[00:00] robert_ancell: with sudo in a terminal possibly, but not sure really. [00:00] The background could be set after the gdm user is created in the postinst. [00:01] Or, that may be better done with the package that has the theme. [00:01] hrm [00:01] I am interested in doing this for studio as well, so I am giving this some thought. === asac__ is now known as asac [00:05] I think we need to abstract theme settings for gdm out somehow... [00:05] So that derivatives can set things the way they want. === onestone__ is now known as onestone === Pici` is now known as Pici === WelshDragon is now known as Fluffles [06:01] Good morning [06:53] Morning pittil. [07:06] hi.. all is alpha3 out? [07:10] Not yet [07:11] TheMuso: ah... thanx... isnt it expected today? [07:14] mac_v: right, but not that early yet [07:14] need to wait until the USians get up, and do some announcement prep before [07:14] ok... ;p [07:19] /c/c [07:20] pitti: I see there are currently no studio alpha images on the tracker to test. Has the tracker even been updated for ubuntu/kubuntu etc images yet, or is it just studio that hasn't been updated? [07:20] Sorry, should have asked that in -devel. [08:12] good morning o/ [08:17] good morning there [08:17] hey seb128, hey didrocks [08:17] hey pitti [08:18] hi seb128 and pitti [08:18] lut didrocks [08:18] seb128: can I ask you to be a patch monkey once more? http://people.canonical.com/~pitti//tmp/gpm/ [08:19] seb128: we are down to 5 distro specific patches now, plus one dubious one [08:19] at the beginning of karmic we had some 25, it's much more manageable now [08:19] pitti, sure [08:20] * pitti hugs seb128 === tkamppeter__ is now known as tkamppeter [08:20] * seb128 hugs pitti back [08:28] seb128: cool, Richard added a gconf key for the suspend failure notification, so that patch of ours will become a trivial schema change now [08:29] nice working with him [08:29] you are doing a good job at working with upstream too ;-) [08:29] thanks [08:29] ie them being nice is mostly due to you rocking ;-) [08:30] hey slomo [08:32] hi seb128 :) [08:32] hey slomo, how are you? [08:33] slomo, I've been playing with totem and DVDs yesterday, works quite fine in karmic now [08:33] slomo, I will try the one which has the sound issue with your patch a bit later today [08:33] oh, speaking of totem, it's somewhat sad that the only video that we ship in example-contents doesn't get along with the default codecs :/ [08:33] slomo, one issue is that the sound, languages entries are not changing nor working ... is that a known issue? [08:34] and searching for it fails and just says something like "gstreamer input blabla" [08:34] pitti, the input thing should be fixed with the gstreamer syncs I did yesterday morning [08:34] ah, cool [08:34] pitti, are you sure you are uptodate? [08:35] seb128: I didn't test it with the very latest ubuntu desktop CD [08:35] will do again [08:35] let me try [08:35] brb switching to my laptop for testing that [08:36] doing as well [08:39] pitti, the spirit of ubuntu one plays fine on the current iso there [08:40] seb128_: confirmed; as usualy, you're ahead of me :) [08:40] lol [08:40] good to see if fixed ;-) [08:41] pitti, btw do you know if there is any move on the polkit devicekit-disks auth dialog at login thing? [08:41] pitti, the upstream bug didn't change for some weeks and davidz said he was blocking on landing his gvfs changes which he did since [08:41] ie could use a small nudge maybe? [08:41] yeah, perhaps [08:45] pitti, seen on #commits [08:45] martin.pitt * r9286af907243 gnome-power-manager/src/gpm-button.c: Make hotkeys work with modifiers [08:45] martin.pitt * r70c79e645544 gnome-power-manager/data/ (2 files): Show .desktop files in XFCE, too. [08:45] \o/ [08:45] seb128_: merci [08:45] you're welcome [08:46] now if somebody could fix gpm to not dim the light after 15 seconds not touching my laptop and not bring it back that would be nice [08:46] hi pitti :) i'm fine but busy as always ;) [08:46] I don't get how that bug can be around for so long with nobody caring [08:46] seb128: no idea if that's a known issue but the dvd language, etc selection needs some work in any case [08:47] I keep fighting the autodiming [08:47] seb128_: new gdm is due on Monday, but I'm on vac; I'd like to do the update, but only on Wednesday, is that okay? [08:47] slomo, ok, and do you know if tracks should be listed in the playlist? [08:47] Amaranth: hi... in regards to your comment on bug 402821 , i'v added a comment , https://bugs.launchpad.net/hundredpapercuts/+bug/402821/comments/3 , do you foresee any problems ? [08:47] Ubuntu bug 402821 in hundredpapercuts "Alt-Tab of a minimized window only shows icon" [Undecided,Confirmed] [08:47] Launchpad bug 402821 in hundredpapercuts "Alt-Tab of a minimized window only shows icon" [Undecided,Confirmed] https://launchpad.net/bugs/402821 [08:47] seb128_: since I want to re-do one of the patches to use the new gconf key which Richard just added [08:48] pitti, s/gdm/gpm ? [08:48] seb128_: yes [08:48] pitti, no problem, nobody here is touching gpm usually [08:48] mac_v: It's a hack that upstream will not accept [08:48] it was ted's and is not really maintained since [08:48] seb128_: ah, so I guess it's stuck on me now :) [08:48] pitti, indeed ;-) [08:49] pitti, btw, http://git.gnome.org/cgit/gdm/commit/?id=384afa69ba2120a32f585fa419dd0fe0b3a8a23d [08:49] speaking about gdm [08:49] seb128_: no, they're not listed in the playlist, that's intentional (you don't want to know about the tracks as they have nothing to do with the user visible structure of the dvd as seen in the dvd menu) [08:49] pitti, I think that's something you discussed with robert_ancell yesterday [08:49] mac_v: currently such a hack would be easy to maintain (but somewhat difficult to implement) but if upstream gets active again it would be a load of work to maintain as well [08:49] mac_v: Until upstream decides how to do it I don't see it happening [08:49] Amaranth: oh ... why they wont accept? or any links to discussions regarding this? [08:49] mac_v: because it's a hack [08:49] slomo, ok, otherwise good work on gstreamer and totem to all of you who working on the changes, it works great ;-) [08:49] seb128_: weird, shutdown/reboot works fine in gdm for me [08:50] seb128_: but indeed we need it if we want to add gdm configuration to the gdm session, so great [08:50] pitti, probably because you have no user logged in [08:50] Amaranth: ah... ;) , could we do this in Ubuntu alone for now ? [08:50] seb128, interesting :) [08:50] mac_v: the thumbnail plugin does it because the thumbnail plugin is controller by one person [08:50] hey robert_ancell [08:50] mac_v: read everything I said [08:50] hey [08:50] Amaranth: oh ok... [08:50] seb128_: ah, of course [08:50] robert_ancell: hey [08:50] robert_ancell, I never know if you are still around, you are so quiet ;-) [08:50] mac_v: If you want to implement such a hack and rebase it frequently go right ahead [08:51] going back to my desktop now [08:51] hey pitti [08:51] mac_v: I suspect by the time you got it implemented in a way that wouldn't require poking it every time upstream makes any changes we'd have a proper solution [08:51] mac_v: also, it uses a _lot_ of memory [08:51] seb128, I prefer to call it "sneaky" [08:51] mac_v: fine in the thumbnail plugin since it isn't enabled by default, not so good in the default install [08:52] mac_v: thus, hack :P [08:52] Amaranth: memory ! ok.. I'll see if i can work something out [08:53] ah... no thunbnails in default... forgot that! [08:53] mac_v: think 1MB or so for every window you have open [08:53] o.o [08:53] and that's for a tiny 512x512 thumbnail === apachelogger is now known as birthdaylogger [08:54] iirc to get good looking thumbnails for the thumbnail plugin you have to boost it to 1024 [08:54] so 4MB [08:55] :( [08:56] or maybe that was for the alt-tab switcher [08:56] most likely [08:56] such a hack was implemented in beryl [08:56] the "proper" solution is to have minimize not unmap the window [08:56] the wastes memory too and makes minimize kind of worthless though [08:57] Amaranth: the present behavior is actually not helpful... Also the M$ guys have done it, so i think we can do it better ;p ... i'll try to hack it and see if it doesnt eat memory... will report back *if* i get a good result [08:58] robert_ancell, right ;-) how are you? enough things to do to keep you busy? [08:58] mac_v: you could on-the-fly convert the thumbnails to png to save room, compiz has a png plugin (enabled by default) that could be used to display them again [08:58] seb128, busy at the moment. Been bogged down getting my head around PolicyKit. The links you guys sent yesterday were a big help, thanks [08:58] mac_v: but you either have to hack core or each plugin needs to save a copy of the thumbnail [08:59] Amaranth: ah... i think developing a separate plugin with such a behavior might be easier ;p [09:00] mac_v: you still have to modify every plugin that wants to use it [09:00] :( [09:00] robert_ancell, you're welcome [09:01] mac_v: compiz and the plugins are so intertwined it becomes a bit of a pain maintaining such patches too [09:01] mac_v: and of course you can at that point no longer forward most compiz bugs upstream [09:02] I hate those blueprint whiteboard change emails [09:02] they would send you the diff [09:02] robert_ancell, what did you change in the gnome3 whiteboard? ;-) [09:02] Amaranth: could you comment this on the bug, for the record, this seems beyond papercut scope [09:02] pls [09:03] seb128, I checked on some upstream packages and one had gtkbuilder already applied in git :) [09:03] it was vino [09:03] robert_ancell, ok thanks [09:04] really not easy to read what changed launchpad just you the whole whiteboard text [09:04] yeah it doesn't email a diff does it? [09:05] robert_ancell, no, just a "Whiteboard changed to" with the whole text [09:06] perhaps the list should move to the wiki page itself? [09:06] pitti, if I can stay away from wiki the better [09:06] the thing is just so slow it's ridiculous and I manage to screw formatting all the time [09:07] ok [09:07] tagged bug reports might work better [09:07] pitti, btw do you want me to change the whiteboard to have each package to change as a work item? [09:07] with the wiki having a canned link to the list [09:07] seb128: I don't particularly mind; don't waste too much time on it [09:08] ok [09:08] I need to chase mvo too [09:08] he's using a weird simpleglade weird thing in his softwares [09:08] he's a gtk hater or something ;-) [09:09] not sure if we should look to convert this things to gtkbuilder [09:09] or just make those applications use gtkbuilder directly [09:09] mac_v: done [09:09] I've seen this thing, but never used it myself [09:09] pygtk is simple enough already :) [09:09] same here [09:09] Amaranth: thanx :) [09:09] especially loading glade files is no rocket science [09:10] we had some concerns yesterday with rick about quickly ubuntu-template licensing (not quickly core itself). As this code is copied and then modified by users to create their own project and we don't want to enforce GPL-V3 to them, what licence should we use? (we don't want to pollute the header of each .py files either and enable people to change it if needed). Bsd-like license would fit, how to make it non intrusive in headers? [09:13] didrocks: public domain? [09:14] BSD 2-clause should also work, yes [09:22] seb128: hm? [09:22] mvo, you use weird things in your softwares! [09:22] mvo, you are a freedom hater I can tell it ;-) [09:23] seb128: its a wonderful piece of technology ;) [09:23] mvo, joke aside you have some weird wrapping class around pygtk glade apparently [09:23] seb128: and makes porting trivial [09:23] and I'm not sure what to do about it in this gtkbuilder world [09:23] seb128: yeah, SimpleGladeApp - now there is SimpleGtkbuilderApp, just sed s/ the one with the other and your port is done [09:23] switching applications from glade to gtkbuilder is trivial [09:23] but switching all this magic I'm not so sure [09:24] oh ok, didn't know that somebody did that already, good! [09:24] switching with the wrapper is even more trivial ;) [09:24] * seb128 hugs mvo [09:24] seb128: take it from e.g. gebi [09:24] seb128: what are you looking to port? [09:24] mvo, right, I was expecting that, I just didn't know there was a wrapper yet for gtkbuilder [09:24] mvo, language selector [09:25] seb128: just go ahead :) if you have any questions, let me know [09:26] hello [09:26] mvo, ok, will do [09:26] pitti: do you have any example of headers for public domain license? [09:26] lut crevette [09:27] didrocks: not really; just say "This file is in the public domain", shouldn't that suffice? [09:28] pitti: I'm more up-to-date on gpl and CC licenses. If you say it should suffice, that's perfect ;) Thanks a lot [09:29] IANAL [09:29] ok, I google it a little too to seek for examples :-) [09:38] pitti, I know you have a really busy week but any chance you manage to fix the retracers today or give me some clues how to do so? [09:40] would be nice to not waste all the alpha3 crashes now that apport is running by default [09:40] seb128: I don't think it's that easy [09:40] I would probably do something like this: [09:40] - reproduce the issue locally in the apport/chroot.py test suite [09:41] - strace the failed test run [09:41] - add some debugging printf() into fakechroot's preloadlib, to see what it's doing [09:41] - fix whatever is wrong [09:41] it might be a new eglibc system call which needs wrapping (it's mostly a copy&paste thing) [09:42] it seems a bit specific to gdb, so I'd pull the gdb source to check the code it uses for opening the binary and debug symbols [09:42] seb128: an easier try would be to check if it works to use jaunty's gdb on karmic and apt-pin gdb for now [09:43] pitti, ok thanks, seems not trivial, I might have a look but no guaranty [09:43] that I can try [09:43] seb128: perhaps try the gdb downgrade first [09:43] doing that now [09:43] pitti, otherwise what do you think about turning apport off again until retracers are fixed? [09:43] the real fix sounds like a day of work, but I'm afraid I won't get to that until next week [09:43] new GNOME is due on monday [09:44] seb128: ok, let's do that if the gdb downgrade doesn't work, ok? [09:44] ie we will probably outdate all the crash before they get retraced [09:44] deal [09:44] I'm trying the downgrade thing now [10:01] bye all [10:01] bye rodrigo_ [10:01] oops [10:02] he's quick to close IRC when he says bye [10:02] pitti, jaunty gdb works! [10:02] \o/ [10:02] so login --save, downgrade, apt-pin and close? [10:02] and restart retracers? [10:02] seb128: so, I guess rebuilding gdb against eglibc, or the new gdb itself changed something then [10:02] seb128: right [10:02] * pitti hugs seb128 for rescuing the retracers [10:03] pitti: bye, but I just arrived :) [10:03] * seb128 hugs pitti for giving good hints [10:03] then I can also remove my hideous and broken workaround [10:03] hey rodrigo_ :) [10:03] pitti, let me try on the amd64 one and restart it to see how it goes [10:05] hmm, did I change some setting, or is firefox in karmic using a google custom search?? [10:06] -? [10:06] it's all very well and good, but there are no links to cached versions of pages, which is a shame [10:09] * pitti points Ng to asac [10:13] Ng: https://bugs.edge.launchpad.net/ubuntu/+source/firefox-3.0/+bugs?field.searchtext=&orderby=-importance&search=Search&assignee_option=any&field.assignee=&field.bug_reporter=&field.bug_supervisor=&field.bug_commenter=&field.subscriber=&field.status_upstream-empty-marker=1&field.omit_dupes.used=&field.omit_dupes=on&field.has_patch.used=&field.has_cve.used=&field.tag=multisearch&field.tags_combinator=ANY [10:13] Ng: if your bug is not there, file one and add th "multisearch" tag please [10:13] ok ;) [10:13] -;)+:) [10:13] Ng: against ffox 3.0 (also add ffox 3.5 task) [10:13] Ng: thanks! [10:14] * Ng confirms with a blank ff profile (mine is *very* far from default ff ;) [10:15] Ng: no need to. the "multisearch" extension was added for alpha3 [10:15] Ng: disable the addon to get back to default behaviour [10:16] asac: oh, I just looked at the extension list and completely failed to see it :/ [10:20] http://people.canonical.com/~bryce//upstreamer-0.png [10:21] (tool to facilitate forwarding bugs upstream from launchpad to freedesktop) [10:22] bryce: awesome! [10:29] pitti: still a work in progress, but at least it looks nice finally :-) [10:30] bryce, how does it work? [10:30] bryce, does it take the launchpad description, etc? [10:31] bryce, I start on a very basic one some time ago http://people.canonical.com/~seb128/lptognome.py [10:32] that one takes a lp bug number, get the title, description, stacktrace and open a bugzilla page (GNOME) with title and description set [10:32] it's done in a hackish way though, using ?title=... etc in the bugzilla url [10:35] seb128, yeah what's in that screenie is unedited; that's what it generates after loading the bug # [10:35] seb128, it grabs the description, title, submitter email address, attachments, etc. from the launchpad bug [10:36] I've not done the logic to submit the multiple attachments to bugzilla yet though... that might be tough [10:36] bryce, using greasmonkey? [10:36] didrocks: "otherwise the effect will be drawned" -- what is that supposed to mean? [10:36] in fact I've not tested it at all against bugzilla yet... [10:37] seb128, no greasemonkey isn't powerful enough; this is a full cgi script [10:37] ah ok [10:37] any reason to do that in a web ui way? [10:38] the version I started sucks because it's buggy but I sort of liked the "give a bug number and get the upstream bugzilla page prefiled with infos" [10:38] ie I'm not sure we win anything to edit infos downstream, you can do that on bugzilla before clicking submit [10:38] "any reason to do that in a web ui way?" > don't understand question [10:39] ahh right [10:39] bryce, your screenshot seems to indicate you do editing on the launchpad side [10:39] where my approch was a "collect everything and open a bugzilla page prefiled with infos and edit there" [10:40] the specific things this would save is that it parses/reformats the description, and it would allow you to select the attachments to be included via checkboxes, rather than individually having to download/upload each attachment individually [10:40] that avoid the need for a server to run the forward tool ;-) [10:40] I guess we are a different workflow [10:40] GNOME bugs have almost never a file to attach [10:41] and I inline stacktrace for easier bug reading and dup finding too [10:41] ie what we usually need on GNOME bug is a copy of the description and the stacktrace if there is one [10:41] but your tool seems powerful indeed ;-) [10:42] ah yeah, your attachment situation is simpler (just the backtrace) [10:42] well I like having logs inline in fact [10:42] * bryce nods [10:42] makes easier to read the bug and search in the page than having lot of logs [10:43] but if you have lot of log that's doesn't really scale [10:43] seb128, gnometris is not longer a default game in ubuntu? [10:43] true, however for X, Xorg.0.logs can be pretty long. But I do like to put stacktraces inline. [10:43] * seb128 declares victory on the retracer [10:43] pitti, bug #403036 \o/ [10:43] Launchpad bug 403036 in scim "scim-launcher crashed with SIGSEGV in scim::FrontEndModule::load()" [Medium,New] https://launchpad.net/bugs/403036 [10:43] seb128: *sniff* beautiful! /me wipes a tear [10:44] pitti, ;-) [10:44] that takes the pressure out of this thing [10:44] right [10:44] seb128, rickspencer3 also came up with a table to map launchpad importance to bugzilla priority/severity, so this does that mapping automatically [10:44] will do the i386 one too now [10:44] bryce, nice [10:45] bryce, I expect you could easily add GNOME bugzilla to your service if the bugzillas behave the same way? ;-) [10:45] pitti, want me to open a bug somewhere about the gdb situation, I'm not sure what the downgrade working tells us [10:46] pitti, I will try on the i386 one the new gdb version from karmic which was built before eglibc to see if that makes a difference [10:46] seb128: against fakechroot, please, as a reminder (assign it to me) [10:46] ie if the issue started with the new gdb or after rebuild [10:46] seb128: it would be interesting to see if it happens with jaunty's gdb built on karmic [10:46] I can do some playing around that [10:46] I rather suspect it's a glibc change, but it could be a gdb change as well (unusual open() ) [10:46] seb128: thanks [10:46] you're welcome [10:47] just throw it all in the bug report, I'll read it when I'm back from vac [10:47] ok [10:47] seb128, well I intend to focus just on freedesktop, but yeah most bugzillas work similarly. I'd definitely accept patches to make it work against additional bugzillas. Of course I'll have to prove it actually useful first :-) [10:47] bryce, let me know when it's in a working state for fd.o I will have a look then [10:48] thanks! [10:48] sure :-) [10:49] (unfortunately the tool is also slow... python + launchpadapi != quick) [10:52] (actually, it's mostly the attachment processing that slows it down) [10:58] pitti: hum, difficult to formulate... "Otherwise, we will be considered immediately outside of the current stenza". Is it clearer to you? [10:59] pitti: thanks for the merge [10:59] didrocks: I just removed it, the code should be clear enough [10:59] didrocks: I have another MP, after that I'll do a new upload [10:59] pitti: thanks :) === yofel_ is now known as yofel [11:06] pitti: before you release, think about udpating python-distutils-extra version from 2.7 to 2.8 in debian/local/python-mkdebian [11:07] didrocks: yep [11:17] dpkg-source: info: applying gdb_6.8-3ubuntu2.diff.gz [11:17] *** stack smashing detected ***: patch terminated [11:17] ======= Backtrace: ========= [11:17] /lib/libc.so.6(__fortify_fail+0x4b)[0xf7f2b97b] [11:17] /lib/libc.so.6(__fortify_fail+0x0)[0xf7f2b930] [11:17] /home/ubuntu-archive/apport-retracer-i386/fakelibs/libfakechroot.so[0xf7f9fb44] [11:17] [11:18] pitti, going to be trickier than expected, I think I will just downgrade, pin and restart for now [11:18] I can't dpkg-source -x the jaunty gdb in the retracer environment [11:18] wow [11:44] grrr [11:44] ubuntuone-client starts being very annoying [11:44] they fixed a nautilus crashed but in exchange of a new one apparently [11:51] pitti, and i386 retracer running and confirmed to do proper retracing [11:51] seb128: good job, thanks! [11:51] you're welcome [11:52] thanks for the tips on how to get those running again [11:59] pitti, oh, are SegvAnalysis and SegvReason new? [11:59] I never noticed those before [11:59] seb128: yes, courtesy of kees [11:59] nice [11:59] he does some analysis from the disassembly [12:09] didrocks: 2.8 uploaded to sid, karmic, jaunty and hardy PPA === MacSlow is now known as MacSlow|lunch [12:37] slomo, the gstreamer change you suggested didn't fix the dvd sound issue [12:38] seb128: : hi... the recent update seems to have removed the volume applet and left the notification icon... is that change to stay? [12:38] mac_v, yes [12:39] mac_v, as described in the update changelog [12:39] seb128: and the volume slider is only going to move up-down? [12:39] mac_v, that I know [12:39] you would expect it to be a 3d clutter cube moving in the z axis too? [12:40] * mac_v confused [12:40] seb128: actually i asked since there was a bug asking for the slider to be up-down rather than horizontal [12:41] mac_v, good, one bug to close then [12:41] :) [12:52] mvo, any idea how gksu crashes can have vorbis code in their stacktrace? [12:53] ie bug #403005 [12:53] Launchpad bug 403005 in gksu "gksu crashed with SIGSEGV" [Medium,Incomplete] https://launchpad.net/bugs/403005 [13:00] seb128: Mind syncing devhelp to fix the FTBFS? [13:00] seb128: no :) [13:00] lool, will do [13:00] seb128: gksu invoking gstreamer to check for codecs? [13:01] mvo, doh, gksu have over 100 bugs in launchpad and lot of recent crasher looking corruption [13:01] Or gksu invoking update-manager invoking gstreamer in a postinst [13:01] lool, because a sudo wrapper need codecs to be fun to use? ;-) [13:01] :-) [13:01] seb128: thakns for the sync [13:01] you're welcome [13:09] thanks a lot pitti :) [13:21] hey pedro_ [13:21] pedro_, retracers are fixed and running again now [13:22] seb128, yay! thanks for the info ;-) === MacSlow|lunch is now known as MacSlow [13:53] seb128: dude! [13:53] pitti, alter! [13:53] bug 123020, look at the status [13:53] Launchpad bug 123020 in soyuz "support shipping verbatim files in the exported tarballs" [High,Fix released] https://launchpad.net/bugs/123020 [13:53] \o/ [14:00] good afternoon everyone:) [14:01] hey chrisccoulson [14:01] hi pitti [14:05] hey pitti [14:06] i just got the update of devicekit-power from the ~ubuntu-desktop ppa [14:06] it seems to have broken suspend [14:07] not sure if it was the removal of devicekit or the changes to devicekit-power that broke it [14:10] kenvandine: devikekit shouldn't matter, it's not used any more [14:10] probably dk-power [14:10] kenvandine: did you reboot after update? or sudo killall devkit-power-daemon ? [14:15] i did reboot [14:16] gnome-power-cmd suspend [14:16] says i can't suspend [14:17] pitti, i am happy to help debug... but it isn't urgent... just wanted you to know it might be broken before you move it to karmic :) [14:17] kenvandine: does gnome-power-manager work? I don't think -cmd still works [14:17] * kenvandine is hacking on desktopcouch atm... which is rather fun :) [14:18] pitti, -cmd has been working [14:18] kenvandine: ok, thanks for the warning; I'll test it here as well [14:18] i use it all the time [14:18] i don't get an option to suspend on shutdown either [14:18] kenvandine: http://git.gnome.org/cgit/gnome-power-manager/commit/?id=c5e34acfbc1b57903bb18d84575b1beaa67e3344 [14:19] kenvandine: right, neither do I; I think this needs a new gnome-power-manager (port to polkit-1 as well) [14:19] probably [14:19] porting the bakcend (dk-p) without the frontend (g-p-m) won't work, I figure [14:19] makes sense [14:19] kenvandine: probably needs http://git.gnome.org/cgit/gnome-power-manager/commit/?id=af19ef18b93abb7f92cb6b1345cf4c633e6f8513 [14:19] the notify-osd patch was accepted to g-p-m the other day :) [14:19] kenvandine: anyway, new g-p-m will release next Monday [14:20] kenvandine: hm? I just did a g-p-m patch cleanup with Richard, and he rejected it [14:20] oh? [14:20] * kenvandine looks again [14:20] "better not apply until this is official gnome" [14:20] kenvandine: I committed some of our patches to upstream, and removed some obsolete ones [14:28] hey chrisccoulson, manager to finish work at a reasonable hour after your early waking up today? ;-) [14:29] dobey, kenvandine: could somebody look at ubuntuone-clients bug and try to bump priority on fixing nautilus crashers there? [14:29] kenvandine: anyway, new g-p-m is due on Monday, I'll package it on Wednesday when I'm back, and test it together with the new dk-p [14:29] dobey, kenvandine: 80% of the nautilus crashes we get are due to ubuntuone issues [14:29] sure [14:30] thanks [14:30] * kenvandine goes searching [14:32] seb128: the only crash i am of aware of being in ubuntuone is fixed. if there is another one, i don't know how to make it happen [14:33] dobey, are you subscribed to ubuntuone-client ubuntu? [14:33] dobey, the usual way when you "don't know how to make it happen", is to ask details to users, not to ignore the bugs [14:33] seb128: yes, and i see the reports, but there is no explanation of what is happening [14:34] dobey, they several crashes in g_object_unref with debug stacktrace sitting there [14:38] pitti, looks like richard did merge that patch... after rewriting it [14:38] bratsche: have you checkout out git master for g-p-m? [14:38] kenvandine: oh? http://git.gnome.org/cgit/gnome-power-manager/commit/?id=d2251e9373add05113623c0a812a86bdd35d9116 looks related, but it doesn't look like notify-osd [14:39] http://bugzilla.gnome.org/show_bug.cgi?id=423186 [14:39] Gnome bug 423186 in gnome-power-manager "Battery info shouldn't be a notification" [Minor,Resolved: fixed] [14:39] commit 3c721b77404f59e23ab712c099b0c179fd1e491c [14:39] well, so much the better [14:39] yup [14:39] kenvandine: ah, *that* patch [14:39] i'll try to make sure it does the job before he releases [14:40] I meant http://bazaar.launchpad.net/%7Egnome-power-manager-team/gnome-power/trunk/annotate/head%3A/debian/patches/02-notify-osd-support.patch [14:40] oh... that is different :)( [14:41] he won't take that [14:41] seb128, wow... there are alot of crashers [14:42] lots are old [14:42] kenvandine, yes :-( [14:42] would be nice to send back the way the message that software quality is important [14:42] ie somebody should look at crashers [14:43] yup [14:44] hey seb128 - yeah, i managed to finish work at lunchtime today:) [14:44] chrisccoulson, good ;-) [14:44] i might sleep for a bit this afternoon ;) [14:45] good choice [14:45] seb128, is there an easy way to find all the crashers for people with u1 installed? [14:45] kenvandine, no [14:45] figured :/ [14:46] kenvandine, I reassign the nautilus one by looking at the stacktrace [14:46] ok, so you have been assigning them to u1? [14:47] kenvandine, ubuntuone-client source package on launchpad yes [14:47] it would be nice if the nautilus apport hook figured out what packages provide nautilus extensions, and ran the apport hooks for all of them as well [14:47] ok [14:47] (and same for any other packages which provide some pluggable interface) [14:48] dobey, that should be doable I think, maybe pitti has some pointer on that [14:48] then at least i could tell which versions of what everyone is using, without having to ask individually and hope that they actually respond [14:50] dobey, the g_object_unref crash are using the current karmic version for sure [14:50] but yes, that would be nice to have the versions for everything shipping something in /usr/lib/nautilus [14:52] dobey: good idea [14:52] dobey, btw [14:53] dobey, apt-cache showsrc gnome-python-desktop [14:53] right, the nautilus apport package hook could just iterate over the plugin directory, find the package with apport.packaging.get_file_package(file), and then call apport.hookutils.attach_related_packages(report, package_list) [14:53] dobey, karmic has each component splitted in a binary, and the only reason we still have python-gnomeprint and libgnomeprint* on the CD is ubuntuone-client [14:54] u-c uses gnomeprint?? [14:54] seb128: ok, i didn't know it was split already in karmic [14:54] no [14:54] dobey, that's why I opened the bug ;-) [14:54] I was looking at why we still had libgnomeprint* on alpha3 [14:55] pitti, no, it depends on python-gnome2-desktop rather than python-gnomekeyring etc [14:55] how do i specify the python depends correctly then? since dpkg doesn't seem to pick it up [14:55] aah [14:55] pitti, so it triggers python-gnomeprint which depends on libgnomeprint* [14:55] seb128: right, I understand [14:55] dobey, add the ones you need in the Depends line of control [14:56] there is no magic that I know to make the list [14:56] seb128: there isn't some special ${python:Blah} thing? [14:56] grep for the import in your source and look at the matching binaries [14:56] not that I know no [14:56] :( [14:57] there is no real way to guess depends with python [14:57] ie you could grep for imports but you don't know if you will find code used or example or things not built, etc [14:58] eh? the way rpm does it works fine afaik [14:59] dobey, which is? [14:59] rickspencer3, hey [14:59] phew, a3 is done [14:59] seb128: but i guess i can do (python-gnomekeyring || python-gnome2-desktop) right? [14:59] rickspencer3: good morning [14:59] * seb128 hugs pitti, good work! [14:59] hi seb128 and pitti [14:59] dobey, yes [15:00] dobey, or rather | [15:00] "depend | alternative" [15:00] seb128: empathy has had the weird habbing of starting up with my user name set to someone else :( [15:00] yesterday I was Cody [15:00] rickspencer3, urg? [15:00] waouh, that's weird [15:00] today I was Quintasan [15:01] rickspencer3: was it the last person you talked with? [15:01] seb128: it greps the installed python files for imports, and adds Requires: Python:module or something like that, and all the python module packages end up doing Provides: Python:module or similar [15:01] didrocks: nope [15:01] it seems pretty f'ing random [15:01] :) [15:01] seb128: it also does similar things for PERL, C#, etc... [15:01] dobey, well you will get depends coming from examples or code you are not using, etc [15:01] * rickspencer3 will log bug this morning [15:02] seb128: i think it skips $docdir and only looks in binary/lib/whatever [15:02] everybody hug pitti for getting alpha3 out [15:02] pitti, \o/ ;-) [15:02] seb128: i already have lots of dependencies for code i don't use, anyway... :) [15:02] congrats pitti [15:02] nice job, a very solid releas [15:02] thanks :) [15:02] * didrocks hugs pitti [15:02] like the majority of kernel modules installed on my system [15:03] this alpha is better than final releases for certain other distros I could mention, but won't ;) [15:03] dobey, I think there is some ongoing work to do a depends system this way in debian, didn't follow what they do recently though [15:03] rickspencer3, you just went to far and are not credible now :-p [15:04] kenvandine, no I haven't, why? [15:04] too [15:05] hrmm [15:06] lol [15:06] bratsche: he asked for you to test it :) [15:06] i just installed ubuntuone-client-gnome from karmic proper and did nautilus --quit like 15 times with no crashers :( [15:06] kenvandine, oh he did? I missed that. :) [15:06] he rewrote your patch :) [15:07] dobey, there could be a difference between a proper --quit and a GNOME session close [15:08] Yeah, I knew he rewrote the patch because he rewrote that entire source file or something. [15:09] seb128: well i got a crash with trunk, but the line #s are different betwen that and the reports, so i can't tell who is running what, and where the crash actually is... even with trunk gdb is telling me the crash is in the line in class_init(), which is obviously wrong [15:32] blah, now i can't make trunk crash [15:37] pitti, https://bugs.edge.launchpad.net/ubuntu/+bugs?field.tag=gnome3-gtkbuilder [15:37] how do i prevent a user from appearing in the new gdm? [15:37] lower uid? [15:37] seb128: ah, nice! much easier to read/autoclose [15:48] ugh, now rhythmbox requires brasero :( [16:10] seb128: is there a supported way to hook into the gnome-session logout process (asking for the updates on shutdown spec) [16:11] mvo, not idea, maybe vuntz or chrisccoulson know better [16:12] pitti, around? I've some build issue and I think it's due to python-distutils-extra [16:12] thanks, I keep wondering if it shouldn't just start its own xsession or something [16:12] mvo, start a new session on logout? [16:13] yeah, a empty one with just the unattended-upgrade-gtk process [16:14] mvo, you could as well patch gnome-session? [16:14] mvo - i don't think there's currently anyway to do what you want to do (without patching gnome-session) [16:15] thanks chrisccoulson and seb128 - when I finished with the actual code, I look into it [16:16] hi [16:16] * rickspencer3 feels like himself again ;) [16:16] wb rickspencer3 [16:29] pitti, unping that was an issue there [16:43] oh, thanks apport for automatically ignoring nautilus crashing :( [16:44] it ignores after 3 crashes [16:44] to not send the same crash 3 zillion times [16:48] unfortunately apport is the only way i can debug nautilus [16:49] how do i make it not do that? [16:59] dobey, clean /var/crash [16:59] dobey, the counted is in the .crash [17:02] thanks [17:57] seb128: hi [17:58] pitti, hello again ;-) [17:58] /away break [17:58] * seb128 ported usb-creator to gtkbuilder [18:02] anybody running kwin or xfce here? [18:02] MacSlow, not me [18:13] MacSlow: try in #kubuntu-devel, likely to be running kwin there [18:14] rickspencer3, bingo [18:15] MacSlow: really nice job on the notify-osd updates [18:15] looks great [18:17] rickspencer3, well thanks.. but a lot of clean up is still to do [18:17] MacSlow: glass half full! [18:17] * seb128 wants text he can read back [18:17] otherwise good work ;-) [18:18] seb128, I've the "green light" for gconf-keys sort of [18:18] \o/ [18:19] rickspencer3, lazy bunch on #kubuntu-devel :) === spc_ is now known as spc [19:05] \o/ patch cleanup day [19:05] yesterday gdm, now consolekit [19:06] they committed all our patches (8), now just 4 debian patches left === mac__v is now known as mac_v [19:18] kenvandine: ah, just saw your couchdb patch; please test it, while dpkg usually retains permissions, it might not do that for conffiles (they are magic and special in lots of ways) [19:19] kenvandine: so while the fix is correct, it might not fix the issue on upgrades (that'd need a postinst snippet) [19:45] I'm off for today, and on holiday until next Tuesday [19:46] have a nice weekend everyone, and try not to break everything so hard :-P [19:46] * pitti -> packing [19:46] I'll disable my IRC proxy, so please give me a phone call on urgencies; I'll also look into my mail every day [19:47] pitti, enjoy! [19:47] pitti, don't worry we will watch for you while you are not there ;-) [19:48] * pitti hugs the team [19:53] * seb128 hugs pitti [20:01] rah, just lost of 3 pages email because evolution crashed on me, damn spell checker [20:07] bug 386139 [20:07] Launchpad bug 386139 in evolution "evolution crashed with SIGSEGV in free() when spell check replaces words (dup-of: 389101)" [Medium,New] https://launchpad.net/bugs/386139 [20:07] Launchpad bug 389101 in evolution "evolution crashed when correct new message" [Unknown,Fix released] https://launchpad.net/bugs/389101 [20:07] Fix released?? [20:10] fta, the fix released status is coming from the upstream task not from the one in ubuntu [20:11] http://bugzilla.gnome.org/show_bug.cgi?id=584243#c4 [20:11] Gnome bug 584243 in Editing "Crash when replacing a mispelled word" [Critical,Resolved: fixed] [20:11] said that the fix was applied to gtkhtml 3.27.5 which is not yet available on Ubuntu [20:13] oh, i've been hurt by that for a while now, any eta to have that in ubuntu? [20:13] seb128, ^ ? [20:14] pedro_, GNOME 2.27.5 due next week [20:14] so before next tuesday [20:14] ok, thanks [20:14] you're welcome [20:15] fta, you should have got a draft of your email though [20:15] it should ask after next start [20:19] i also a bunch of tiny regressions with my desktop, for example, when i untray rhythmbox, it no longer re-appears on the front if it was trayed lower, it's confusing [20:22] i have a few keys bound to play / pause / next etc (it's not a multimedia keyboard, so i mapped F5/F6/F7/F8), if rhythmbox is idle for too long (not sure it's the cause though), all the bound keys no longer do anything [20:22] xchat systray does not appear on 1st run (new gnome session), i've already mentioned that here [20:23] and a few other tiny-but-annoying regressions.. [20:25] fta, patches are welcome [20:25] if i knew what it was, i would not even ask [20:28] fta - i had a look at your xchat issue a couple of weeks ago [20:28] i can't figure it out though. it works with an old gtk version [20:29] but i can't see what xchat does differently to other applications that work correctly [20:29] (i assume we're thinking about the same issue here anyway) [20:29] chrisccoulson, i remember. it seems xchat is the 1st to appear on the desktop, while the panel is still empty [20:31] fta - i ran xchat with GTK_DEBUG=plugsocket, and then killed the panel. you see some messages when the notification area disappears, but xchat never notices when it appears again [20:31] i did the same with nm-applet, and saw some debug messages everytime i killed and reloaded the panel [20:31] i've had a look through the gtk code, but i can't figure out whats going on [20:32] so, i got a bit stuck there [20:37] i will add a wrapper and see when i next restart my desktop [21:11] re [21:11] chrisccoulson, your g-s-d upgrade still has 2 files listed in dh_install --list-missing, one .glade and one .xml [21:12] hmmm :-/ i'll look at that now - i should do these updates when i'm more awake;) [21:38] seb128 - i've pushed g-s-d with the missing files included now [22:01] I usually use auto-login in gdm so I didn't notice this before. To test a bug I disabled auto-login and did a re-login. When I fist saw the greeter I didn't know how to login as I just had 3 buttons but no user to select which took a few seconds to appear. is that normal? a notice would be great that gdm is collecting the user list or so [22:02] geser, there is some bugs about the list being too slow to be computed [22:02] I don't think we need a different bug for the animation [22:02] ok [22:02] either we make it fast or add a visual clue but that's the same bug [22:03] as long as it's known, it's fine for me [22:31] kenvandine: in reference to your code review ... [22:31] damn it! === Quintasan_ is now known as ricspencer3 [22:32] kenvandine: ok, it' [22:32] ricspencer3, I understand what you meant now [22:33] hi seb128_ [22:33] you mean about empathy making me impersonate people? [22:33] ;) [22:33] ricspencer3, you did a typo in your nickname ;-) [22:33] ricspencer3, yes [22:35] seb128_: did it happen to you? [22:35] or am I the only one using empathy for IRC? [22:35] ricspencer3, no but I'm not using an IM for IRC, sorry I know we should be testing what we ship but I need a decent IRC client ;-) [22:35] heh [22:35] well ... I'm testing it [22:36] I'll try to repro the bug when I have chance [22:36] open a bug even if it lacks information [22:36] it always sets me to the same two people, Code or Quintasan [22:36] maybe upstream will have a clue [22:36] will do [22:42] note to self, suspend on this desktop doesn't ask for password on the login screen and crash the box [22:44] what IRC client do you use seb128? [22:44] chrisccoulson, xchat-gnome [22:44] i haven't tried that. i just use pidgin for everything currently [22:44] i probably get testing empathy really [22:45] I don't like using IMs to do IRC [22:45] i haven't tried xchat-gnome before, so i don't know the features of a decent IRC client [22:45] I think that im clients are made to send messages to people [22:46] you don't want the same interface for IRC, you want a user list, IRC functions, not thing to pop up on screen as private messages do etc [22:46] yeah, that makes sense [22:47] i certainly don't want the same interface for channel-based chat, as for private messages [22:50] hmmm, yeah, xchat-gnome looks ok [22:50] i could probably get used to it;) [22:52] chrisccoulson2, I'm not trying to convince anybody, I was just using IRC clients before any of my IM client was doing IRC [22:52] and I get used to have different workflows, etc for those === asac_ is now known as asac