=== Amaranth_ is now known as Amaranth === asac_ is now known as asac === robbiew1 is now known as robbiew [07:29] Good morning [07:35] Hey pitti. [07:35] hey TheMuso, had a nice weekend? [07:37] pitti: Very much so thanks. Yourself? [07:37] pretty well, thanks! [07:38] TheMuso: could you please have a look at bug 681412; can you confirm that this isn't a regression, and that gksu has always worked that way? [07:38] Launchpad bug 681412 in at-spi "Can not enter password for Administrative tasks using Onboard Keyboard" [Medium,Triaged] https://launchpad.net/bugs/681412 [07:43] pitti: Oh I thought I commented on that... I really don't know to be honest, will take another look. [07:43] TheMuso: thanks [07:46] pitti: Oh there is a gconf key that can be set to prevent gksu from grabbing keyboard et al focus, its in the assistve technologies preferences window. [07:46] TheMuso: so perhaps if a11y is on, it shouldn't do that by default? [07:47] TheMuso: so grabbing the keyboard/mouse is the reason why you can't use the on screen keyboard, right? that's how I understood it [08:04] pitti: I suspect so. [08:11] Morning all! [08:12] hey Sweetshark, good morning [08:17] good morning [08:17] bonjour didrocks [08:17] Guten Morgen pitti, how was you week-end? [08:18] didrocks: pretty nice; we went ice skating again, and we celebrated the Bachelor title of my wife (so had a full house) [08:18] didrocks: how was your's? [08:20] pitti: it was fine as well. Still assembling some furnitures (for the kitchen this time) and long walks in Lyon :) [08:20] didrocks: do you like your new home city? [08:20] pitti: yeah, it's so better than Paris! I'm really happy to be back to Lyon. Really feel "home" again :) [09:07] pitti: do I uderstand bug 725763 correctly? That on Ubuntu we want --with-arm-target=7 on armel too? [09:07] Launchpad bug 725763 in libreoffice "FTBFS on armel, built for ARMv4, should be ARMv7" [Undecided,New] https://launchpad.net/bugs/725763 [09:08] Sweetshark: I have no clue about arm, but if Jani says it, I guess it's correct [09:08] Sweetshark: btw, sorry for the oo.o hiccup on the weekend [09:09] Sweetshark: so you want to build the transitional from a separate oo.o source, but modify them to work for ubuntu? or put them back to libo? [09:09] pitti: np, I should have been more clear about that [09:11] hey pitti Sweetshark [09:11] pitti: I want to build the transitionals from a oo.o source (ubuntufied). This keeps our diff to debian in libreoffice smaller, also it keeps the logic out of the libreoffice package (which has already a way too huge debian/*) [09:13] good morning everyone [09:13] Sweetshark: *nod* [09:13] hey chrisccoulson [09:13] chrisccoulson: had a nice weekend? [09:14] hey pitti. my weekend wasn't too bad thanks, although i did quite a bit of work on the firefox menu [09:14] how are you? [09:16] chrisccoulson: pretty well indeed, thanks [09:26] hey pitti, Sweetshark [09:26] hey chrisccoulson [09:26] pitti, I'm trying to restart the retracers, they crashed on fridays again [09:27] *sigh*, thanks [09:27] hi seb128, how are you? [09:27] I'm fine thanks [09:27] how are you? [09:28] trying to see the fallout from last week dx updates so we can get things fixed today [09:28] so first step is to have things retraced [09:33] monirng [09:36] pitti: I have a fix for the arm issue, which I unfortunately cannot very easily here (no arms or legs around here yet). Should we dput a 3.3.1-1ubuntu4 to see what the buildd say about it? Also: I tried to do a libreoffice-l10n build, but could not upload it. The reason seems to be because the changes file generated from a changelog (via debuild -S) with two different packages in it was a bit confusing for dpkg-genchanges .... [09:36] seb128 - yeah, i'm good thanks. had quite a productive weekend! [09:36] lut huats === ogra_ is now known as ogra [09:46] Sweetshark: for -arm, yes, let's upload it [09:47] Sweetshark: for -l10n, I'm afraid you need to ask doko how to generate this correctly [09:49] is there a policy in Ubuntu regarding which apps should or should not be visible in the applications menu? [09:50] there's an onboard bug in papercuts that requires it to be shown, but I am not sure if omitting it is part of having less clutter in menus [09:50] pitti: so I will upload to the ppa again and you take it from there? [09:50] didrocks, seb128 ^ [09:50] Sweetshark: or just put to chinstrap again [09:51] janimo: are you talking about a GNOME menu or an unity place? [09:51] janimo, there is policy but I think this one was hidden on purpose for less menu clutter [09:51] pitti: ah, ok. Will do. (There are only i386 and amd64 in the ppa anyway ...) [09:51] didrocks, /usr/share/applications .desktop file, wherever they are read [09:51] at least classic gnome IO [09:51] UI [09:52] seb128, this is one of those interesting apps, onscreen keyboard cannot really be launched via typing in alt-f2 window if user cannot type [09:52] so the request is valid, but I am not sure it has been discussed already outside the bug in LP [09:53] janimo, I think there was a button to start it in the a11y control center dialog or something [09:54] ok [09:54] pitti: btw, how does the (closes: #XXX) magic distinct between debian and launchpad bugs? does it? [09:54] Sweetshark: it doesn't -- for Luanchpad it's (LP: #XXXX) [09:55] pitti: yup [09:55] pitti: yuck [09:55] * Sweetshark fixes changelog [09:57] morning [10:04] pitti, could you check the amd64 retracer crash? [10:05] KeyError: 'DistroRelease' [10:05] broken bug, I'll untag [10:05] pitti, it seems an apport bug or rather lack or tolerance to invalid formatting [10:06] pitti, ok, I asked in case you wanted to make apport just skip those rather than crash or something [10:06] yeah, it shoudn't completely crash like that [10:06] pitti, thanks [10:06] right, I'll try that [10:06] ProblemType: CrashDistroRelease: Ubuntu 11.04 [10:06] bah [10:14] good morning [10:14] seb128: hey [10:15] hey kklimonda [10:15] seb128: could you add the comment to my application? If not, I may end up having to drag you to the meeting ;} [10:15] oh right, what was the wiki url again ? [10:15] https://wiki.ubuntu.com/KrzysztofKlimonda/MOTUApplication [10:16] ok [10:17] thanks [10:20] seb128: crash fixed in trunk, and amd64 lock removed [10:20] pitti, you rock, thanks! [10:26] "gnome-keyring-daemon[3295]: couldn't allocate secure memory to keep passwords and or keys from being written to the disk" ??? [10:31] chrisccoulson, is bug #724202 still on your list? [10:31] Launchpad bug 724202 in libdbusmenu "nautilus crashed with SIGSEGV in image_notify_cb()" [Medium,New] https://launchpad.net/bugs/724202 [10:32] chrisccoulson, it just got a retracing with a debug stacktrace, I commented on the bug [10:32] seb128 - yes, but i'm struggling to figure out what might cause that [10:32] i'll take a look at the new trace though [10:32] chrisccoulson, http://launchpadlibrarian.net/65275023/Stacktrace.txt [10:34] pitti, the stacktrace quality is excellent since the gdb upgrade btw [10:34] seb128: that's great! [10:34] so it seems I was chasing phantoms all along [10:34] yes, that's much better1 [10:34] ! [10:34] oops ;) [10:34] chrisccoulson: shift keys are overrated [10:35] pitti: libreoffice_3.3.1-1ubuntu4_source.changes is on chinstrap [10:35] pitti: caps lock is cruise control for cool? [10:36] seb128, didrocks: hi, shouldn't bug 717600 be assigned or something? [10:36] Launchpad bug 717600 in nux "compiz crashed with SIGSEGV in std::_List_node_base::_M_hook()" [Undecided,Confirmed] https://launchpad.net/bugs/717600 [10:38] fta2: should be nice if we can get a retraced stack [10:38] (and we use milestone to target, before assignement) [10:39] didrocks, i can upload my crashfile from earlier this morning if it helps [10:40] i even have it resolved locally [10:40] fta2: yes please :) [10:40] if it's resolved, you can just put on this bug report :) [10:41] chrisccoulson, bug #725064 is collecting some duplicates [10:41] seb128: Error: Could not parse data returned by Launchpad: list index out of range (https://launchpad.net/bugs/725064) [10:42] chrisccoulson, is that yours? [10:42] couchjs crash [10:42] seb128 - yep [10:42] ABI break (again) ;) [10:42] didrocks, hmm, did you notice that unity FTBFS on armel ? [10:42] chrisccoulson, can you take the bug assignement then? ;-) [10:42] just like with every single upload i've done this cycle :( [10:42] seb128 - yeah, i can take that [10:42] ogra, it's fix commited on launchpad [10:43] ogra: of course, I opened a bug even [10:43] bug #724615 [10:43] and yeah, it's fixed [10:43] Launchpad bug 724615 in unity "unity FTBFS on armel" [High,Fix committed] https://launchpad.net/bugs/724615 [10:43] ogra, ^ [10:43] ah, sweet, thanks [10:43] chrisccoulson: yeah, oneconf is broken as well :/ [10:44] yeah, this is getting annoying now :( [10:44] well, won't be an issue next cycle [10:44] moving away from desktopcouch seems to be the only solution [10:44] is that just for oneconf? desktopcouch isn't going away is it? [10:45] chrisccoulson, right [10:45] it will just solve it for didrocks [10:45] not for you ;-) [10:46] chrisccoulson: sorry for the false hope :-) [10:46] i want a separate libmozjs package really, but i think there's going to be push-back from the security team [10:46] hmpf, it says fix committed, but doesnt say where :P [10:46] if we can't have that, then we really need to just drop everything from the archive that depends on it ;) [10:47] * ogra wishes jani would link his fixes to bugs :/ [10:48] ogra: ask for jani, he doesn't link his branch to it. it's not in trunk even [10:48] ogra, http://bazaar.launchpad.net/~unity-team/unity/trunk/revision/894 [10:48] didrocks, ^ it is? [10:48] yes [10:49] seb128: ok, I need glasses when looking at bzr visualise :/ [10:49] ah, thanks [10:50] ok, couchdb uploaded [10:50] ogra, there will be an unity update in natty today btw [10:50] ogra, so no need to bother about backporting [10:51] seb128, thanks, though i still would prefer if he would link his fixes :) [10:52] ogra, right, tell him please so he fixes that [10:52] yep, on it [10:57] didrocks, done [11:18] Sweetshark: uploaded, thanks! [11:20] fta2: thnks [11:25] didrocks, hello, the unity commit to FTBFS arm is 894 to trunk. I thought bzr would link the bug since I used LP: #xxxx and did not check if it did [11:25] janimo: no, I have a wishlist opened against bzr for that [11:25] janimo: you should --fixes lp: [11:26] didrocks, thank will use that next time. Will it set it to Fix commited even id the arg name is --fixes? [11:26] as that is the corrdct bug status [11:26] janimo: that would be nice, but no :) [11:26] in this case commits to trunk are not fit for --fixes, only if trunk of debian packing I guess [11:26] janimo: so, you still have to manually ensure you set it to fix committed and ping us to set it to next milestone target (so that we close it automagically) [11:26] I'lll do it manually [11:27] yay automation :) [11:27] ok I'll do that next time [11:27] janimo: thanks a lot and thanks for the fix :) [11:27] heh [11:27] np, will keep an eye on armel related stuff or ping me if I seem to miss one [11:28] janimo: sure, can I assign you armel FTBFS bugs if we get some? [11:28] janimo: as now, I'm opening one as soon as it fails only on armel [11:28] didrocks, most definitely [11:28] janimo: ok, will do then! Thanks :) [11:29] thank you too [11:29] nux as well [11:29] yeah, all "unity-related" [11:54] Sweetshark: (*#$(#U( hrmpf armel [11:55] debian/rules:1270: *** Recursive variable `CONFIGURE_FLAGS' references itself (eventually). Stop. [11:55] Sweetshark: ^ hmm, this doesn't look very armel specific, though? [11:59] pitti: it is [11:59] * Sweetshark is to blame. [12:01] why the heck does CONFIGURE_FLAGS have deferred expansion. *grmbl* [12:16] Sweetshark: ? [12:16] doko_: pong [12:16] Sweetshark: http://launchpadlibrarian.net/65277930/buildlog_ubuntu-natty-armel.libreoffice_1%3A3.3.1-1ubuntu4_FAILEDTOBUILD.txt.gz [12:16] debian/rules build [12:16] debian/rules:1270: *** Recursive variable `CONFIGURE_FLAGS' references itself (eventually). Stop. [12:16] dpkg-buildpackage: error: debian/rules build gave error exit status 2 [12:17] doko_: https://pastebin.canonical.com/44016/ [12:19] doko_: see also: [12:19] 13:01 < Sweetshark> why the heck does CONFIGURE_FLAGS have deferred expansion. *grmbl* [12:19] ;) [12:19] Sweetshark: this is wrong too [12:19] elaborate [12:20] lucid has v6, versions before v5. please just use the defaults, no need for an extra option === MacSlow is now known as MacSlow|lunch [12:23] How do I disable the globalmenu thing? It plays very, very poorly with focus-follows-mouse. [12:24] doko_: next try: https://pastebin.canonical.com/44017/ [12:24] I tried setting both of /apps/gnome_settings_daemon/gtk-modules/globalmenu-gnome-panel and /apps/gnome_settings_daemon/gtk-modules/globalmenu-plugin to false, but it made no difference. [12:25] + ifeq "$(ARCH)" "armhf" [12:25] CONFIGURE_FLAGS += --with-arm-target=7 [12:25] + endif [12:25] Sweetshark: still wrong [12:26] why the option? [12:26] doko_: read carefully, it is still in the diff [12:26] s/diff/if-endif/ [12:28] Sweetshark: I did. still wrong. why hard code the target for armhf? [12:29] doko_: if debian wants that for their build i aint gonna stop them [12:48] njpatel, heyo, saw that both my datetime/ido branches got merged [12:50] hey mterry [12:50] seb128, hello! [12:50] mterry, dbarth rolled an indicator-datetime tarball and I said we would backport the ido commit we need [12:51] seb128, OK, I can do those now unless someone else is on it [12:51] Sweetshark: I don't care about Debian, but it should default to sanity for an Ubuntu armhf port [12:51] seb128, pitti kindly approved FFe [12:51] mterry, nobody is so feel free [12:52] mterry, then we have quite some libdbusmenu and friends crashers [12:52] ;-) [12:52] hrm [13:00] hi mterry [13:01] seb128: so you need to ffe bugs for both ido and the indicator, right? [13:01] dbarth, hello, thanks for doing some indicator-datetime work for me [13:01] dbarth, I did last week, it's approved [13:01] mterry: that's fine dude, that was not the hardest part of the work ;) [13:02] ah ok for the ffe: do you have the links to the bugs, so that i can attach the rev. numbers and tarballs for seb to pick up [13:04] dbarth, bug #725121 [13:04] Launchpad bug 725121 in indicator-datetime "Feature Freeze Exception for indicator-datetime" [Undecided,Confirmed] https://launchpad.net/bugs/725121 [13:04] dbarth, https://bugs.launchpad.net/ubuntu/+source/ido/+bug/725121 for both [13:04] Launchpad bug 725121 in indicator-datetime "Feature Freeze Exception for indicator-datetime" [Undecided,Confirmed] [13:04] Need to debug a networking bug I forget who to ask [13:05] libido is up [13:18] Sweetshark: hmm, for natty and newer, I think the patch can be dropped. the builtins are properly expanded with the linaro gcc, and in fsf gcc-4.6 [13:19] pitti: https://bugs.launchpad.net/ubuntu/+source/ido/+bug/725121/comments/2 [13:19] Launchpad bug 725121 in indicator-datetime "Feature Freeze Exception for indicator-datetime" [Undecided,Confirmed] [13:19] seb128, mterry ^^ [13:19] so this one is done hopefully [13:20] what's next [13:20] dbarth, right [13:20] * dbarth goes looking for some more [13:20] doko_: ok, so I will drop the patch and try a build on natty. Do we have to do backports for maverick/lucid on arm at all? [13:21] Sweetshark: do we *have* to make backports on other archs? [13:22] I guess not. My inbox tells me a different story ... [13:28] seb128, dbarth: both uploaded now [13:29] mterry, excellent, thanks! [13:29] mterry, do you have things to work on next? [13:29] seb128, I can take some crash bugs, but I should do piloting for the second half of today, unless the crashers are real bad? [13:30] mterry, can you upload https://code.edge.launchpad.net/~mterry/libappindicator/fix-crash-708188/+merge/51356 to natty? [13:30] seb128, sure [13:30] mterry, so we can see during a3 testing what crashers remain [13:30] mterry, thanks [13:31] mterry, well, the libdbusmenu crashers are annoying since they take down random applications and not only indicators [13:31] mterry, but it's only an a3 so no hurry either [13:31] mterry, btw can you try if pidgin shows up in appmenu for you? [13:31] it stopped working for me some days ago it seems [13:31] seb128, it was on friday... [13:31] seb128, yeah, still works [13:32] mterry, does the account menu lists accounts? [13:32] today it's showing up but the accounts don't [13:34] seb128, it lists the one account I have... [13:34] ok, so probably a local issue, this box is lacking some updates [13:34] mterry, in any case nothing urgent so please do your half a day sponsoring [13:35] mterry, I will let you know if there are crashers we should try to get sorted for a3 [13:35] libdbusmenu has some annoying ones [13:35] seb128, well, point me at bugs so I can work on them until sponsoring [13:35] but since chrisccoulson is working on some the issue you could end duplicating efforts [13:35] ah [13:35] feel free to help out with bug 724202 though ;) [13:36] Launchpad bug 724202 in libdbusmenu "nautilus crashed with SIGSEGV in image_notify_cb()" [Medium,Confirmed] https://launchpad.net/bugs/724202 [13:36] chrisccoulson, do you think you will work on those libdbusmenu issues? [13:36] this ones not making much sense to me atm [13:36] seb128 - yeah, i'm still looking at those [13:36] mterry, ^ if you have any clue about that one [13:36] bug #726153 got some duplicates as well [13:36] Launchpad bug 726153 in libdbusmenu "bluefish crashed with SIGSEGV in dbusmenu_menuitem_property_set_shortcut_menuitem()" [Medium,New] https://launchpad.net/bugs/726153 [13:36] if one of you want to claim it [13:37] no tedg yet? :( wouldn't mind him taking a look at my patch for bug 718926 [13:37] Launchpad bug 718926 in indicator-appmenu "Some apps don't integrate to appmenu after having their windows closed" [High,Triaged] https://launchpad.net/bugs/718926 [13:37] that ones really annoying for empathy ;) [13:38] hello everyone! [13:38] chrisccoulson, he's at scale today [13:38] hi nessita! [13:38] chrisccoulson, he will be back tomorrow [13:38] hi chrisccoulson [13:38] seb128 - ah, ok. thanks [13:38] how are you? [13:38] mterry, ^ can you review chrisccoulson's patch? [13:38] nessita, i'm good thanks, but very busy. how are you? [13:38] chrisccoulson, wait on mterry to ack it and distro patch for a3 if you want? [13:38] hey nessita [13:39] chrisccoulson: pretty much the same ;-) I have to finish closing all UI related bugs to make it to the UI freeze, since I'll be on holidays for 2 weeks :-D [13:39] hey seb128! how are you? [13:39] nessita, I'm fine thanks, what about you? [13:39] pretty good! [13:39] nessita, oh, enjoy holidays! [13:40] chrisccoulson, ok, looking [13:40] seb128: thanks! this Saturday I'm traveling to attend to PyCon USA, which is very exciting [13:41] nessita: sorry the upload got rejected :( [13:41] mdeslaur: is ok, it was my fault. I added the COPYING file directly to the branch package :-/ [13:41] mdeslaur: I fixed that though, shall I wait for Jamie to re-review? [13:42] nessita: it needs to get uploaded again for him to take a look [13:43] nessita: you can ask whoever is on patch piloting today, or else let me know and I'll look at it again [13:43] nessita, it's not really holidays, it's a work conference ! [13:43] mdeslaur: ok, new branch is located at lp:~nataliabidart/+junk/ubuntuone-file-storage-api-0.0.2 [13:43] Sweetshark, hello, will natty have LibO 3.3.X or 3.4 RC? [13:43] nessita: ok, I'll take a look [13:43] seb128: I'm going to the conference by my own, so is kinda holidays :-) [13:44] seb128: and I'm taking a couple of free days as well [13:44] mdeslaur: thanks! [13:46] nessita: there's no debian directory in that branch [13:48] nessita: oh, wait a sec...I'm looking wrong [13:48] janimo: natty is in feature freeze, and 3.4.0 rcs wont show up before 2011-04-04 so: natty will have 3.4.X [13:48] 3.3.X that is [13:49] doh [13:49] Sweetshark, thanks [13:50] (and I would assume the rcs not too stable before branchoff which is 2011-04-18 [13:51] Sweetshark: hi, how does one install LibO in maverick? from the ppa, it seems i have to install each package to get each application.. [13:51] Sweetshark: there seems to be no single meta package which pulls everything for LibO and removes OOo , am I right? [13:52] or maybe I'm not lookin at the right package.. :) [13:53] mdeslaur: don't scare me :-D [13:54] nessita: just one comment so it properly gets processed: could you add "(LP: #723139)" at the end of the debian/changelog "Initial release" line, and could you remove two spaces from the indent of the "- Implemented" part? [13:55] mdeslaur: yessir! [13:55] good morning! [13:55] nessita: I don't want it to get rejected again :) [13:55] vish: the libreoffice package is what you want [13:57] hmm, i tried that ,but seemed to miss some packages.. tries again … [13:57] mdeslaur: so, the - Implemented should be left aligned with the 'I' in '* Initial release'? [13:57] nessita: yes! === MacSlow|lunch is now known as MacSlow [14:01] Sweetshark: this is what i get in maverick » http://paste.ubuntu.com/573459/ [14:04] vish: could you try https://bugs.launchpad.net/ubuntu/+source/libreoffice/+bug/715144/comments/1 and comment on the bug, if that helped? [14:04] Launchpad bug 715144 in libreoffice "Running Lucid (10/04) tried to install Libreoffice via ppa and could not do so...report of broken dependencies and refused to install" [Undecided,New] [14:05] * vish tries.. [14:14] mdeslaur: pushed to the same branch (revno 6) [14:21] nessita: ok, uploaded...let's see if it works this time :) [14:21] mdeslaur: thanks! [14:29] * pitti feeds the hamsters to make webkit build faster [14:37] chrisccoulson, I'm suspicious that bug 724202 is really a theme_changed_cb bug [14:37] Launchpad bug 724202 in libdbusmenu "nautilus crashed with SIGSEGV in image_notify_cb()" [Medium,Confirmed] https://launchpad.net/bugs/724202 [14:37] Sweetshark: nope.. that dint help. i removed ure, uno-libs3 , libhyphen0 , and everything i got for search "openoffice" except libhunspell-1.2-0.. and still it doesnt install [14:37] mterry, yeah, me too [14:37] the trace is weird [14:37] chrisccoulson, everything but the top of the stacktrace is about the theme changing. image_notify_cb has no right being called given the rest of the stack [14:37] mterry, yeah, that's the bit i've been struggling to make sense of [14:39] chrisccoulson, I just pushed a new libappindicator for the theme_changed_cb bit. I'd say let's assume it's the same bug until a new crash gets filed with the new libappindicator [14:40] mterry - i think i already fixed a similar bug like that in dbusmenu too :/ [14:40] mterry, you fixed that? awesome! [14:40] unless there's somewhere else that the signal handler should be being disconnected [14:43] chrisccoulson, I thought the usage of that signal was safe in dbusmenu... maybe I didn't fix it correctly in libappindicator if you found something wrong with the dbusmenu code === dbarth is now known as dbarth-otp [14:44] chrisccoulson, oh, that was because you fixed it [14:44] that was a while ago, I see [14:44] ok, everything is good [14:44] :) [14:44] mterry, yeah, i did this one last week: http://bazaar.launchpad.net/~dbusmenu-team/dbusmenu/trunk/revision/231 [14:44] that *should* ensure it gets disconnected now though :/ [14:44] i am really amazed how fast shotwell starts... [14:46] does gtk_icon_theme_get_default always return the same GtkIconTheme? [14:47] chrisccoulson, I looked it up, and it does [14:48] chrisccoulson, well, as long as the default screen doesn't change? [14:50] chrisccoulson, bug 726153 looks like a classic case of not disconnecting when the dbusmenuitem goes away [14:50] Launchpad bug 726153 in libdbusmenu "bluefish crashed with SIGSEGV in dbusmenu_menuitem_property_set_shortcut_menuitem()" [Medium,New] https://launchpad.net/bugs/726153 [14:52] mterry, yeah, bug 726153 looks like quite a simple fix :) [14:52] Launchpad bug 726153 in libdbusmenu "bluefish crashed with SIGSEGV in dbusmenu_menuitem_property_set_shortcut_menuitem()" [Medium,New] https://launchpad.net/bugs/726153 [14:53] chrisccoulson, I'm making a patch [14:53] right, i need to grab some food before my head explodes :) [14:53] mterry, get it reviewed by chrisccoulson when you get it and backported to natty ;-) [14:53] let's try to get crashes fixes in if we can [14:54] so we see from a3 what remains to work on [14:54] hey kenvandine [14:54] hey seb128 [14:57] chrisccoulson, bug #725759 [14:57] Launchpad bug 725759 in evolution "evolution crashed with SIGSEGV in dbusmenu_menuitem_property_set_variant()" [Undecided,New] https://launchpad.net/bugs/725759 [14:58] is another crasher [14:58] bug #725980 [14:58] seb128: Error: Could not parse data returned by Launchpad: list index out of range (https://launchpad.net/bugs/725980) [14:59] chrisccoulson, please review: https://code.launchpad.net/~mterry/dbusmenu/disconnect-more-signals/+merge/51550 [15:01] pitti, bah, bug #719861 is not fixed, I hate it [15:01] Launchpad bug 719861 in gdk-pixbuf "After installation icon theme default to gnome-icon-theme and cannot be changed" [High,In progress] https://launchpad.net/bugs/719861 [15:04] seb128: would it help to add a few additional dependencies to libgdk-pixbuf2.0-0, such as libgl1 and librsvg2-common, to ensure that the latter are configured first? [15:04] pitti, it would be wrong, the gdk loader doesn't depend on those [15:04] pitti, you might have configs where people use gdk without gl or svg [15:05] just on pixmaps [15:06] seb128: it's weird that it broke due to ldconfig [15:06] pitti, see comment #9 [15:06] libraries usually don't need that to work, unless they aren't in /usr/lib/ [15:06] chrisccoulson, my branch will also fix bug 725980 [15:06] Launchpad bug 725980 in libdbusmenu "evolution crashed with SIGSEGV in dbusmenu_menuitem_property_set_variant()" [Medium,New] https://launchpad.net/bugs/725980 [15:06] seb128: yeah, I was reading that [15:06] pitti, libGL is not in /usr/lib [15:07] /usr/lib/mesa/libGL.so.1.2 [15:07] oh, that [15:07] ah, has a /usr/lib/mesa/ld.so.conf [15:08] seb128: so as far as I understand, the issue is that it doesn't catch the libGL based loaders, because libGL isn't available yet? [15:08] seb128: what about adding a trigger interest in /etc/ld.so.conf.d/ ? [15:09] pitti, why? [15:09] pitti, the update I didn't which didn't work makes the libgdkpixbuf run to not be delayed [15:09] seb128: that woudl cause /usr/lib/gdk-pixbuf-2.0/gdk-pixbuf-query-loaders to be called again if a library like libGL gets installed -- wasn't that the problem? [15:09] so the libgdkpixbuf post installation should have it [15:10] but I thought that'd run too early? [15:12] pitti, well I though the loader list was built before the libGL unpack and the ldconfig which is why I didn't find it [15:12] but it seems it's unpacked after [15:13] so I'm not sure why now [15:13] mterry - your change looks good. do you want me to comment on the merge proposal too? [15:14] seb128: perhaps we should temporarily add an extra /usr/lib/gdk-pixbuf-2.0/gdk-pixbuf-query-loaders to stderr to the postinst, for debugging? we can then check the output in the livefs build log [15:14] chrisccoulson, sure. I'll push a distro-patched version up [15:16] mterry, ok, done [15:17] pitti, one way to workaround it would be run the gdkpixbuf update from the librsvg postinsts rather than relying on a trigger which is not set on first install [15:18] seb128: hm, I still don't understand the bug then -- librsvg2-common ships the svg loader in the dir that libgdk-pixbuf2.0-0 triggers on [15:18] pitti, yes [15:18] pitti, but the trigger doesn't work, see my second bullet point in #comment 2 [15:18] ups [15:18] comment 9 [15:19] it's a first install issue only [15:19] librsvg is unpacked before the trigger is set [15:19] so it will not trigger [15:19] seb128: hm, why doesn't librsvg2-common depend on any libgl package? [15:19] because it doesn't use libgl [15:19] $ ldd /usr/lib/gtk-2.0/2.10.0/engines/libsvg.so|grep libGL [15:19] libGL.so.1 => /usr/lib/mesa/libGL.so.1 (0x00007ffa6c5be000) [15:19] seb128: it does [15:20] $ objdump -x /usr/lib/gtk-2.0/2.10.0/engines/libsvg.so | grep GL [15:20] $ [15:20] $ objdump -x /usr/lib/libcairo.so | grep GL [15:20] NEEDED libEGL.so.1 [15:20] NEEDED libGL.so.1 [15:20] ah ok; so, bad linking/shlibdeps then [15:20] pitti, it's coming through libcairo since we build it with the gl backend (needed for wayland) [15:21] but librsvg2-common doesn't depend on libcairo either [15:21] no, it depends on gtk [15:21] which depends on cairo [15:22] $ objdump -x /usr/lib/gtk-2.0/2.10.0/engines/libsvg.so | grep NEEDED === dbarth-otp is now known as dbarth [15:22] seb128, you didn't merge this https://code.launchpad.net/~rodrigo-moya/ubuntu/natty/nautilus/fix-leak/+merge/51307 right? [15:22] ok, so at the time when librsvg2-common gets configured, libcairo and libGL should both be configured [15:23] but apparently the trigger runs after unpack, not at configure? [15:23] doko_: http://libreoffice.pastebin.com/etTyZc5R <- patch for master, looking good? [15:24] pitti, well I think the trigger watches files on disk, since librsvg is unpacked before libgdk-pixbuf is configured (which is when the trigger is set) it doesn't work [15:24] i.e it will not notice the librsvg2-common .so [15:25] pitti, but the libgdk-pixbuf postinst does run the gdk-pixbuf-query-loaders [15:25] seb128: so I guess your suggestion of explicitly re-triggering libgdk-pixbuf2.0-0, or just doing the loader cache update directly is reasonably correct? [15:25] I still don't get why my previous fix didn't work [15:26] # don't delay the ldconfig call, see lp #719861 [15:26] export LDCONFIG_NOTRIGGER=y [15:26] Launchpad bug 719861 in gdk-pixbuf "After installation icon theme default to gnome-icon-theme and cannot be changed" [High,In progress] https://launchpad.net/bugs/719861 [15:26] ldconfig [15:26] seb128: so you are saying that libgdk-pixbuf2.0-0 is already the last pacakge in teh chain that gets configured [15:26] is that correct? [15:26] oooooooooooooooooooooooooooooh [15:26] pitti, no, I say everything is unpacked [15:26] seb128: libgdk-pixbuf2.0-0.postinst doesn't actually update the cache on $1 = "configure" :) [15:27] pitti, since libgdkpixbuf runs ldconfig with nottriger and then do the update it should resolve the libgl [15:27] but certainly it should [15:27] Sweetshark: looks good. I assume the default is implemented somewhere else [15:27] seb128: right, but triggers don't run until the triggered package is configured [15:27] pitti, what trigger? [15:27] chrisccoulson, any other crashers I could be useful with? [15:28] seb128: ah, ignore me; on the bottom of the postinst you have a cache update which isn't guarded by a $1 check [15:28] so that's not it [15:28] rodrigo_, no, I will merge it, thanks [15:28] that looks a bit confusing [15:28] seb128, no hurry, just wanted to make sure, so thanks :) [15:28] pitti, right, it does LDCONFIG_NOTRIGGER=y ldconfig which should make the libGL be found [15:28] pitti, then it runs the cache update [15:28] pitti, or my [15:28] mterry, there is also bug 720895, but it doesn't have a particularly useful trace [15:28] Launchpad bug 720895 in libdbusmenu "bluetooth-applet crashed with SIGSEGV in g_variant_builder_add_value()" [High,Triaged] https://launchpad.net/bugs/720895 [15:28] export LDCONFIG_NOTRIGGER=y [15:28] ldconfig [15:28] is a fail [15:28] doko_: yes, there are two more fallbacks ;) [15:29] seb128: right, but I don't understand why the ldconfig is even the problem here; you said earlier that libgl gets configured earlier than rsvg or pixbuf [15:29] pitti, no, it's unpacked earlier [15:29] pitti, it's configured after [15:29] pitti, depends garanty it's unpacked first, not that it's configured first [15:29] seb128: it does both [15:29] hum [15:30] chrisccoulson, the description of what the user was doing (toggling radio buttons) makes it sound like the toggle_checked callback I just disconnected in my recent merge [15:30] seb128: a depends: b guarantees that a gets configured only after b is configured [15:30] mterry, yeah, i guess that's a possibility too [15:30] chrisccoulson, oh heh. Sorry, not toggling a radio button, but the radio kill switch. Different. :) [15:30] ricotz, ping [15:30] ah [15:31] pitti, ok, so I'm at fail of why we get "libpixbufloader-svg.so: libGL.so.1: cannot open shared object file: No such file or directory" [15:31] perhaps i should try my radio killswitch ;) [15:31] i misread that too [15:31] pitti, by the time the cache update run libgl is installed and ldconfig has been run [15:31] chrisccoulson, the stack that we can see looks like someone passed a NULL string to variant_builder_add or some such [15:31] seb128: ah, does that appear in the livefs build log? [15:32] chrisccoulson, for some dumb reason, that causes a crash instead of interpreting it as "" [15:32] pitti, yes [15:32] pitti, the line after "Setting up libgdk-pixbuf2.0-0 (2.23.0-1ubuntu3) ..." [15:32] pitti, http://people.canonical.com/~ubuntu-archive/livefs-build-logs/natty/ubuntu/current/livecd-20110228-i386.out [15:33] seb128: perhaps because the /etc/ld.so.conf.d/GL.conf file is only installed in libgl1-mesa-glx.postinst [15:33] [15:33] pitti, is it? DOH [15:33] seb128: i. e. calling ldconfig before libgl1-mesa-glx is configured doesn't buy you anything [15:33] pitti, I though it would be in the deb and unpacked [15:34] seb128: no, it's an alternative; so I guess we need a fake depenency [15:34] pitti, or we should just put a cache update in the librsvg post installation script... [15:35] seb128: and as libgl's postinst already does a forced ldconfig, we don't really need the one in pixbuf, I think [15:35] seb128: so you think a GL dependency on rsvg-common is wrong? [15:35] pitti, right, i did that because I though it was unpackaged and not configured [15:36] pitti, well I see no reason svg rendering should require gl [15:36] we didn't have that depends until libcairo got built with the gl backend for wayland [15:36] seb128: ok; then let's take your suggestion; I don't see a better alternative either [15:37] seb128: and sorry for all those questions; took me a while to understand the bug [15:37] pitti, I've no strong opinion either way, we can try putting the depends and see if someone complains [15:37] pitti, oh no worry, I has been taking me a while as well to get the issue, thanks for helping me to sort it, I didn't notice that the libgl ldconfig snippet was set in the libgl postinst [15:38] * pitti hugs seb128 [15:38] * seb128 hugs pitti [15:40] dpm: btw, firefox translations in po2xpi weren't magically fixed in the meantime -- rosetta_xpi_to_sources still throws tons of errors, and the generated de.tar.gz is empty except install.rdf [15:40] dpm: I even have a data/11.04/ firefox de xpi for 4.0, so that's not it; I'll try harder [15:41] (doesn't work in devmode either) [15:48] mterry, do you think bug #725631 is a duplicate of bug #723463? [15:48] Launchpad bug 725631 in indicator-appmenu "unity-panel-service crashed with SIGSEGV in dbusmenu_menuitem_send_about_to_show()" [Medium,New] https://launchpad.net/bugs/725631 [15:48] Launchpad bug 723463 in indicator-appmenu "indicator-applet-appmenu crashed with SIGSEGV in dbusmenu_menuitem_send_about_to_show()" [Medium,Confirmed] https://launchpad.net/bugs/723463 [15:54] mterry, did you see my last comments on bug 720895? [15:54] Launchpad bug 720895 in libdbusmenu "bluetooth-applet crashed with SIGSEGV in g_variant_builder_add_value()" [High,Triaged] https://launchpad.net/bugs/720895 [15:54] i think we should check the return value of gdk_keyval_name, just in case [15:57] chrisccoulson, mmm, yeah probs. or it could be weird fallout of https://bugs.launchpad.net/dbusmenu/+bug/726153 but a check makes sense anyway. You want to whip that patch up? [15:57] Launchpad bug 726153 in libdbusmenu "bluefish crashed with SIGSEGV in dbusmenu_menuitem_property_set_shortcut_menuitem()" [Medium,Fix released] [15:58] seb128, I think so [15:58] mterry - yeah, can do [15:58] mterry, ok thanks [16:08] ok, the gnome3 ppa is now all ok for the gtk3 packaging rename [16:08] hello, was hoping someone could explain nautilus/totem behavior to me. the default media player in ubuntu is totem, so double-clicking an mp3 file in nautilus launches totem. if i right-click on the file->properties, and say, "open with rhythmbox", that seems to change the association properly [16:08] only the *themes packages not building [16:08] when i enter gnome-control-center, i now see that the default media player is rhythmbox. when i double-click a movie (say, .m4v or .mov) in nautilus, the proper thing happens and totem is launched [16:09] pitti, thanks for looking at po2xpi and the FF4 translations and for the update [16:09] rodrigo_, great [16:09] so i guess i'm just trying to understand, where are the preferences stored per file type? i poked around with gconf-editor some and can't seem to find anything interesting [16:09] bigon, ping [16:09] mterry: any clues? [16:10] achiang: you mean, file associations? they are not stored in gconf [16:10] achiang, I know that [16:10] achiang, if you're talking natty... [16:11] well, i'm talking lucid/maverick. [16:11] dobey: where are they stored? [16:11] achiang, I think pre-natty they were stored in gconf [16:11] achiang: in ~/.local/share/applications/defaults.list [16:12] mterry - what do you think we should do about bug 723873 for now? it currently breaks the menus in firefox and thunderbird [16:12] Launchpad bug 723873 in indicator-messages "indicator menu items missing since the recent natty update" [Undecided,Fix released] https://launchpad.net/bugs/723873 [16:12] oh well, url handlers were [16:12] mterry: mime type associations weren't ever stored in gconf [16:12] mterry: url handlers and "preferred applications" were, but that is not the same as the mime associations [16:12] which can be confusing i guess [16:12] dobey, yup [16:12] dobey: ah, you mean mimeapps.list [16:13] oh, yes mimeapps.list [16:13] dobey: got it, so if i wanted to make that default for new users, i guess i could put a file in /etc/skel ? [16:13] and defaults.list [16:14] uhm, i think you would change /usr/share/applications/defaults.list maybe [16:14] oh, i see [16:14] but [16:14] the "default media player" in the preferred applications dialog, really has nothing to do with what happens when you double-click a file in nautilus [16:15] dobey: ok. i don't really want to change that; the real goal is much simpler: "when i double-click an mp3, rhythmbox should launch, other defaults are fine" [16:15] dobey: thanks [16:16] achiang: ok; you are aware that rhythmbox is not installed by default on 11.04 right? [16:16] dobey: yep, that is fine for my purposes [16:16] dobey: i appreciate the help, thanks! [16:16] sure [16:17] chrisccoulson, not sure. how badly does it break them? is that why I can't get to my message filters? [16:17] mterry, yeah, that's the reason you can't access the message filters [16:18] for firefox, it's even worse. you can't restore your saved session ;) [16:18] chrisccoulson, is it a complicated fix do you think? has it started to be looked at? [16:18] i don't think anyone has started looking at it yet [16:18] argh! F10 in gnome-terminal is captured, no matter if I have the option unset in keyboard shortcuts [16:18] nessita, it's captured by unity ;) [16:19] chrisccoulson: bring me back my F10! :-) [16:19] heh :-) [16:19] chrisccoulson: do you know how can I disable that? I can't code without F10 [16:19] nessita, i'm not too sure. you probably can't disable that ;) [16:20] that makes no sense! I don't like people telling me what to do, much less desktops telling me what key I can't use :-D [16:20] nessita, it is captured by appmenu-gtk, based on your keyboard settings [16:20] nessita, but you say it's not respecting that? [16:21] mterry: pressing F10 in gnome-terminal is being captured by the terminal like if I had the option enaled [16:21] chrisccoulson, well, I guess best solution is to fix it, else we should probably disable the menus for A3? [16:21] nessita, only gnome-terminal? [16:21] mterry: is the only place where I use F10... so I can't confirm other apps [16:22] mterry, btw when you do sponsoring you can check the top items on http://people.canonical.com/~platform/desktop/versions.html as well [16:22] nessita, maybe I misunderstand which way the problem is. What did you expect to happen when you press F10? [16:22] seb128, k [16:22] mterry: not to be captured by the terminal so I can use midnight commander as I please :-) [16:22] nessita, and what is happening instead [16:23] mterry: when I press F10 inside gnome-terminal I expect that the terminal itself will not capture, and instead send it directly to the program is running in the terminal [16:24] mterry: the terminal is acting like if the second checkbox here http://ubuntuone.com/p/fJV/ was checked [16:25] mterry: is the issue a bit clearer? [16:25] nessita, right, but I'm curious what IS happening when you press it, just so I understand who is responsible fro the capture [16:25] mterry: the menu for the gnome-terminal is opened (just like that checkbox was checked) [16:26] nessita, ok, cool. I get it. Can you file a bug? I wrote the code to do the F10 bits and I probably didn't consider the case of gnome-terminal overriding its behavior [16:27] nessita, the bug is against appmenu-gtk [16:27] mterry: sure! thanks [16:27] mterry: any idea if I can hack some source file to get F10 back? [16:28] nessita, oh whoops, not appmenu-gtk, but gtk itself [16:29] argh, already filled [16:29] nessita, yeah, you can comment out what I did in my patch if you're comfortable patching ubuntu packages [16:29] nessita, no probs, point me at it, I'll change the bug [16:29] mterry: I'm not comfortable with patching ubuntu packages, but I love sudoing and editing files [16:30] rodrigo_: yes? [16:30] bigon, you did the last update of gnome-shell to the gnome3 ppa, right? [16:30] nessita, nope, not that easy, sorry. :-/ [16:30] dobey: hm, modifying /etc/gnome/defaults.list to make audio/mpeg=rhythmbox.desktop doesn't seem to have any effect [16:30] nessita, you'll just have to nag me to fix it. ;) [16:31] bbl [16:31] mterry: feel yourself nagged! [16:31] :-) [16:31] bigon, seems to be broken -> gnome-shell : Depends: gir1.2-upowerglib-1.0 but it is not installable [16:31] this was broken before [16:31] I've requested a sync for upower [16:32] bigon, ah, ok [16:32] seb128: thanks for the svg fix! (note that you typoed the bug # in the changelog, you need to close manually) [16:32] rodrigo_: https://bugs.launchpad.net/ubuntu/+source/upower/+bug/725229 [16:32] Launchpad bug 725229 in upower "Sync upower 0.9.8-3 (main) from Debian experimental (main)" [Wishlist,Confirmed] [16:32] dobey: ah, ok. i guess i have to do a few of them, like audio/x-mp3 [16:32] I'm waiting archive admin [16:32] pitti, thanks, already closed the bug since I reassigned from gdk-pixbuf to librsvg after upload [16:32] bigon, ok [16:33] mterry: bug #726639 [16:33] Launchpad bug 726639 in appmenu-gtk "F10 is captured even if gnome-terminal setting is explictly set to not capture" [Undecided,New] https://launchpad.net/bugs/726639 [16:34] seb128: will you manage to add the comment in the next few hours, before the meeting? [16:34] kklimonda, yes [16:39] rodrigo_, did you forward your libcanberra 01-dont-use-gtk-quit-add.patch patch upstream when you did it? [16:42] rodrigo_, oh, seems you did add it on https://bugs.freedesktop.org/show_bug.cgi?id=32839 [16:42] Freedesktop bug 32839 in Unspecified "gtk_quit_add removed in GTK+ 3.0" [Blocker,New] [17:02] rodrigo_, how busy are you? [17:04] seb128, yes, filed it [17:04] seb128, 60/70% busy :-) [17:04] seb128, what's up? [17:04] rodrigo_, can you look at updating 91-resolve-symbols.patch in libcanberra to do it properly (i.e configure.ac change) and upstream it? [17:05] seb128, yes, is there a bug #? [17:05] rodrigo_, lennart pointed by email some issues that was one of those [17:05] ok [17:05] rodrigo_, the other one was that your patch for gtk_quit_add() was buggy as he pointed in the bug [17:05] rodrigo_, I've backported the git commit to replace it so that's solved [17:07] oh, didn't see the change in the upstream bug [17:09] kenvandine, could you have a look to bug #724917? [17:09] Launchpad bug 724917 in libappindicator "Importing appindicator from python crashes with ImportError on undefined symbol" [Medium,Triaged] https://launchpad.net/bugs/724917 [17:10] kenvandine, it might be worth backporting for alpha3 no? [17:10] yeah [17:10] looking [17:11] kenvandine, thanks [17:12] kenvandine, I have just updated the bug with a debdiff [17:12] artfwo, great, thx [17:12] rodrigo_, can you try to see if you can sort bug #718805 as well? [17:12] Launchpad bug 718805 in libcanberra "[Natty] system-ready and login sounds do not play" [Medium,Triaged] https://launchpad.net/bugs/718805 [17:12] seb128, sure [17:14] seb128, oh, just moving canberra-gtk-play to a better subpackage? [17:14] rodrigo_, not sure where is was but yeah put back the gtk2 version as it was before [17:18] kenvandine, bug #721562 as well btw [17:18] Launchpad bug 721562 in libappindicator "libappindicator3-dev does not include headers" [Undecided,New] https://launchpad.net/bugs/721562 [17:19] humm [17:19] we don't install gtk3 specific headers [17:23] kenvandine, should you depends on the gtk-dev then? [17:23] kenvandine, i.e libappindicator-dev [17:23] i thought it did [17:23] i'll confirm [17:23] Depends: libdbusmenu-glib-dev (>= 0.1.8), libdbus-glib-1-dev (>= 0.76), libappindicator3-1 (= 0.2.96-0ubuntu1) [17:23] here [17:25] ok, i'll fix it [17:27] kenvandine, thanks [17:28] and fix the .pc file [17:28] ugh === tkamppeter_ is now known as tkamppeter [17:31] is it a known unity bug that I can't type anything in the 'Search' text entry from the thing that is opened when clicking on the ubuntu icon on top of the unity launcher? [17:33] nessita, the think is called the dash [17:33] nessita, not known that I know about [17:34] seb128: thanks for pointing me to the proper term. Would you know who I can ask about it, in particular? to see if I file a bug [17:36] nessita, njpatel [17:36] seb128: thanks! [17:36] chem cho? [17:36] or didrocks when he's not under alpha update rush which is the case now [17:37] njpatel: hi there! I was wondering about me not being able to type anything in the Search box in the dash [17:37] njpatel: shall I file a bug? [17:37] nessita, should be fixed as part of today's release [17:37] njpatel: ok, thanks! [17:37] nessita, if it's the same tomorrow, then yes, please file a bug :) [17:37] njpatel: sure [17:38] njpatel, is that your reply to any bug? ;-) [17:39] hi all! I just upgraded to natty and now "magnet:"-type links don't open anymore, they are associated with "transmission" in firefox, but open clicking on one, there is http://paste.ubuntu.com/573552/ in .xsession-errors. Where would I start looking at the cause? [17:39] if someone has time to do two reviews: https://code.edge.launchpad.net/~mathieu-tl/evolution/2011-02-25/+merge/51352 and https://code.edge.launchpad.net/~mathieu-tl/evolution-exchange/lp694884/+merge/51219 [17:39] the patch pilot of the day ;-) [17:39] hahah [17:40] Ampelbein, the new glib, gvfs use mimetypes for url handlers [17:40] seb128, yep :) [17:40] mterry: ^ if you have some time, could you review the above please? :) [17:40] Ampelbein, you need an x-scheme-handler/magnet in the .desktop I guess [17:40] does anyone know how udev works on an ARM system, where there is no dmidecode? [17:40] cyphermox, k, will look [17:40] mterry, thanks [17:41] what tool provides values for things like DMI_VENDOR, etc.? [17:41] seb128: ok, thanks. would that be considered a bug or is it the user's responsibility to add those mime-types? [17:42] achiang, dmidecode? [17:42] Ampelbein, it's a bug, it's the package .desktop which should have it [17:42] artfwo: on ARM, i mean. [17:42] pitti, ^ the udev question might be for you [17:44] achiang, afaik udev does directly, no? e.g. ENV{DMI_VENDOR} == "blah blah" [17:45] achiang, I'm basing my analysis on what's in /lib/udev/rules.d/95-keymap.rules [17:45] cyphermox: i guess i must just be confused as to what supplies the values for those environment variables [17:46] duh, i read bad [17:46] achiang: udev doesn't use dmidecode [17:47] achiang, ENV{DMI_VENDOR}="$attr{[dmi/id]sys_vendor}" [17:47] in general, it's not needed at all any more, as the values are in sysfs [17:47] /sys/class/dmi/id/ [17:47] pitti: ah, thank you [17:47] sorry pitti ;) [17:48] of course that doesn't necessarily exists on arm [17:48] pitti: right, i think i just discovered that [17:48] cyphermox: sorry what for? you are completely right :) [17:49] rickspencer3: do you get weird gstreamer playback errors in pithos after listening for a while? [17:49] pitti: maybe i should ask my real question -- i'm trying to write a multimedia keyboard quirk for an arm machine, but getting stumped as to how the udev rule would fire, since we don't have any DMI information anywhere [17:49] it basically gives up after like 20 minutes for me [17:50] achiang: you don't really need to use DMI information, that's just generally most appropriate on x86 [17:50] achiang: you can match on anything really [17:50] pitti: ok, just gotta find something unique, i guess [17:51] achiang: in general the rules shoudl be specific to the keyboard, not to the machien [17:51] achiang: on many laptops we need to fix the internal keyboard, so the DMI info is the best thing we have [17:52] achiang: but look at the top of /lib/udev/rules.d/95-keymap.rules [17:52] pitti: this is an internal keyboard too; and i guess i've found some identifying information about it in dmesg [17:52] pitti: so i could match on that [17:53] jcastro, I haven't used it lately, sorry [17:53] achiang, we usually match against /proc/cpuinfo on arm [17:53] achiang: that does a few fixes for USB based external keyboards, where we read the vendor/product of the keyboard itself [17:53] beyond that you usually have lsusb [17:53] achiang: actually I wasn't entirely correct above [17:53] achiang: for laptop internal keyboards, it is the BIOS that defines the "meaning" of the hotkeys and their scan codes (not the keyboard itself as in "the piece of hardware" [17:54] is the info you need board or actually kbd specific [17:54] ogra: keyboard specific [17:54] ah, thats bad [17:55] ogra: this is an embedded system, so i guess i could just match on the board [17:55] pitti: nod, understood that it is firmware that defines the values of the keys/scancodes [17:55] yeah, thats what i was thinking [17:55] ogra: what i'm working on won't be suitable for upstream anyway... [17:56] just have a rule that fires off a script ? [17:56] which in turn sets what you need [17:56] yeah [17:56] probably even without involving udev but just using an upstart job [17:56] given that its likely that you have a custom image :) [17:57] interesting, i could do it via an upstart job too. i suppose that makes sense [17:57] ok, thanks for all the suggestions [18:04] achiang: (udev rule vs. upstart job should be fairly irrelevant to how you identify the machine, FWIW) [18:05] pitti: got it, thanks. also, do you know where the various XF86* hotkeys are defined? === chaotic_ is now known as chaotic [18:09] achiang: for X? /usr/include/X11/keysymdef.h [18:10] pitti: ah, thanks [18:10] achiang, some are defined in xkeyboard-config [18:10] dunno if that's relevant to your problem though [18:11] well, one bug i'm trying to solve is the ability to launch a chat client with a multimedia hotkey [18:12] but i'm not seeing anything similar to XF86WWW, e.g., that could do the job [18:14] but you have the key existing ? [18:14] why dont you just use xev to catch it by keycode [18:15] ogra: yes, i do have the keycode. but g-c-c's keyboard shortcuts app doesn't show anything interesting for launching a chat client [18:15] oh, but i guess i could just add a custom command [18:16] you could even define an xmodmap :) [18:16] heh, yeah [18:16] key { [ XF86Messenger ] }; // KEY_CHAT [18:17] bryce_: ah, thanks [18:19] whee, webkit finally built; that took as long as OpenOffice back then, argh [18:19] * pitti feels Sweetshark's pain [18:20] pitti, wait until we give him a beagleboard to fix all the arm issues ;) [18:20] *then* you can feel pity [18:28] jasoncwarner: so on natty, we trigger a build for libreoffice (european morning) and if the amd64 is finished before the i386 build, you get broken packages because some of the packages used on amd64 are actually build by i386. [18:28] rodrigo_, ping === MacSlow is now known as MacSlow|dinner [18:28] * Sweetshark tries to find the buildlogs, but launchpad times out on him :/ [18:29] Sweetshark: I was getting this error starting, I think, yesterday morning my time, until just now...not sure if that helps with timing... [18:29] so for a good 24 hours I was probably getting the error [18:30] jasoncwarner: now it works? [18:30] Sweetshark: only b/c I think I removed openoffice manually and then did the dist-upgrade after [18:31] Sweetshark: I think anyone that has stock 10.10 system might still get the error when upgrading currently (just a guess) [18:32] jasoncwarner: ah, you tried an upgrade? [18:32] yes, that's known broken [18:32] LibO doesn't build the transitional openoffice.org-* pacakges any more [18:32] and we don't have a replacement for those yet [18:32] pitti: Might that be because we do not have new transitionals .... [18:32] ... ;) [18:33] rodrigo_: do you want me to sponsor bug 718805 ? [18:33] Launchpad bug 718805 in libcanberra "[Natty] system-ready and login sounds do not play" [Medium,In progress] https://launchpad.net/bugs/718805 [18:33] jasoncwarner: will be fixed soon. [18:34] Sweetshark: ack..thanks [18:36] jasoncwarner: I am currently on fixing the arm build, after that I will take care of the transitionals and the l10n stuff. [18:37] Sweetshark: thanks...:) [18:40] good night everyone [18:41] 'night pitti! [19:10] hi again. I now added "MimeType=application/x-bittorrent;x-scheme-handler/magnet;" to the desktop.in file of transmission-gtk, rebuilt it and installed that package. what works: transmission opens - but the dialog for adding the torrent does not. If I manually run "transmission-gtk "magnet:?xt=urn:btih:SOSPTVEPXJT3I7ZKVDRUNQAM2C6CN3YI&dn=FILENAME&tr=http://tracker.openbittorrent.com/announce", the add dialog comes. Any ideas where to l [19:10] ook? === zyga is now known as zyga-afk [19:11] could it be a problem with the special characters in the magnet url? how do I find out what command is actually called when I try opening magnet links? === MacSlow|dinner is now known as MacSlow === bdrung_ is now known as bdrung [19:45] is there any chance to have a sponsorship for https://code.launchpad.net/~nataliabidart/ubuntu/natty/ubuntuone-control-panel/ubuntuone-control-panel-0.9.0/+merge/51318 ? === ogra is now known as Guest25882 === Guest25882 is now known as ogra_ [20:17] nessita, hello! I see you request for an ubuntuone-control-panel update. I'll look at it, but we're after FF, so depending on what it does, it may need an FFe? [20:19] mterry: they add no new features but bug fixes [20:19] gah [20:19] what is this debian-changes-foo getting output in debian/patches and added to debian/patches/series nonsense? [20:20] dobey, source format 3.0 [20:20] dobey, it takes any changes to the source and rolls it into a patch [20:20] mterry: actually it appears that it is a patch which reverts the patch which i am trying to add :( [20:21] and it is threatening my sanity :( [20:21] dobey, :) your patch is quilt? not sure why format 3.0 would fight with itself like that [20:22] mterry: well banshee is already using quilt it seems, and i added my patch in and edited the series file [20:22] and when i try to build a source package it generates this new debian-changes patch and shoves it in series too [20:23] all i want to do is get a sane debdiff so i can attach it to the bug and get it sponsored :) [20:25] mterry: any ideas? do i need to go to the farmers' market and see if i can get a live chickengoat? [20:28] dobey, um [20:28] dobey, is the package using quilt too? (like, are both quilt and source format 3.0 active?) [20:28] maybe they're fighting [20:28] (since source format 3.0 re-implements quilt) [20:29] debian/rules isn't including the patshsys-quilt.mk [20:29] so i don't think so [20:29] and i'm having to work from an apt-get source copy, because the bzr branch seems to be really old [20:30] it's banshee i'm working on though [20:31] dobey, looking at man dpkg-source, maybe provide --unapply-patches ? [20:31] dobey, besides futzing with debian/source/options, the only thing I can suggest is to investigate why the patches aren't being unapplied/cleaned before building a source package [20:32] hmm, ok === ogra_ is now known as ogra [20:40] nessita, uploaded the new control panel [20:40] mterry: thanks! [20:47] dobey: If that is threatening your sanity, you should see libreoffice where there are patches patching patches in a quiltlike, but not quite quilt implementation ... [20:49] Sweetshark: i've seen, and done, a lot worse than that. :) [20:59] anyway.. is there any compiz 0.9.4 being port to maverick ? [21:01] didrocks: [21:01] ejat: if someone wants to do it, they can :) [21:02] just be aware there is a huge dependency stack to backport [21:02] yeah .. [21:02] thats what im thinking .. the dependency [21:02] * ejat breaking my desktop .. after upgrading it to 0.9.2 :( [21:04] should i downgrade back to 0.8 or anything that i can do .. [21:05] hey, is gnome-shell going to be supported in natty+1? [21:06] kklimonda: define supported [21:07] micahg: in main archive, with patches provided for 18/36 months. [21:07] the question isn't even about natty+1 but more general - are we going to support it when it gets to archive just like old good gnome? [21:08] kklimonda: as long as it's build against mozjs, probably not [21:08] mterry, i think i know what's causing bug 723873 now [21:08] Launchpad bug 723873 in indicator-messages "indicator menu items missing since the recent natty update" [Undecided,Fix released] https://launchpad.net/bugs/723873 [21:08] chrisccoulson, ooh! [21:09] chrisccoulson, what, pray tell? [21:10] mterry - thunderbird is resetting some menuitem properties back to defaults when the menu opens, and this just deletes them in the new version. In DbusmenuClient, when a property is deleted from a menuitem, it just calls dbusmenu_menuitem_property_remove [21:10] but that doesn't signal to anything that a property is deleted [21:10] so changing a property from non-default -> default never triggers any updates [21:10] i think, anyway ;) [21:11] i'm about to verify that in gdb [21:13] chrisccoulson, awesome. I hope that's it [21:15] mterry, yeah, it sorta makes sense [21:15] should be fairly trivial for me to fix too [21:16] i already had a look with dbus-monitor, and i can see the properties being deleted where they were previously updated in the old version [21:16] so DbusmenuClient just needs updating to handle that too, and set the defaults in the client [21:22] didrocks: y i cant login to my unity desktop in maverick ? is it becoz of the compiz ? [21:22] ejat: you shouldn't try the new unity in maverick, we have a lot of dep on natty [21:23] owh ok then .. ill take note that .. [21:33] kenvandine: hey, can you sponsor the patch i attached to bug #723960 ? [21:33] Launchpad bug 723960 in banshee "[FFE] u1ms: links are not handled by banshee" [High,In progress] https://launchpad.net/bugs/723960 [21:33] dobey, sure [21:33] thanks! [21:39] dobey, you also enabled u1ms in debian/patches/07_enable-u1ms-by-default.patch in this change too? [21:39] if so, please add it to the changelog [21:42] dobey, also can you give me a same u1ms url for testing? [21:45] s/same/sample [21:46] i just downgrade my compiz to 0.8.x from 0.9.x .. i try to enable back my desktop effect .. its work but the toolbar disappear ... [21:54] dobey, whoops... i guess you just merged those two patches into one? [22:05] I've noticed a bit of an issue with OpenOffice/LibreOffice. It's a bit impossible to shut down or restart if the QuickStarter is open. Is this an Ubuntu issue or probably just in Oo/Lo? [22:07] dobey, doesn't seem to be getting applied, i guess a quilt problem? [22:07] no time to look now, will try later on [22:07] gotta run out now [22:19] JackyAlcine, that's bug 562027 [22:19] Launchpad bug 562027 in openoffice.org "[ooo-build] OOo QuickStarter: unable to shutdown / reboot / logout when quickstarter is active" [High,In progress] https://launchpad.net/bugs/562027 [22:24] unity decided to hide all my windows decorations. Does anyone know how to maximize a window using keyboard shortcuts? [22:24] nessita, perhaps try restarting unity-window-decorator ;) [22:24] it might have crashed [22:25] nessita: alt-drag the window up to the top, it should maximise :) [22:25] and that too ;) [22:25] chrisccoulson: very likely, the process in charge of my window decoratuions has been crashing all day long [22:25] RAOF: let's see... [22:25] yeah, me too! [22:25] i think a recent upload should have fixed those though [22:25] RAOF: nothing happens, just a window move is the result [22:26] RAOF: ah! it did work, only after I stop pushing the ALT key [22:26] (but I had to keep dragging until the screen went blueish) [22:26] Right. [22:26] That's it saying ‘I'll maximise if you drop me here’ [22:27] Similar things happen if you drag to the far left or right. [22:27] RAOF: right, thanks! [22:29] Does anybody else occasionally lose window border/title bars? [22:29] TheMuso: Yeah; the decorator crashes. [22:29] I'm running traditional GNOME with compiz, but this is the second time in as many days I've lost title bars/borders. [22:29] ah ok. [22:30] TheMuso: yeah, the same happes to me both on unity and classic desktop [22:32] the compiz upload from today should fix some of the crash issues [22:32] trying upgrading [22:33] Will do so in a bit, thanks. [22:33] * nessita will too [22:34] chrisccoulson: I've probably found a real cause for bug 723664, so don't work on it for now. I'll comment on it in few hours, or tomorrow, after I get a response from T dev. [22:34] Launchpad bug 723664 in firefox "When using the "Open with..." option in the Download dialog, Firefox launches Transmission but doesn't add the torrent." [Medium,Triaged] https://launchpad.net/bugs/723664 [22:37] kklimonda, ok, thanks [22:37] mterry, i have a working patch for bug 723873 now :-) [22:37] Launchpad bug 723873 in indicator-messages "indicator menu items missing since the recent natty update" [Undecided,Fix released] https://launchpad.net/bugs/723873 === soren_ is now known as soren === chris|| is now known as chris| === Daviey_ is now known as Daviey [23:10] is this lovely or what? :-) [23:10] [24115.509129] glade[7517]: segfault at 30 ip 00007f4da0fda5b8 sp 00007fff6542cd20 error 4 in libgtk-x11-2.0.so.0.2400.1[7f4da0dc4000+436000] [23:29] So... Now after updating, unity still wants to load even when I have classic desktop selected at gdm. [23:31] ...and simple-ccsm is broken. [23:33] ;3~;3~...but ccsm works. [23:33] gah [23:34] /b 46