12:19] desrt, well that's sort of my point, even if the way it was done was invalid doesn't make it less of a compatibility break to change it [12:20] but anyway let's see, I will try backporting the gtk patch and see if anything else break [12:20] seb128: if we tolerate abuses of our API and fear making changes because of their existence then we sort of miss the point of having a defined interface [12:20] seb128: if you backport that patch make sure you introduce a glib dependency [12:20] desrt, yeah, I wouldn't mind breaking a few applications, breaking gtk is a bit different ;-) [12:20] because that patch will break on the old glib [12:20] i will prepare one that is safe for both [12:20] desrt, right, I will lock both together [12:21] well I will just backport for local testing [12:21] and release that on the stable series [12:21] k. [12:21] I will probably wait your updated version for the official update [12:21] really sorry for not having NEWS about this [12:21] desrt, I want to make sure you that doesn't break qt [12:21] because I don't want to tight glib, gtk and qt updates together [12:21] that will be some kind of nightmare, indeed :) [12:21] desrt, that's ok, I did catch it before upload ;-) [12:21] seb128: oh. great. [12:22] seb128: so just wait until we have the new gtk, i guess [12:22] desrt, you might want to drop an email on the gtk-devel list or something about it though [12:22] seb128: sure. will do. [12:22] thanks [12:22] hey mterry [12:22] now that i know the world didn't explode i can take some time to consider the proper fix [12:22] seb128, hello! [12:23] mterry, can you look into bug #814034 [12:23] Launchpad bug 814034 in desktop-file-utils "desktop-file-install rejects .desktop files with "OnlyShowIn=Unity;"" [High,New] https://launchpad.net/bugs/814034 [12:23] yup [12:23] mterry, thanks [12:24] mterry, desktop-file-install is a bit stupid, it errors out of unknown values [12:24] chrisccoulson: is it known/wanted that thunderbird isn't explicity set at the default mailto app? [12:24] seems like it should warn only, but still Unity should be added to desktop-file-utils [12:25] didrocks, I guess it needs to be changed in desktop-file-utils, maybe mterry can fix that for you while he's doing work on the source [12:25] * rodrigo_ lunch [12:26] seb128: indeed, it's not right now, so just checking :) [12:26] well or just commit the change to the d-f-u so it's in the next upload [12:27] let's do that quickly [12:29] and done :) [12:30] mterry: you should pull if needed [12:30] didrocks, was evolution ever set as the default in desktop-file-utils? [12:30] didrocks, ack [12:30] i think it just worked by virtue of being the only mail client installed [12:30] chrisccoulson: maybe not, but explicit is better than implicit :) [12:30] yeah [12:44] pitti: cool. is there some awesome way to do magical apport stuff for plug-ins in apps like nautilus/rbox/banshee/etc? [12:46] seb128: we decide to use glib_check_version() to work with both old and new glib with the minimal possible amount of changes [12:46] seb128: then we'll land the fix on the stable branch [12:46] desrt, right, I read #gtk [12:46] desrt, thanks ;-) [12:46] ah. didn't see you there :) [12:46] desrt, enjoy lunch [12:47] didrocks is making rice :) [13:28] mterry, just curious but why was that deja-dup upload needed? is that likely to be an issue for other things? [13:28] i.e do we need to rebuild other rdepends? [13:29] rodrigo_, s-c-p 1.3.3+20110712-0ubuntu2 (with your patch) is uploaded now. [13:29] seb128, yes, I'll do indicator-datetime too [13:29] seb128, do we have others? [13:29] seb128, (just panels need to be rebuilt) [13:29] mterry, gnome-bluetooth [13:30] mterry, why do they need a rebuild? [13:30] seb128, those are both working fine... [13:30] mterry, there was no recent g-c-c update [13:30] seb128, there was 3.1.3, which wasn't recent, but I didn't try to enter the deja-dup preferences since then [13:31] mterry, oh ok [13:31] seb128, it added a new virtual function to the panel class [13:31] mterry, alright, thanks [13:31] seb128, and there is no buffer room on the class (since it's private! :)) === MacSlow|lunch is now known as MacSlow [13:42] tedg, does unity or do the indicators strip their environment in any way? [13:43] mterry, No, just what dbus gives them. So it's the environment very early in the boot process. [13:43] (or I guess login) [13:43] k [13:43] mterry, If the variable isn't set before dbus runs, it won't be seen. [13:45] hrmm, why am i getting gpg sig error from dput when uploading to my ppa? boo. [13:45] dobey, ignore it [13:45] dobey, the upload likely worked [13:46] seb128: it worked, but it's annoying because i can't pass multiple source.changes arguments with one dput call :/ === m_conley` is now known as m_conley [13:58] hey seb128, that account adding problem you had... can you try that again from a branch? [13:58] kenvandine, sure [13:58] seb128, i added a print statement... and haven't been able to reproduce it again :/ [13:59] lp:~ken-vandine/gwibber/seb128 [13:59] and just run ./bin/gwibber-accounts [13:59] kenvandine, ok [14:00] kenvandine, oh, question for you, is libindicate-gtk on gtk3 required? [14:00] or is that deprecated? [14:00] we still have a wi for it [14:02] seb128, i keep harassing tedg about it... [14:02] we need it for tp-indicator at least [14:02] to display the icon [14:02] ok [14:04] kenvandine, is that anything in the debug line you are looking for? [14:04] the account string that gets printed [14:04] kenvandine, http://pastebin.ubuntu.com/649144/ [14:04] kenvandine, I've mandled the numbers because I didn't know how much they can be used to do something with my account :p [14:05] kenvandine, nothing segfaulted, it returned to the account dialog but the left list is empty [14:05] those are very publically available :) [14:05] like it didn't add any account [14:06] yeah, it has an "id" [14:06] should I try again? [14:06] which means gwibber thinks you are editing an existing account [14:06] no... one sec [14:06] ok [14:06] echo "select * from accounts;" | sqlite3 ~/.config/gwibber/gwibber.sqlite [14:07] you might need to install sqlite3 [14:07] $ echo "select * from accounts;" | sqlite3 ~/.config/gwibber/gwibber.sqlite [14:07] $ [14:07] so no accounts [14:07] wtf! [14:08] ugh it stores accounts in sqlite3? :( [14:08] dobey, yes... not passwords :) [14:08] dobey, how would you suggest storing them? [14:09] i thought gnome had this amazing new thing for that :P [14:10] dobey, we can't use it yet [14:10] i was being facetious [14:11] kenvandine, let me know if you need extra debug infos [14:11] seb128, did this line print more than once "account before checking is_new" [14:11] kenvandine, http://pastebin.ubuntu.com/649153/ [14:12] kenvandine, that's the full log [14:12] ah ha! [14:12] damn!.... it blows up if there is no username returned! [14:12] kenvandine, the first time is before I entered my email and password in the webkit [14:12] I think [14:13] oh... interesting [14:13] I saw the Couldn't find field: username on the stdout while I was typing my infos [14:13] ok... that is very helpful! [14:13] kenvandine, doh, I know! [14:13] kenvandine, it prints it when I validate the color I selected [14:13] kenvandine, it happens because I pick a color! [14:14] yeah... [14:14] :) [14:14] I didn't like the default purple ;-) [14:14] it is the whole autosaving changes [14:14] we need to not do that if the account is new :) [14:15] kenvandine, well, at least it seems that you have a clue about the issue ;-) [14:15] yeah... easy fix :) [14:17] \o/ [14:20] seb128, ok, can you pull and try again? [14:20] make sure you change the color first too :) [14:20] don't worry ;-) [14:20] I still don't like your default purple :p [14:21] * kenvandine didn't choose it :) [14:21] i don't either [14:21] bah, often the webkit view switch to the password field after I type one letter [14:22] that's facebook trying to be smart... [14:22] if I was on the laptop keyboard I would blame the palms clicking but I'm on an USB keyboard [14:22] they aren't :) [14:22] ;-) [14:22] waouh, I got account! [14:22] kenvandine, works, thanks! [14:22] woot! [14:22] thanks for finding that! [14:23] seb128, note... replies are broken right now in oneiric :) [14:23] kenvandine, yw [14:23] kenvandine, no worry, I'm a lurker, I don't do replies :p [14:23] i should have another release today :) [14:23] hehe [14:24] is rhythmbox working for anybody? [14:24] lots of changes since friday... i've been hacking on gwibber every night until like 2am for over a week... [14:24] hello btw [14:24] hey pedro_! [14:24] you guys really want me to switch to the new stuffs heh? [14:24] urg, scrolling and refresh are not working really great in that gwibber [14:24] * kenvandine removed rb ages ago [14:24] first evolution got broken and now rhythmbox? :-P [14:24] pedro_, how is it broken? [14:25] seb128, what is wrong with scrolling? [14:25] seb128, for me at least it doesn't start [14:25] kenvandine, first there is no orange scrollbar indicator [14:25] should be... [14:26] kenvandine, second I often get part of the screen "empty" [14:26] like grey and not filled [14:26] yeah, we know that :) [14:26] i'll fix that after FF :) [14:26] we try to always align a message at the top and keep the view full [14:26] it's weird, I get also a scrollbar on the left like if there was a left bar but there is none [14:26] but there are bugs [14:26] or it's empty and small [14:26] oh! [14:26] ok... that is a bug too [14:27] which may be fixed in trunk :) [14:27] actually, not trunk [14:27] scrolling with the overlay scrollbars by clicking on the thumb leads to the bottom post entry to flicker a bunch of time [14:27] then it updates [14:27] it's weird [14:27] flickers? [14:27] humm [14:28] like show,hide,show,hide [14:28] well, that is probably related to the streams not being lined up right [14:28] maximize the window [14:28] is it better? [14:28] basically all the streams are there, side by side... and we move between them [14:28] that's completly broken [14:29] there seems to be some calculation problems there [14:29] close it and start again [14:29] do you want a screenshot of the completly broken? [14:29] nah [14:29] i've seen it [14:29] i think the trigger is a post that can't be sized properly [14:30] makes the stream too wide... and it doesn't recalculate it [14:30] kenvandine, ok, it's better now [14:30] I've the scrollbar orange indicator [14:30] max thought right? [14:30] scrolling doesn't flicker [14:30] kenvandine, right [14:30] unmax now [14:30] is it still ok? [14:31] it was [14:31] I tried to resize but that broke it [14:31] now I just get a grey squar [14:31] empty gwibber [14:31] ok, after the resize... close and start over? [14:32] kenvandine, ok, it's good now [14:32] I get the orange hint and the scrolling works [14:32] it's a bit slow but works [14:32] should be very fast... [14:32] but at least it's working :) [14:32] kenvandine, I think part of the issue is because I've only 10 entries [14:33] and trying to scroll over the end of the stream seems to screw it [14:33] oh... you need more friends :) [14:33] hehe :) [14:33] kenvandine, well that's gwibber which did fetch only 10 [14:33] so a small stream is something we probably don't handle well [14:33] I've over 300 in the web ui [14:33] I just connected to the website to see [14:34] interesting... [14:34] I've over 300 non-read ones [14:34] look at the other streams [14:34] like attachments [14:35] those should look mostly empty [14:35] they do ;-) [14:35] but you have plenty of them? [14:36] no [14:36] the metadata for those weren't all in the model [14:36] images and video are empty [14:36] there are a few links [14:36] humm [14:36] kenvandine, well don't bother, I will wait for the new version and see how it works [14:36] well a new version won't fetch more data [14:36] it could just be because you just added the account [14:37] facebook limits how many entries get downloaded [14:37] right [14:37] so over time it will get more data [14:37] I will try again a bit later and let you know how it does [14:37] it handles not very well too short steams [14:37] the scrolling goes over the list easily and bugs in different ways [14:37] yeah, i expected that [14:38] that's the kind of thing i'll work on after FF :) [14:39] kenvandine, it's weird, the first entry is one of you which is empty [14:39] yeah... :) [14:39] i was working on replies... and posted an empty message :) [14:39] oh ok [14:39] it didn't get your picture either [14:39] but he got dholbach's and barry's ones [14:40] you'll get them cached over time... neil is adding async loading of those [14:40] ok [14:40] kenvandine, anyway I got my account working well done [14:40] same for thumbnails of images [14:40] I will see how it behaves over time when I get a bit more datas fetched in [14:40] cool [14:41] thx! [14:41] kenvandine, thank you! [14:43] desrt, the "change fileselector on a stable update" will let make it fun for distro to figure if they want to bring the "stable update" in or not, especially for those who have a "no ui change" rule for stable updates [14:43] seb128: #gtk+ please [14:43] tremolux: awsome news on the startup time improvements for the direct deb open case \o/ [14:44] desrt, that was a side comment, I'm not sure we will win a lot from a discussion there especially if federico is not there [14:44] mvo: :) [14:44] seb128: it's a good point. [14:44] i'm quite scared to make this release, to be honest [14:44] desrt, I can raise it when federico is around if you want [14:45] seb128: i was already planning to hold off on the release until i could talk to him [14:45] mvo: makes a difference for sure! [14:45] desrt, but I know that Debian or Ubuntu would not take an update changing the ui this way, especially that the new ui didn't even get testing in an unstable tarball yet [14:45] seb128: meanwhile you could do a lot of people a favour to take my most recent patch as a vendor fix on your existing gtk package... [14:45] desrt, yeah, that's what I was thinking to do [14:45] tremolux: yeah, really good. time for a new release :P [14:45] desrt, you should probably drop an email on the distributors list about it [14:46] mvo: sounds good :) [14:46] seb128: i was planning to wait until after the release for that but maybe i will do it now [14:46] desrt, well I'm sure some distro will go for the "backport the commit rather than take a new fileselector" way [14:46] indeed [14:50] seb128: out of curiosity, why a new fileselector? what does it do? [14:50] what does it look like? [14:51] mvo, https://live.gnome.org/DocumentCentricGnome [14:52] mvo, https://live.gnome.org/DocumentCentricGnome/Help%20the%20user%20choose%20a%20place%20to%20put%20a%20new%20file [14:53] mvo, http://people.gnome.org/~federico/news-2011-07.html#01 in fact [14:53] it has screenshots [14:54] thanks [14:54] thats interessting [14:54] eww [14:56] isn't that what gnome has now? [14:58] I had the same feeling and opened up gedit to test [15:04] dobey, mvo: I really dislike the current oneiric way [15:04] it default to the previous folder rather than the current one [15:04] seb128: that depends on the app [15:04] which means that usually I get my fileselector open in the dir of what I was packaging before and not what I'm working on next [15:05] well in gedit at least [15:05] right, i think gedit remembers the last one [15:05] I usually want to open things when I'm working [15:05] i don't use gedit :) [15:05] emacs ftw [15:08] and i don't understand how those are "dcoument centric" at all [15:08] they're "file location centric" [15:40] seb128, didrocks, I confirm the link between the leak in compiz and the video driver [15:40] jibel, no leak on intel? [15:40] I switched from an nvidia to intel card on the same system and the leak is /only/ ~6MB/hour against 40 to 50MB with nvidia [15:40] waow [15:41] jibel: nvidia blob driver isn't it? [15:41] didrocks, yes [15:41] njpatel, ^ [15:41] well pitti was having issues on intel though [15:42] seb128: with the memleak? [15:42] 6MB is close to what I see on an inactive system [15:42] pitti, yes [15:42] I don't actually see it growing [15:42] I just thought that 700 MB was inordinately large [15:44] mvo, hey [15:45] mvo, when i call simulate on a valid transaction shouldn't i then see a properties changed callback [15:45] maybe its my callback signature [16:03] good night everyone! [16:05] 'night pitti [16:12] lol, my daughter has stuck a Qt sticker on her high-chair [16:12] she must have raided my laptop bag [16:14] chrisccoulson, heh, my nieces have their bikes full of ubuntu stickers :) [16:14] heh :) [16:14] seb128: dconf dump/load on master now [16:15] using the same keyfile format as /etc/dconf/db/ [16:15] desrt, you should stay in Lyon, seems it fits you well, having a productive week ;-) [16:15] seb128: breaking lots of things :D [16:16] breaking things is fun [16:16] ;-) [16:16] chrisccoulson, qt!? what's wrong with you :p [16:16] desrt, in lyon? I thought you were in paris? [16:16] rodrigo_: i'm an international man of mystery. you have to keep up! [16:17] * desrt has paris, toronto and berlin on the list for the next couple of weeks [16:18] and montreal very briefly, i think [16:20] desrt, :) [16:21] chrisccoulson: congrats to the new Qt maintainer! [16:21] desrt: ping me when you're in montreal if you want to go for a beer or something [16:22] didrocks, i'll tell her. she'll be overjoyed! [16:22] :) [16:22] chrisccoulson: I guess she designated you as it's in your house! [16:22] lol [16:23] cyphermox: i'm just passing through YUL :) [16:23] ah ;) [16:24] offer still stands for anyone visiting montreal anyway [16:24] chrisccoulson, my son loves the green Qt towel i got at desktop summit [16:25] you can both maintain it :) [16:25] kenvandine: the one that leaves green fuzz all over everything it comes in contact with? [16:25] no need to fight!! [16:25] :) [16:25] desrt, it doesn't anymore [16:25] desrt: back to working on the parallel ipv4/ipv6 committing of IPs; we got some feedback from dcbw and it mostly all works; just missing something broken for stateless dhcpv6 [16:25] desrt: well, you are in green right now :-) [16:25] it did though... made a mess of the washer... but he kept using it [16:25] is this the sort of stuff i miss out on by not going to the desktop summit? [16:25] no green towels for me :( [16:25] didrocks: only because you were [16:25] it's nice to see connections establish in <5 seconds instead of > 35 [16:26] oh wait! :) [16:26] cyphermox: most would agree :) === _LibertyZero is now known as LibertyZero [17:07] didrocks, just curious but why a Breaks on unity-common rather than a versionned depends? [17:08] didrocks, usually depends create less upgrade and apt resolver issues [17:08] seb128: will change that for next time, I was tired, indeed a dep will be better, staging the change [17:08] didrocks, thanks ;-) [17:08] seb128: thanks for notifying :) [17:08] yw! [17:10] didrocks, are those wanted? [17:10] - const string PREFIX = "/usr"; [17:10] + const string PREFIX = "/usr/local"; [17:10] didrocks, in the config.vala [17:11] didrocks, just asking because I crossed them while looking the Breaks and that looked weird [17:11] not sure if that's a static config or will be picked up at build by a configure option or something [17:11] ? [17:12] seb128: it's taking the real one on rebuild IIRC [17:12] didrocks, excellent ;-) [17:18] w00t - https://code.launchpad.net/~chrisccoulson/messagingmenu-extension/expose-main-pref/+merge/68722 [17:19] i need french translation though ;) [17:19] (and everything else) [17:19] great [17:19] kenvandine, gwibber is being weird still [17:20] seb128, wait for the next upload and see if it is still weird [17:20] :) [17:21] kenvandine, ok, it has count in the indicator not matching what is the ui [17:21] is "in" the ui [17:21] but yeah, will try the next version [17:22] the count in the indicator is since the last time you clicked it [17:22] well it says "10" and I just started it and I've like 3 new posts since I closed it [17:22] kenvandine: gwibber doesn't use gtk2 any more right? [17:22] jbicha, the gwibber client doesn't [17:23] gwibber-accounts still does [17:23] i might try to port that with gi [17:23] but keep it python for now [17:23] \o/ [17:23] would be nice ;-) [17:23] i would like to make it use gtk.application [17:23] should be easy enough for the account dialog [17:24] yeah [17:24] so to get rid of the wnck dep, i am going to take a swing at that [17:24] ok, because the libgwibber-gtk2 confused me :-) [17:24] soname... [17:24] we have lots of packages like that [17:24] libindicate-gtk2... [17:25] wait that's still gtk2 :p [17:25] you mean version 2 of libgwibber, not the gtk2 version? [17:25] seb128, that reminds me... i've merged libgwibber into the gwibber source... so we don't need the libgwibber source package anymore [17:25] ok, great [17:25] seb128, and there was a bug filed that libgwibber FTBS [17:25] there is a ftbfs bug, recycle it to a "please drop the source" [17:25] cool [17:26] and assign to who? [17:26] that's confusing [17:26] jbicha, i know... we have lots of packages like that [17:27] kenvandine, subscribe archive-admin [17:27] ok [17:27] ubuntu-archive I mean [17:28] kenvandine, how come you merged it back? (just curious) [17:28] it never was [17:28] it is far easier to develop in the same tree [17:28] oh ok [17:28] and gwibber depends so tightly on it now [17:28] I though it was gwibber code splitted in a lib at first [17:29] yeah, makes sense to keep them in sync in the same source [17:59] hrmm [17:59] are the ubuntuone packages not part of the ~ubuntu-desktop upload privileges domain? [18:08] i guess https://wiki.ubuntu.com/DesktopTeam/Components is outdated [18:41] Ok, who broke my Launcher and Menu panel? :> [18:41] * BigWhale bursts into tears. [19:03] Uhm... [19:03] bigwhale@thefish:~$ setsid unity [19:03] bigwhale@thefish:~$ execvp: No such file or directory [19:04] Am I doing something wrong? :/ === Zdra-n900 is now known as xclaesse [20:10] tedg, are you the best person to review indicator-messages merge requests? [20:10] i assigned you as reviewer for https://code.launchpad.net/~chrisccoulson/indicator-messages/dont-hardcode-evolution/+merge/68741, but i'm not sure how busy you are ;) [20:10] chrisccoulson, best looking, does that work? [20:10] tedg, yeah, i'm using it here now [20:11] it's nice to have the Mail entry pointing to Thunderbird :) [20:11] chrisccoulson, I'm planning on going through merge requests today, but I've gotten waylayed on some HR stuff :-/ [20:11] tedg, sure, no problem === m_conley is now known as m_conley_away [21:12] so RAOF, I knew toilets drained backwards when flushed in Australia, but didn't realize waterfalls also went the opposite direction down there. http://www.bbc.co.uk/news/world-asia-pacific-14217256 [22:15] bryceh: Oh! You've discoverered our secret anti-gravity water research, then? :) [22:16] apparently! [22:19] Man, that's totally awesome! [22:49] morning everyone [22:50] anyone know why I can't load unity3d this morning? getting pushed to 2d after an update. Also, my nvidia driver seemed to get removed without me knowing...which might have something to do with it, but I can't tell yet ;) [22:51] what causes apport-gtk to popup when apport writes out a report file? it seems like something must be monitoring /var/crash, but whatever it is doesn't seem to be in apport-gtk itself... [22:59] jasoncwarner_: Good morning! I strongly suspect your two issues are the same - Didier was complaining about nvidia not being built against the -5 kernel yesterday, too. [22:59] thank, RAOF...am I then in a 'wait until nvidia builds against the right kernel' scenario? [23:01] jasoncwarner_: I've not investigated why nvidia isn't building against the right kernel. Once it does Unity3d will probably work for you again, though. [23:04] RAOF: alright. if you find anything, i'd be interested to know when I can use it agian! :) [23:19] bug 814346 [23:19] Launchpad bug 814346 in xorg "xorg-common depends on xdiagnose which pulls in all of gtk3" [Undecided,New] https://launchpad.net/bugs/814346 [23:35] jbicha, hmm good point I hadn't considered kubuntu [23:36] jbicha, alright we'll move xdiagnose down to ubuntu-desktop so it doesn't pull gtk into kubuntu [23:36] bryceh: cool, it was mentiond in #ubuntu+1 so I just helped it along [23:38] bryceh: 1 guy on the forums was upset that Ubuntu was forcing him to install apport [23:39] apport-gtk looks to be only a recommends on ubuntu-desktop so perhaps xdiagnose should be too [23:42] jbicha: apport-gtk, or just apport? [23:43] Because I think it's unreasonable to complain that apport is installed. [23:43] (And will be even more unreasonable when we get a crash database) [23:44] RAOF: I agree that apport is very useful, however he does have a point that this is a change to force it [23:44] recommends does the same thing, right? [23:44] Yes, but with some very important corner-cases. [23:45] well it doesn't matter too much to me either way, I just thought I'd bring up the extreme-minority [23:45] viewpoint :-) [23:45] I'm frankly surprised that apport isn't a dependency of *-desktop already ;)