[00:03] cyphermox, i uploaded the gtk patch from ted for fallback icons [00:03] in unapproved queue [00:04] awesome, thanks [00:05] np [00:44] hello everybody. [00:44] i need some help,please! [00:45] i installed the gnome3 in ubuntu 10.10 like this: [00:45] sudo add-apt-repository ppa:gnome3-team/gnome3 [00:45] sudo apt-get update [00:46] sudo apt-get install gnome-shell [00:46] but the desktop didn't change.but why? [00:48] TopWinStudio: (1) I'm not sure if the gnome3 ppa actually supports 10.10 (rather than the current development release that will become 11.04), (2) You'd need to select the gnome-shell session at GDM. [00:50] thanks.is the 11.04 stable now? === marrusl is now known as marrusl_afk [01:36] hi. what package do i assign a bug to when there's an issue with language packs? [01:39] vish: did someone notice that the file copy icon in the indicator area (the nautilus one) is now a mouse pointer? === marrusl_afk is now known as marrusl [07:14] jcastro: yup, i've noticed it, maybe its a MarkDecision™ ? I dont know and dint ask.. ;) [07:44] Good morning [08:11] good morning [08:11] hey didrocks, good morning [08:11] hey mvo, how are you? [08:12] good, thanks! not fully awake yet, but the tea is redy now so it won't take long [08:17] :) [08:19] mvo, morning [08:20] mvo, just fixed another refactoring leftover in watch_lock [08:20] glatzor: awsome! [08:20] * mvo hugs glatzor [08:20] mvo, how can I avoid that acquire mounts a cdrom? I am working on a test case for ProvideMedium [08:23] glatzor: apt_pkg.config.set("acquire::cdrom::nomount", "1") should work [08:24] glatzor: then you also need to point it to the right directory "acquire::cdrom::mount" "/path/to/dir" should work [08:24] glatzor: let me know if there are problem,s you may have to turn off udev automounting too, but I think it will just work with the two above config options [08:28] mvo, apt-get install silly-base -o 'Dir'=/tmp/tmpq8796Btmp -o 'Acquire::cdrom::/tmp/tmpq8796Btmp/media/cdrom::mount'=False -o 'Acquire::cdrom::mount'=/tmp/tmpq8796Btmp/media/cdrom -o 'APT::CDROM::NoMount'=1 -o 'Dir::State::Status'=/tmp/tmpq8796Btmp/var/lib/dpkg/status -o 'Acquire::cdrom::AutoDetect'=False" [08:29] /tmp/tmpq8796Btmp ist das chroot [08:29] so that works now? [08:29] no [08:29] out of curiosity what is/was the problem with the provide medium [08:29] keeps asking for the medium [08:29] oh, hold on a sec then, let me check [08:30] a moment i can share my branch [08:31] adding the cdrom works nicely: [08:31] + os.system("echo 'Test CDROM' | apt-cdrom add -m -d %s " [08:31] + "-o 'Acquire::cdrom::AutoDetect'=False " [08:31] + "-o 'Dir'=%s" % (mount_point, self.path)) [08:33] the original provide medium error seems to be a python-dbus issue [08:34] what does Debug::Acquire::cdrom "1" print? if you add that to the test? [08:35] 0% [Wird verarbeitet]CDROMMethod::Fetch cdrom:[Test CDROM]/repo/./silly-base_0.1-0update1_all.deb [08:35] Looking for CDROM at /tmp/tmpomDDIVtmp/media/cdrom/ [08:35] ID 2 30789ebfc1ce94dc618fd4b6547ca0a2-2 [08:35] ID 1 5ec68c75c4024592762a172ee9b1fdcc-1 [08:35] Medienwechsel: Bitte legen Sie das Medium mit dem Namen [08:35] »Test CDROM« [08:35] in Laufwerk »/tmp/tmpomDDIVtmp/media/cdrom/« ein und drücken Sie die Eingabetaste (Enter). [08:36] does the id match? the one that got printed by add? [08:38] ach in der tat [08:39] Verwendeter CD-ROM-Einbindungspunkt: /tmp/tmpOtJb3ytmp/media/cdrom/ [08:39] Identifizieren ... [2f96a8a50411c379e304302e4878845b-2] [08:39] Durchsuchen des Mediums nach Index-Dateien ... [08:39] 1 Paketindizes, 0 Quellindizes, 0 Übersetzungsindizes und 1 Signaturen gefunden [08:39] Bitte geben Sie einen Namen für dieses Medium an, wie zum Beispiel »Debian 5.0.3 Disk 1«: [08:39] Name: 'Test CDROM' [08:39] Dieses Medium heißt: [08:39] »Test CDROM« [08:39] Reading Package Indexes... Fertig [08:39] Schreiben der neuen Quellliste [08:39] Quelllisteneinträge für dieses Medium sind: [08:39] deb cdrom:[Test CDROM]/repo/ / [08:39] Wiederholen Sie dieses Prozedere für die restlichen Disks Ihres Satzes. [08:39] the ids are different [08:39] it will use the .disks/ dir on a writable medium to identify the disk, but it needs to stay the same, i.e. no creating/recreating in a tmpdir [08:40] well, the "disk" ;) [08:40] so the more "frozen" the whole cdrom dir stays, the happy this will be [08:40] mvo, what should the content of .disks be? [08:40] mvo, I could also use a symlink [08:41] the mount point dir doesn't get modified between the apd-cdrom and apt-get call [08:41] glatzor: usually it contains build info and stuff, let me check how much its actually using it, for now it may be enough to just populate iwth with a info file [08:41] glatzor: hmmm [08:41] glatzor: ok, if it does not get modified, there goes my theory :/ [08:42] still worth tryint to add .disk/info if that does not work I would appreciate if you could push the branch and I dig a bit deeper into it [08:43] * glatzor hugs mvo! [08:43] info file is enough! [08:43] excellent :) [08:43] * mvo hugs glatzor [08:44] * mvo HUGS glatzor again for all the nice fixes that are currently flooding in [08:44] * glatzor is just fixing his own bugs :) [08:45] … and makes millions of users happy with this! [08:51] morning [08:53] hey glatzor, rodrigo_! [08:57] hello desktopers [08:57] hey didrocks rodrigo_ [08:58] how are you? [08:58] hi seb128, didrocks [08:59] salut seb128 [09:03] bonjour seb128, ca va didrocks [09:04] hey rodrigo_ [09:04] ça va bien pitti ;) you? [09:04] glatzor: servus! [09:04] didrocks: feeling old! [09:05] pitti: oh right! Happy birthday :-) [09:05] merci! [09:06] hey pitti, ca va bien! et toi ? [09:06] pitti: is that what you asked for your birthday? a working and shiny beta2? :-) [09:06] pitti, oh? happy birthday! [09:06] didrocks: yeah, it's a nice gift, isn't it? :-) [09:06] hey pitti, happy birthday! [09:06] seb128: merci! [09:07] rodrigo_: gracias [09:07] :) [09:07] * pitti sobs "31" [09:07] I was 24 when I started all this. working on Ubuntu makes you grow old :) [09:08] ;-) [09:08] same here [09:08] well I'm not 31 yet but that's coming :p [09:23] bah [09:23] compiz just locked [09:23] didrocks, those locking issues are keybinding related right? I never had a lock before using super to open a place which I just did [09:23] which doesn't work anyway [09:24] seb128: you mean global freeze? [09:24] no, I mean compiz freeze [09:24] not sure what difference you make [09:24] yeah, UI freeze anyway [09:24] I switched to a vt and stopped compiz and restarted it and I'm fine [09:24] so it's only compiz [09:24] it's not a box freeze [09:25] so yeah, it's a freeze where compiz wait on a key receive signal by xlib [09:25] ok [09:25] right, but the whole ui is frozen then as compiz doesn't refresh [09:25] I'm happy I don't use the keyboard ;-) [09:25] seb128: clap your hands, it's the future :-) [09:25] btw in case I didn't tell it before be opening the places with the keyboard sucks [09:26] I sit on super key and did press letters like 10 times [09:26] seb128: how? [09:26] yeah [09:26] this is a compiz issue [09:26] then I unpressed super and it showed the place... [09:26] sam proposed a fix which broke the rest of key handling [09:26] (in compiz) [09:26] ok [09:26] so the fix has been "backlog" [09:27] so I guess "known bug but not on the natty list" [09:27] yes [09:27] thanks [09:27] I will stop complaining there ;-) [09:27] yw, sorry for that, but I can't do anything from the unity side :) [09:27] * didrocks tests a natty installed in // on his netbook. First ubiquity slideshow and "waow" [09:28] (my netbook was stucked, couldn't type in gdm, nothing, last week-end) [09:31] pitti, bah, I though for a minute that we unfroze and that the queue got flushed when reading the other channel :p [09:32] hehe, me too; my heart stopped for a second [09:54] didrocks, bug 711561 seems only half fixed. if i enable the desktop cube plugin, i'm getting the following: http://paste.ubuntu.com/593917/ - after disabling unity & co, the desktop will be a mess [09:54] Launchpad bug 711561 in unity (Ubuntu) "Compiz won't allow Desktop Cube plugin to load with unity" [Wishlist,Fix released] https://launchpad.net/bugs/711561 [09:55] however, after enabling the cube, i can re-enable unity and the cube/rotation plugins work fine [09:55] [i tested this after a unity --reset, just to be sure] [09:56] htorque: yeah, hence my latest comment. This seems to be a bug in ccsm though or libcompizconfig, patch welcomed [09:56] htorque: it's something we don't support at all, I just relaxed the dependency. [09:57] htorque: basically, cube and wall provides a feature, called "largedesktop" IIRC [09:57] unity depends on it [09:57] then, it seems that when you switch to enable one plugin with that feature (which conflicts with other plugins in that features) ccsm doesn't really know to handle this (it should try to remove the conflict first) [09:57] * Sweetshark reads backlog ... [09:58] pitti: Happy Birthday! [09:58] * Sweetshark mumbles: "31? young grasshopper!" [09:58] thanks Sweetshark ! [09:59] htorque: hence the fact it removes unity. Anyway, I've enough to fix with what support already :) but a volonteer contribution on that (shouldn't be that complex to make the check of deps once the whole "transaction finished") is more than welcomed :) [09:59] didrocks, but this is definitely no unity bug? [09:59] ohhh, happy birthday pitti :) [09:59] mvo: :) danke [09:59] pitti, happy birthday! [10:00] htorque: no, this is a ccsm or libcompizconfig one [10:01] pitti: could you approve the FFe for bug 753584? [10:01] Launchpad bug 753584 in humanity-icon-theme (Ubuntu) "[FFe] New Libre office icons are used in the launcher but not in Nautilus / Dash search results" [Low,Triaged] https://launchpad.net/bugs/753584 [10:02] mvo, the test is in rev 637. the provide_medium method works :) [10:02] mvo, but the package manager seems to fail and I cannot identify the reason. [10:02] didrocks: good, thanks! as for "patches welcome" i guess i'm the wrong guy unless it's a one-liner :P [10:03] htorque: heh :) === davidcalle_ is now known as davidcalle [10:07] ping rodrigo_ [10:12] hey lucazade [10:12] lucazade, thanks for coming! [10:12] :) glad to help [10:13] lucazade, ok, so can you please 'bzr get lp:~rodrigo-moya/ubuntu/natty/gnome-settings-daemon/really-fix-649809' first? [10:14] lucazade, and edit debian/patches/series and comment out all the lines (adding a '#' at the beginning of each line) except for the last one (93_wait_for_xsettings...) [10:15] [10:15] rodrigo_ ok [10:15] lucazade, then rebuild, with 'bzr bd -- -b' [10:15] lucazade, then install the packages it will create under ../build-area and try that [10:15] perfect I'll try (didn't know this 'bzr bd') [10:16] lucazade, you'll need to install bzr-builddeb [10:16] lucazade, with this, we'll know if it's one of the ubuntu patches [10:17] lucazade, oh, before loging in, can you go to a vt and: [10:17] rodrigo_, one of the user said that the issue is also on debian when you asked to uninstall ubuntuone-client-gnome [10:17] rodrigo_, but I guess we can still verify [10:17] glatzor: thanks, I will have a look in a bit [10:18] $ gdb /usr/lib/gnome-settings-daemon/gnome-settings-daemon $pid_of_gdm'sgsd [10:18] rodrigo_ ok [10:18] seb128, yes, but I'm out of ideas, so trying everything [10:18] lucazade, on the (gdb) prompt just type 'c' and then log in and inmediately go back to the vt and enter 'thread apply all bt' in the (gdb) prompt [10:19] rodrigo_, can't we simply make g-s-d takes over any running xsettings manager when starting rather than bailing out? [10:19] seb128, it's not possible afaik, not until the X property is gone from the gdm's gsd [10:19] ok === zyga_ is now known as zyga [10:20] seb128, but the problem lucazade and another guy are still having seem to really be g-s-d is taking too long to finish [10:20] which might be a gdm bug if so [10:20] since it should wait for it to finish [10:20] rodrigo, when I try to build I get this: "bzr: ERROR: Unable to find the needed upstream tarball: gnome-settings-daemon_2.32.1.orig.tar.gz. [10:20] pitti, sorry about the "opinion", small bug in my greasemonkey it does that sometimes but's meant to be "invalid" [10:20] " [10:21] seb128: how can it be invalid? [10:21] pitti, it's still not a gdm bug, it's a gtk one and there is a fix in the natty queue [10:21] seb128: that icon clearly is broken.. [10:21] seb128: ah, can we make it a duplicate then? [10:21] pitti, because it's a duplicate but I can't be bothered to search for numbers for every single duplicate filed or I would do nothing else from my day [10:21] lucazade: try bzr bd -S [10:21] seb128: ah, ok [10:21] lucazade, wget http://ftp.gnome.org/pub/GNOME/sources/gnome-settings-daemon/2.32/gnome-settings-daemon-2.32.1.tar.gz and rename it to gnome-settings-daemon_2.32.1.orig.tar.gz on the parent folder of the branch checkout [10:21] seb128: so I misunderstood that then, sorry [10:21] pitti, no worry, sorry about the suboptimal handling [10:22] not sure why, but debian/watch fails also sometimes for me to get the upstream tarball [10:22] I guess it hits an out-of-date mirror [10:22] pitti, http://launchpadlibrarian.net/69317309/gtk%2B2.0_2.24.4-0ubuntu1_2.24.4-0ubuntu2.diff.gz [10:22] pitti, you have the bug reference in there ;-) [10:22] ah :) [10:22] (easier when the uploadis in the queue than searching though launchpad) [10:23] * pitti dupes it then [10:23] dpm: Just a heads up - I just added a string freeze exception rq on https://bugs.launchpad.net/ubuntu-translations/+bug/754424 [10:23] Ubuntu bug 754424 in Unity Foundations "Dash: only returns first 5–6 "available to download" results; misleading because many more are in the Software Centre" [Medium,In progress] [10:24] pitti, thanks [10:24] lucazade, oh, before typing 'c' in (gdb) prompt, please type this: 'set logging on' [10:24] Sweetshark: bug updated (it's a bug fix, not an FFE) [10:25] lucazade, that would save the output to gdb.txt [10:25] lucazade, so that you can pastebin that [10:25] rodrigo_ ok [10:27] glatzor: it looks like its trying to run the debconf preconfigure stuff in the wrong directory (i.e. not in the chroot) [10:27] glatzor: let me check further [10:31] pitti: Hi Martin! [10:31] pitti: Could you please sponsor https://code.launchpad.net/~gunnarhj/gdm/profile/+merge/57218 ? kirkland was supposed to do it, but he seems to have got busy with other stuff, and I'm worrying about "Main Release Freeze". (It's about changes to three lines of code, so it shouldn't take long.) [10:31] hey GunnarHj [10:32] test -f "$HOME/.profile" && ( . "$HOME/.profile" ) && . "$HOME/.profile" [10:32] GunnarHj: hmm -- would you mind changing that to just || true? [10:32] first it only takes half the time, second there might be stuff in .profile which doesn't get along well with being run twice [10:33] mvo, the debconf message are fine. it just cannot open the passwords.dat - but this is just a warning [10:34] pitti: You are the third person who suggests that. ;-) I succeeded in convincing kirkland that it's not working. Pls see comments on the MP. [10:35] glatzor: ok, I have in the chroot log "required operation requires superuser priviledges" (from dpkg). so something with the wrapper, I'm digging [10:35] mvo, ahhhh I know [10:35] mvo, thanks for pointing me to this [10:35] GunnarHj: right, just ( . profile) won't work, that's a new shell [10:35] GunnarHj: I thought about [10:35] set +e [10:36] . profile [10:36] set -e [10:36] kamstrup, thanks for the heads up. Do you think you could give me an updated .pot file I could manually upload into LP, so that translators don't have to wait for the upload and can start translating straight away? [10:36] glatzor: ok, cool - what is it :) I'm curious as well! [10:36] GunnarHj: sourcing .profile twice sounds error prone, I'm afraid I'm not willing to merge this that late [10:36] GunnarHj: wrapping in +e/-e sounds safer to me [10:36] dpm: possibly - i'm just checking with the release team whether they +/-1 it [10:37] kamstrup, ok, when/if you get the confirmation, just ping me with the template if possible and I'll upload it and send a notice to translators [10:37] thanks! [10:38] in the aptdaemon.test.Chroot.setup() method the dpkg-wrapper is configured. but aptdaemon, which is launched in the background, cannot access the apt_pkg.config object in the actual test file [10:38] dpm: sure thing [10:38] lucazade, everything going all right? let me know if you have any problem [10:39] rodrigo_ still building! [10:39] lucazade, ok :) [10:42] pitti: If you hesitate about patch 43_..., it's the change to patch 36_... I'm anxious that it gets uploaded. [10:42] pitti: But set +e sounds like it should work, so I can change to that. [10:42] pitti: If I do, should I really "set -e" afterwards, considering that -e is not set in Xsession by default? [10:42] GunnarHj: oh, it's not? why does it fail then? [10:43] oh, due to syntax errors, not failed programs [10:43] kamstrup, you might want to ask about your ffe on #ubuntu-release [10:43] GunnarHj: set +e won't help with syntax errors indeed [10:43] pitti: no need UIFe either for Bug 753584 ? [10:43] Launchpad bug 753584 in humanity-icon-theme (Ubuntu) "New Libre office icons are used in the launcher but not in Nautilus / Dash search results" [Low,Triaged] https://launchpad.net/bugs/753584 [10:43] pitti: i think Sweetshark meant to tag it UIFe instead of FFe [10:43] GunnarHj: but I don't think we should penalize everyone by running stuff twice just to try and guard against broken .profiles IMHO [10:44] vish: still, it's a bug, isn't it? [10:44] at least I consider it as that [10:44] rodrigo_ ... failed at the end : "Now signing changes and any dsc files... [10:44] Could not find a signing program (pgp or gpg)! [10:44] debuild: fatal error at line 1256: [10:44] running debsign failed [10:44] pitti: not a bug,IMO, its just something new and shiny they want to have ;) [10:44] bzr: ERROR: The build failed. [10:44] " [10:44] lucazade, yeah, don't worry about that, the .deb's are under ../build-area [10:45] correct! [10:45] pitti: Ok. Would it be possible for you to upload only the 'safe' revision, or should I remove the .profile part from the MP? [10:45] pitti: icons are already there, but they wanted to use the new LibO icons [10:45] GunnarHj: as you prefer; if I merge it and drop the xsession stuff from it, it will appear as merged although it isn't really, but I guess it doesn't matter? [10:46] pitti: Aha, think I change it then. Easily done. :) [10:46] pitti: Getting back in a few minutes. [10:50] rodrigo_ installed... i've tried without gdb for the moment and the issue is no more present.. it looks like it is something in ubuntu patches.. now i'm going to use gdb [10:50] lucazade, wait, can you please try a few times to re-confirm? [10:50] ok [10:51] lucazade, then, if you have time, enable patches one by one [10:51] lucazade, well, I'll let you know some patches that can be enabled safely, to save you some builds [10:51] perfect [10:51] lucazade, well, I don't want to steal lots of time from you, so do you have time to do the rebuilds (5 or 6)? [10:52] maybe not now, in the afternoon I can do some tries [10:52] no need to do a new build each time [10:52] lucazade, ok, let me find the safe patches [10:52] you can go in the build dir, apply a patch and run "debuild binary" [10:52] it will not clean and build again but just do "make" and "make install" again basically [10:52] seb128, ah, cool, and that creates the new .deb's? [10:52] yes [10:52] seb128, ah, great [10:53] lucazade, ^^ [10:53] it just start from "make" rather than doing it all over [10:53] dpkg-buildpackage -b -us -uc -nc [10:53] -nc -> "no clean" [10:53] pitti, "debuild binary" is easier to type ;-) [10:53] apply a patch in series file ? [10:53] this is so exceptionally useful that I have a "dreb" alias for it :) [10:53] seb128: ^ not for me :) [10:53] lol [10:53] ok, you win ;-) [10:53] and how does he apply a patch on the build dir? [10:54] lucazade, you can usually just patch -p1 < debian/patches/num_patch [10:54] ah, ok [10:54] seb128 ... ok good to know I'll try [10:55] rodrigo_ anyway Im doing some reboot and logout ... system is stable [10:56] lucazade, ok, so you can safely enable 01__xrdb.patch, 02_fix_randr.patch, 02_missing_libs.patch, 03_maintainer_mode.patch, 05_disable_corner_tapping.patch, 05_use_xrecord.patch, 07_dont_display_autostart.patch, 12_monitor_network_fs.patch, 30_pkgconfig-path.patch, 91_update_gvc_source.patch and 92_datetime_fixups.patch [10:57] lucazade, ok, let me know what you find out, I'll have a look at the other patches myself in the meantime to see what might be wrong there [10:57] I guess the guy complaining about the bug being in debian is because debian doesn't have the upstream fix [10:57] * rodrigo_ checks [10:58] rodrigo_ ok ... I'll let you know asap [10:58] lucazade, no hurry, and thanks a lot for your help :) [10:58] np [11:01] lucazade, oh, and btw, no need for the gdb part of the instructions for now, just enable the patches, and when you see the problem again, let me know which patch and we'll do some debugging if I can't find the issue myself, ok? [11:02] lucazade, that would make things quicker for you [11:02] rodrigo_ yes, of course.. [11:02] ok, thanks again :) [11:03] rodrigo_ first issue! "Applying patch 92_datetime_fixups.patch [11:03] patching file plugins/datetime/gsd-datetime-mechanism.c [11:03] Hunk #1 succeeded at 628 (offset -241 lines). [11:03] Hunk #2 succeeded at 639 (offset -241 lines). [11:03] Hunk #3 FAILED at 891. [11:03] 1 out of 3 hunks FAILED -- rejects in file plugins/datetime/gsd-datetime-mechanism.c [11:03] Patch 92_datetime_fixups.patch does not apply (enforce with -f) [11:03] make: *** [debian/stamp-patched] Errore 1 [11:03] dpkg-buildpackage: error: debian/rules build gave error exit status 2 [11:03] debuild: fatal error at line 1335: [11:03] dpkg-buildpackage -rfakeroot -D -us -uc -b failed [11:03] bzr: ERROR: The build failed. [11:03] " [11:03] lucazade, yes, start from the 1st patch, since the patches build on each other [11:03] well, some patches like this one do [11:04] i've enabled all the patches you told me were ok [11:04] ah, right [11:04] ok, disable that one then, it needs some previous ones [11:04] ok [11:05] brb [11:09] kamstrup: responded to the bug with some alternative suggestions how to avoid the string break [11:18] pitti: Ok, dropped the .profile stuff. https://code.launchpad.net/~gunnarhj/gdm/profile/+merge/57648 [11:20] GunnarHj: merged and uploaded, thanks! [11:29] pitti: Thanks! Think I'll still look into the +e idea. I'll let you know. [11:29] mvo, mei was für a schwere geburt. aber jetzt sollte der test laufen :) [11:29] GunnarHj: it will help with the most common problem of having failing program calls, but if you say that Xsession already isn't set -e, it won't help [11:30] mvo, I wonder why I get the unbalanced action groups warnings [11:30] GunnarHj: you wanted to guard against syntax errors, right? [11:34] pitti: Right. But my assumption has been that there are occurrences of configurations that somehow makes Xsession run with -e being set, even if the Xsession code does not include it. Otherwise I don't understand bug 678421. [11:34] Launchpad bug 678421 in gdm (Ubuntu) "Error in ~/.profile halts the X startup" [Low,In progress] https://launchpad.net/bugs/678421 [11:35] GunnarHj: in this case it would help [11:36] pitti: That's what I thought. But what about my assumption? Does it make sense to you? [11:38] rodrigo_ applied patch 6-appindicator and the issue is now present (going to check again) [11:38] lucazade, yeah, that seemed the culprit to me, so looking at the code [11:39] lucazade, can you just enable the rest and leave that one disabled to check it's indeed the culprit? [11:39] yes yes was going to do [11:39] lucazade, ok [11:39] GunnarHj: off the top of my head I wouldn't know how, unless you also modified the Xsession script [11:41] pitti: Maybe, In that case, I should just ask a question to the bug reporter and set the status "Incomplete"? [11:41] GunnarHj: that bug talked about syntax errors, not failed program calls [11:42] mvo, a simple with self._cache.actiongroup(): [11:42] pkg.mark_install() results in an unbalanced action group error [11:42] pitti: Yes. Not sure what you mean now. I'm also talking about syntax errors. [11:43] pitti, hi, any idea what bug 759635 could be about? [11:43] Launchpad bug 759635 in chromium-browser (Ubuntu) "apport ERROR: hook /usr/share/apport/package-hooks//chromium-browser.py crashed" [Medium,New] https://launchpad.net/bugs/759635 [11:43] GunnarHj: right, and set -e doesn't help you there :) [11:43] or set +e [11:44] pitti: As far as I understand, as long as -e is not set, syntax errors does not make the process stop. [11:44] pitti: Doesn't +e disable -e, so to say? [11:45] GunnarHj: right [11:45] GunnarHj: syntax errors always kill the program, as you can't even parse the shell script then [11:45] GunnarHj: try e. g. [11:45] echo begin [11:45] echo ${HOME/m/x/} [11:46] echo end [11:46] i na file [11:46] in a file [11:46] and then run it with bash (works) and dash (fails, as ${x/from/to} is a bashism) [11:47] pitti: Will do. Getting back in a while. [11:47] GunnarHj: it does fail, it was just an example to point out the difference between a failing program call and a syntax error [11:50] pitti: Does it mean that my examples in the comments on https://code.launchpad.net/~gunnarhj/gdm/profile/+merge/57218 actually tested for failed program calls, not syntax errors? [11:51] GunnarHj: calling "wrongsyntax" is not a syntax error, it's just a failed program call [11:51] which set +e would ignore [11:52] GunnarHj: so there a set +e/set -e wrapper would do fine (but shouldn't be necessary really as Xsession isn't set -e) [11:52] GunnarHj: the original bug mentioned a bashism in ~/.profile [11:52] rodrigo_ enabled all patches but #6 ... no issues [11:52] pitti: I see. Then I give up. For now, that is. :) Thanks for the lesson! [11:53] lucazade, ok, looking at the code, so as soon as I have some fix, I'll post a link to a package on the bug [11:53] lucazade, so not stealing more time from you anymore :) [11:53] rodrigo_ ok .. I'll look at the bug report, ask me if you need testing [11:54] lucazade, yes, I'll need some testing as soon as I've got a fix, as the issue does not happen to me [11:54] lucazade, but yes, just keep track of the bug, so thanks again for your time [11:54] rodrigo_ ok see you! [11:54] lucazade, bye! [11:56] pitti: thanks! I added a reply [11:56] kamstrup: me too :) [12:01] vish: ping? [12:02] hmm, I remember jcastro posting some URL for unity tricks, does anyone know where is it? also, is that where we should redirect people? [12:02] didrocks, ^ [12:02] rodrigo_: which kind of tricks? his video or shortcuts? [12:03] rodrigo_: there were something else IIRC on askubuntu with a "unity tag" [12:03] didrocks, shortcuts and other things, iirc [12:03] oh, will try askubuntu [12:04] http://askubuntu.com/questions/28086/unity-keyboard-mouse-shortcuts [12:04] ok thanks [12:04] I guess the best thing is to send people to http://askubuntu.com/questions/tagged/unity [12:04] right? [12:04] seems to have a lot of stuff [12:05] yeah [12:05] that's what he pointed [12:13] seb128: the accessibility icon beeing the generic "unknown icon" in gdm is a known issue I guess? [12:18] Sweetshark: pong [12:20] didrocks, yes [12:20] seb128: ok thanks [12:20] didrocks, there is a fixed package in the queue waiting for the freeze to relax [12:20] yw [12:20] ok, excellent then :) [12:22] vish: I will ask about getting bug 753584 sponsored from https://code.launchpad.net/~ubuntu-art-pkg/humanity/release any final words about that one? [12:22] Launchpad bug 753584 in humanity-icon-theme (Ubuntu) "New Libre office icons are used in the launcher but not in Nautilus / Dash search results" [Low,Triaged] https://launchpad.net/bugs/753584 [12:22] Sweetshark: nope, we just need to point the changelog to Natty and release :) [12:24] vish: great! [12:25] rodrigo_, what patch is creating the gsd issue? [12:25] seb128, the appindicator one [12:25] weird [12:25] rodrigo_, are we sure? [12:25] seb128, I think the keyboard one, still checking [12:26] seb128, lucazade enabled all patches but that one and it works, but with that one enabled, he sees the problem again [12:26] vish: about bug 758854 and related issues: IMHO the icons in our humanity theme in LO and outside of it should be the same. So if there are new icons in humanity-icon-theme, the same should also be updated to the LO theme. Right? [12:26] Launchpad bug 758854 in humanity-icon-theme (Ubuntu) "Humanity does not support document type distinction by color" [Undecided,New] https://launchpad.net/bugs/758854 [12:26] rodrigo_, ok, weird, did he get a stacktrace of where it hangs with it? === MacSlow is now known as MacSlow|lunch [12:27] seb128, no, didn't want to steal more time from him, but I will ask him again in the afternoon if I don't see where the problem is [12:27] vish: (not for natty, but for O) [12:28] rodrigo_, ok, thanks for investigating it [12:28] seb128, thanks go to lucazade :) [12:28] Sweetshark: right now, its not an issue they are same.. actually he got confused by the first bug and thought we needed to create new icons for humanity in LibO colors.. not sure if we will be changing those icons too much, but if we do i'll let you know [12:30] Sweetshark: the humanity icons in libO package are mainly rendered png versions of what we have in humanity, just that it is a PITA searching and using the right icons names :D [12:33] vish: indeed. [12:38] pitti, ping [12:39] fta, you should better ping with context, he's taking a break but other people might be able to reply [12:40] seb128, it's above, melted within a thread :( [12:41] the chromium apport bug you listed? [12:41] apport hooks (bug 759635) [12:41] Launchpad bug 759635 in chromium-browser (Ubuntu) "apport ERROR: hook /usr/share/apport/package-hooks//chromium-browser.py crashed" [Medium,New] https://launchpad.net/bugs/759635 [12:41] no idea what's going on there [12:43] glatzor: maybe something for juliank? [12:58] fta - you seem to be misusing the second parameter to add_info in your script [12:58] chrisccoulson, yep, just figured it out [12:58] that's not a string, but is an object that allows you to interact with the user (ask questions, and stuff like that) [12:58] ah :) [12:59] it's a apport.ui.HookUI instance (now?). i wonder why it worked before [13:00] i'm not sure why it worked before. perhaps that changed, but i don't think so :/ [13:04] hum [13:04] "We're Sorry. Firefox had a problem and crashed." [13:04] oh? [13:05] did you get the crash dialog? (i assume so) [13:05] it's still on screen [13:05] can you submit it? [13:05] I'm pondering telling mozilla or not [13:05] i want to make sure it's nothing to do with me ;) [13:05] I don't like much sending browser datas over the internet [13:06] "Submitting your report" has a "broken icon" icon btw [13:06] bah [13:07] I submitted the crash but the stupid thing didn't give me an url about where it landed or anything [13:07] is the broken icon on gdm for a11y already reported somewhere ? [13:07] seb128 - that's ok. if you go to about:crashes, you can get the URL from there [13:07] ogra_, yes [13:07] (while you talk about broken icons) :) [13:07] ah, great, thanks [13:07] chrisccoulson, https://crash-stats.mozilla.com/report/pending/f61a67ff-f55d-4863-b2d8-28e672110414 [13:08] thanks [13:08] well it's spinning and has a retry counter [13:08] I did ctrl-W while in the natty queue with a filter on shotwell if that's of any use [13:09] there was no other tab open [13:09] chrisccoulson, https://crash-stats.mozilla.com/report/index/f61a67ff-f55d-4863-b2d8-28e672110414 [13:09] hmmm [13:09] i really need to start generating breakpad symbols from gtk and glib, and submitting those too [13:10] unfortunately, it's crashing on a thread with no symbols at all :( [13:10] there is no apport crash or anything I can deal with it locally? [13:11] no, it's blacklisted in apport [13:15] mvo, mpt: I made a small path which would allow to overwrite the lintian warning if a package of bad quality should be installed [13:15] patch [13:15] glatzor: ok [13:17] seb128 - oh, i see some gtk criticals on shutdown if i do ctrl+w with a single tab on the natty queue [13:17] could be a clue ;) [13:17] (firefox-bin:8216): Gtk-CRITICAL **: IA__gtk_clipboard_get_for_display: assertion `!display->closed' failed [13:18] ok, lunch time, bbl [13:20] mvo, lp:~glatzor/software-center/force-bad-quality [13:25] thanks glatzor, looking [13:26] ogra_, broken a11y icon: I think it's supposed to be in the queue too, or at least the fix would have been yesterday, if it does fix it ;) [13:26] cyphermox, yeah, i just wanted to know if i should file it [13:27] ogra_, it was filed already [13:27] yup [13:39] pitti, ok, sorry to ask but do you have any clue when uploads might start being reviewed and get in? any likely it will be today? [13:39] seb128: not yet, as we are still testing; I certainly hope/think tonight [13:40] ok [13:42] pitti - can i get an ACK for bug 758895? :) [13:42] Launchpad bug 758895 in mozjs (Ubuntu Natty) "couchdb needs rebuilding" [Medium,Triaged] https://launchpad.net/bugs/758895 [13:48] chrisccoulson: ack for what? it's just a bug fix? [13:48] hm, what have I broken since the unity menu search doesn't find anything [13:49] pitti - oh, i wasn't sure whether i should ask, or just go ahead and upload it :) [13:49] chrisccoulson: go! [13:50] seb128 - i had a look at the greasemonkey menu issue yesterday (the multiple separators) [13:50] it looks a pain to solve with my current menu design :( [13:50] i might have to leave it that way for natty ;) [13:51] and super-key stops working [13:51] meh [13:53] tjaalton: as a first measure, try unity --reset ? [13:54] it seems there's still a lot of churn - for me, tiling broke again === KenEdwards is now known as KenEdwards-afk [13:56] chrisccoulson: would you mind having a quick look at http://people.canonical.com/~platform/workitems/natty/canonical-desktop-team.html#chrisccoulson ? [13:56] chrisccoulson: "Figure out how to provide system-wide settings in /etc in the new world" seems we should postpone this or just drop? (I remember that you said that it's not really supposed to work that way) [13:57] pitti - i can guess we can close that one as DONE. i've already figured it out, i just haven't implemented it yet ;) [13:57] heh, it says "figure out", so fine :) [13:57] the mozilla-devscripts one can be postponed [13:58] and i guess i should probably fix webfav [13:58] was about to ask, is that actually necessary? [13:58] probably not, now we got rid of most of our extensions [13:58] we just ship packaged extensions for e. g. langpacks, which seems fine? [13:58] chrisccoulson: so feel free to switch to DROPED [13:58] DROPPED, too [13:58] does anyone actually use webfav? [13:59] nobody reported a bug about it not working [13:59] we used to in the past on netbook [13:59] -- natty/main armel deps on webfav: [13:59] ubuntu-netbook [13:59] -- natty/universe amd64 deps on webfav: [13:59] tbh, i've not been tracking extensions that much this cycle (they aren't on https://wiki.ubuntu.com/DesktopTeam/Specs/Natty/Firefox4/XULRunner20Transition) [13:59] netbook-launcher-efl [13:59] that's it [14:00] I'm looking at http://people.canonical.com/~platform/workitems/natty/canonical-desktop-team.html#mozillateam now [14:00] and wonder if I can just kill the ones without rdepends [14:00] micahg: ^ [14:01] we can kill libgtk2-mozembed-perl [14:01] pitti: didn't help, plus now the launcher is stacked below every window :) [14:01] gtkmozembed has been removed from mozilla now anyway [14:01] which means, anything using gtkmozembed will be disappearing next cycle [14:02] chrisccoulson, micahg: libjavascript-perl, ubuntu-it-menu, mediatomb, libgtk2-mozembed-perl have no rdepends whatsoever, can I kill? === MacSlow|lunch is now known as MacSlow [14:05] pitti - libjavascript-perl and libgtk2-mozembed-perl can definitely go [14:05] i'm not so sure about mediatomb. i think some people are still using that [14:05] and i'm not sure about ubuntu-it-menu. [14:06] pitti - i'd get rid of it. it's easily available from addons.mozilla.org [14:06] it doesn't really fit the criteria for keeping it in the archive :) [14:07] it's also easily discoverable from the addons-manager [14:07] right [14:07] oh, nice. i did a search for "ubuntu" in the addons manager and got this: http://www.webgapps.org/addons/askubuntu [14:07] jcastro ;) [14:10] tedg, morning! [14:11] chrisccoulson: actually, what's the problem with mediatomb? [14:11] tedg, just FYI for today's releases hopefully: https://code.launchpad.net/~mterry/indicator-datetime/update-after-resume/+merge/57574 and https://code.launchpad.net/~mterry/indicator-applet/global-desc-update-callback/+merge/57533 [14:11] pitti - i don't want to make myself too unpopular ;) [14:12] chrisccoulson: no, I mean, I don't actually see a xulrunner 1.9.2 dep [14:12] pitti - oh, ok. i'm not sure why that's on the list then :) [14:12] mterry, Cool, definitely! [14:13] chrisccoulson: http://paste.ubuntu.com/594001/ [14:13] huh, why is conkeror on there? i thought that was done already, unless someone reverted it [14:14] tedg, look at that second one especially, as I would imagine it would cause problems all the time, instead of the relatively rare crash, so I might have not understood something [14:14] * kenvandine grumbles at how long it takes for dh_girepository to run on dbusmenu [14:14] tuxguitar loks fine as well [14:14] yeah, tuxguitar was fixed with the swt-gtk change [14:14] kenvandine, I have a testing version of the packaging that strips out a lot of the packaging, including the gir packages for dbusmenu :) Runs super fast [14:14] i am sure :) [14:15] it takes much longer on dbusmenu than any of the others i build regularly [14:15] pitti - moonlight-plugin-mozilla is bogus (there's a recommend on xulrunner-1.9.2, but it depends on firefox) [14:15] that should probably be cleaned [14:16] mozplugger is ok. that depends on firefox | xulrunner-2.0 | xulrunner-1.9.2 (plus many others) [14:16] xulrunner-1.9.2 <- libjdic-java <- paros (only user of libjdic-java) [14:17] libgtk-mozembed-ruby1.8 is real, though [14:17] libgtk-mozembed-ruby1.8 has no real rdepends (just the libgtk-mozembed-ruby metapackage) [14:17] so in theory we could just drop this entirely, or at least ignore for natty [14:18] chrisccoulson: want me to do a mozplugger upload with dropping the recommends? [14:18] pitti - from popcon "34884 paros 308 13 248 8 39 (Unknown)" [14:18] i'd probably just drop that ;) [14:19] fine for me [14:19] pitti - yeah, feel free to do a mozplugger upload to drop the recommends [14:19] i can do a moonlight upload to drop the recommend there too [14:19] chrisccoulson: sorry, mozplugger is fine [14:19] no recommends [14:19] just alternative depends [14:19] ah, yes [14:20] chrisccoulson: also a non-default recommends, so no worries [14:20] conkeror is also ok - those are alternative depends too [14:21] the wrong way around, but yeah [14:21] (it should depend one the latest one first) [14:22] pitti - oh, i need to fix that then [14:22] conkeror users should be installing the latest version [14:23] chrisccoulson: so by and large it's eclipse now, right? [14:23] hello all [14:23] chrisccoulson: webfav doesn't depend on xulrunner -- is the problem that it doesn't work with ffox 4.0? [14:23] hey rickspencer3 [14:24] pitti, right [14:24] it probably just needs maxVersion changing, it's quite a simple extension [14:24] hey rickspencer3 [14:24] hey didrocks [14:25] pitti I hear there were some hickups overnight with building a couple of ISOs? [14:25] hey rickspencer3 [14:25] rickspencer3: yes, the DC pretty much broke down due to some hw failures, so building the latest images got delayed quite much; but all fine now [14:25] heya seb128 [14:25] pitti thanks man [14:26] what's the word on the street regarding beta 2? === KenEdwards-afk is now known as KenEdwards [14:26] nothing OMG any more, AFAICS [14:26] :) [14:26] looking pretty good to me [14:27] chrisccoulson: want me to fix conceror, and you look into webfav? [14:27] pitti - sure, can do [14:27] thanks [14:30] chrisccoulson: uploaded, I'll send the fix to Debian [14:30] thanks [14:39] pitti - webfav tested and uploaded [14:40] oh, assuming i can upload that :/ [14:41] oh, i can. that's ok [14:41] "Archive Upload Rights for ubuntu-mozilla-uploaders: archive 'primary', package set 'mozilla' in natty" \o/ [14:43] i'd like to get bug 709125 fixed :) [14:43] Launchpad bug 709125 in AptUrlRedirector "User agent doesn't include Ubuntu in it so apt.ubuntu.com doesn't work" [Critical,Fix committed] https://launchpad.net/bugs/709125 === seb128_ is now known as seb128 [14:45] chrisccoulson: splendi! [14:46] ... d [14:46] heh :) [14:49] bah [14:49] pitti: happy birthday ! # just got notification from skype ;) [14:49] wtf - on the firefox-4 blueprint, someone has requested feedback from me "I am a user, have Linux on 3 computers in daily use, I find some of the changes in Firefox not convenient,in particular the location of the tab bar, it is more difficult to useand not at all logical. Also the date and time app has lost it's temperature setting..." [14:49] edwardc: oh, thanks :) [14:49] natty on the 10v ssd starts is 29:78 [14:49] it was like 15 seconds in lucid [14:49] seb128: it's hilarious, isn't it? [14:49] seb128: we had it down to 10.3 in the best case.. [14:49] asking for help on a blueprint is strange enough, but what on earth is the last sentence about? it seems a bt random [14:50] pitti, right, for some definition of "hilarious" [14:50] seb128 - lucky you. i can go and make a coffee whilst my laptop boots in natty [14:50] chrisccoulson, it's about indicator-datetime not doing what the GNOME clock was doing ;-) [14:50] it seems that startup time on rotary disks has regressed quite significantly :( [14:50] it's slower than it has ever been ;) [14:51] well not only rotary [14:51] like it's twice slower on the 10v ssd [14:51] that's pretty bad [14:51] humm... smoking fast on this thinkpad [14:51] it's start to gdm in 10s and desktop in 20 seconds [14:51] we were down to like 6 seconds [14:52] yeah, desktop start is the real killer for me too [14:53] bah, I don't like the u1 indicator integration [14:53] like when enabling my account on a new install I get 3 system user lines in the indicator [14:53] which don't really make sense [14:55] chrisccoulson, If a debian package has "Depends: iceweasel | firefox | abrowser", do we change that in Ubuntu to anything? [14:55] didrocks: "Remerge the Quickly template with new API" -- can we safely consider this "postponed"? [14:55] mterry: not necessary to have a delta for it, as we don't have an iceweasel package [14:56] mterry, no, there's no need to change anything there [14:56] kenvandine, bug #760680, have fun [14:56] Launchpad bug 760680 in indicator-applet (Ubuntu) "Translations need to be exported and package needs to be rebuilt" [Undecided,New] https://launchpad.net/bugs/760680 [14:56] pitti: yeah, the template is even not in ubuntu and we now have python places support, so it will be even better! [14:56] pitti, chrisccoulson: that's what I thought, cool [14:56] mterry, which package is it? [14:56] chrisccoulson, pytrainer, some sport training app in universe [14:56] didrocks: sorry, does that mean "postpone" or not? [14:56] mterry, oh. that's doomed anyway [14:57] chrisccoulson, ? [14:57] pitti: yeah, postpone, can do that once natty is released [14:57] mterry, it's using gtkmozembed, which is now dead [14:57] (it's been dropped from mozilla entirely) [14:57] seb128, oh joy [14:57] ok [14:57] chrisccoulson, ah, right. :-/ well, I can fix it for natty at least [14:58] (it's currently not launching on startup) [14:58] kenvandine, https://translations.launchpad.net/ubuntu/natty/+source/indicator-applet/+export [14:58] didrocks: mind if I steal bug 595008 from you? [14:58] Launchpad bug 595008 in gnome-pkg-tools (Ubuntu) "should port the GNOME cdbs custom rules to dh7" [Wishlist,Triaged] https://launchpad.net/bugs/595008 [14:58] kenvandine, let me spare you searching for the url ;-) [14:58] :) [14:58] didrocks: (in oneiric I'd like to merge with Debian and add dh_translations to it) [14:58] kenvandine, oh, it was in the bug... anyway ;-) [14:58] didrocks: and also update our cdbs rules to use dh_translations [14:58] yup :) [14:58] pitti: well, you did the work, so it's more than fair :) [14:59] didrocks: all of this pretty much belongs together [14:59] seb128, need to figure out how to build the package with them... [14:59] didrocks: this bug/WI is different from "write dh_translations", but related of course [14:59] mterry, http://hg.mozilla.org/mozilla-central/rev/2038ba37a2e8 ;) [14:59] kenvandine, get ted to import them in trunk and roll a tarball? [14:59] that sounds easy [14:59] :-D [14:59] ;-) [14:59] tedg_, ^^ [14:59] chrisccoulson, yeah, I remember hearing about it. Just didn't notice this guy was using it [15:00] pitti: sure, but you did the hard part, so do not hesitate to pick this one :) [15:00] didrocks: ok [15:00] mterry, there's not much in the archive using it. the other thing is chmsee, but that is already buildable with webkit too [15:00] tedg_, kenvandine: you might need a LINGUAS file in the po dir which list the po available as well [15:00] which is good, because i use chmsee :) [15:00] heh [15:07] seb128: I think your WI on https://launchpad.net/ubuntu/+spec/packageselection-desktop-n-improve-photo-printing is probably done or obsolete or so; do you have a current status for this? [15:08] pitti, it's still somewhat on my list but I didn't found anybody who cared or had a clue to reply [15:08] pitti, so postponed I guess [15:08] seb128: ack, tahnks [15:23] lucazade, around? [15:24] Sweetshark, do you think you'll have the chance to look at bug 725746 after the beta? Do you need more info on the report? [15:24] Launchpad bug 725746 in libreoffice (Ubuntu) "Missing language on LibreOffice -Catalan (Valencian)-" [Medium,Confirmed] https://launchpad.net/bugs/725746 [15:25] rodrigo_ yes [15:26] lucazade, have time for some gdb fun? [15:26] lucazade, I think I know what the fix is (building a package for you to test), but if I could get confirmation from gdb, that would be great [15:26] ok.. you have to guide me [15:26] lucazade, yes [15:27] lucazade, ok, so log out, and on a vt -> gdb /usr/lib/gnome-settings-daemon/gnome-settings-daemon $pid_of_gdm_gsd [15:27] lucazade, then (gdb) set logging on [15:27] (gdb) c [15:28] and then log in and inmediately switch back to the vt and CTRL-C + (gdb) thread apply all bt [15:28] lucazade, then just send me the gdb.txt file you will get on the dir where you run gdb [15:28] s/send me/pastebin [15:29] ok I'll try [15:29] ok thanks [15:29] I've to switch back to vt as soon as panels appear? [15:30] lucazade, as soon as you can, so once you press ENTER on gdm login screen, switch to the vt [15:30] rodrigo_ ok [15:31] lucazade, it took a few seconds for the gdm's gsd to die, right? [15:31] maybe 10sec [15:31] ok [15:32] lucazade, then no need to be superfast, but yes, switch asap [15:34] after thread apply all bt? [15:34] ctrl+c? [15:34] lucazade, no, before [15:34] gdb is still running [15:34] when you switch to the vt, you get gdb running [15:34] so CTRL-C to get to the prompt [15:35] done [15:35] and then ¡thread apply all bt' on the prompt [15:35] 'thread apply all bt', sorry [15:35] yep.. and I still have gdb prompt [15:36] should I exit to get log [15:36] ? [15:36] yes, after some output, right? [15:36] just press 'c' to let it finish [15:36] and then 'q' to quit gdb and yes, get the log [15:38] no output.. infact gdb.log is empty..."the program is not being run." [15:38] gdb.txt [15:38] oh, empty? [15:38] yes, i meant gdb.txt [15:39] no other files there? [15:39] no [15:39] hmm [15:40] lucazade, you did the 'set logging on' right? [15:40] rodrigo_ yes [15:41] and no output at all when you entered 'thread apply all bt' ? [15:41] no [15:41] when I start gdb i get this [15:41] http://i.imgur.com/U8qVn.jpg [15:42] lucazade, ah, right, you need to run it with sudo gdb ... [15:42] :) [15:42] default in ubuntu is to not allow users to use gdb [15:42] lucazade, so, can you repeat, please? [15:43] rodrigo_ yes [15:43] lucazade, ok, so on the vt: sudo gdb /usr/lib/gnome-settings-daemon/gnome-settings-daemon [15:43] (gdb) set logging on [15:43] (gdb) c [15:43] then log in and switch back and: CTRL-C and 'thread apply all bt' on the (gdb) prompt [15:44] and q? [15:44] after 'thread apply...' -> 'q', yes [15:45] lucazade, you're italian? :) [15:45] yes [15:45] where from? [15:45] roma [15:45] this time I saw the output in gdb [15:45] cool, that's ok then [15:45] but has not been written to txt [15:45] hmm [15:46] lucazade, ah, probably it's in /root/ because you ran it as sudo [15:46] "the program is not being run. Countinuing. [thread 0xb74dbb70 (LWP 1880) exited] program exited normally. quit [15:47] this is gdb.txt [15:47] that's all? [15:47] yep [15:47] oh, are you running the gsd you built with the patches disabled? [15:47] yes [15:47] only 6 disabled [15:47] ok, that's it then, you need to re-enable #6 [15:47] or wait for my package to build in the ppa [15:48] this is why I don't gather the log? [15:48] thus we can test if the fix does it [15:48] lucazade, that's why when you get to gdb prompt the program is done, since there is no bug with that patch disabled [15:48] :) [15:48] obviously [15:48] yeah :) [15:49] is your package in building? [15:49] yes, but I can push my branch, if you want to build locally [15:50] i386 is almost done on my ppa, if you use that [15:50] ok.. give me a link [15:50] to the branch? [15:51] to ppa [15:51] right, and it's now built, so https://launchpad.net/~rodrigo-moya/+archive/ppa [15:51] that is -> deb http://ppa.launchpad.net/rodrigo-moya/ppa/ubuntu natty main [15:52] it's got 2.32.1-0ubuntu15 [15:52] ok thanks [15:52] lucazade, please do the gdb thing again if possible, just in case the fix is not good, so that we can get the gdb stuff [15:53] rodrigo_ ok understood [15:56] lucazade, https://launchpad.net/~rodrigo-moya/+archive/ppa/+files/gnome-settings-daemon_2.32.1-0ubuntu15_i386.deb if you want to just download it instead of adding the ppa and apt-get update [15:56] ugh... not fixed [15:57] :( [15:57] and what does gdb says? [15:57] gdb [15:57] trying [15:57] ok [16:01] rodrigo_ [16:01] http://paste.ubuntu.com/594064/ [16:02] lucazade, ok, so it's in the spot I thought (xrandr plugin) but not in the exact one [16:03] lucazade, looking for a better fix, mille grazie again [16:03] no problem [16:05] it's gconf blocking it seems :( [16:15] tedg_, when were you planning on doing a dbusmenu release? [16:18] mterry, I was just getting started on it... do you have a patch planned? [16:18] mterry, No API changes, so I can do things in any order. [16:18] tedg_, I feel kinda close to that xemacs fix... I'll know more in a bit [16:18] * tedg_ usually starts from the bottom of the stack and works up -- just an FYI [16:19] mterry, Cool, I'll do it last then. [16:19] tedg_, it also may be indicator-appmenu, FYI [16:19] both :-) [16:19] tedg_, if I don't get it, we might want to consider blacklisting it? [16:21] mterry, Okay. [16:26] lucazade, only thing I can find is a memory leak, which I'm not sure it would fix the problem, so can you build my branch and run it (no gdb)? [16:27] lucazade, bzr get lp:~rodrigo-moya/ubuntu/natty/gnome-settings-daemon/really-fix-649809 [16:27] rodrigo_ ok going to try [16:27] lucazade, ok thanks [16:28] rodrigo_, why would indicator call hang? [16:28] seb128, it doesn't it seems [16:29] I thought app_indicator_set_status did some more work than expected, but that wasn't the problem [16:29] now the memory leak is on gnome_rr_config, from gnome-desktop [16:29] which has some gconf stuff, so maybe that's the issue, since we are leaking it in the patch [16:30] rodrigo_ don't know if this help: I've tried to remove all indicator from panel but the problem persist [16:30] rodrigo_, oh ok [16:30] lucazade, that won't help, since the indicator stuff is still used, even if you don't have the applet in the panel [16:30] ok [16:30] we should really check at runtime if there are indicators or not, but I guess that's not the problem here [16:31] or everyone would see the problem [16:31] as gdm doesn't have the indicator applet [16:32] and I haven't been able to see the problem, not only once after 100s of tries [16:32] strange, is always reproducible on different machines here [16:33] building.. [16:35] why-oh-why is mimetype handling in firefox such a mess [16:35] rodrigo_ if I use autologin issue is not present [16:35] lucazade, right, no gdm's gsd in that case [16:37] lucazade, do you set anything non standard on your boxes? [16:37] I never got it in any boxes there [16:38] removed ubuntuone and some other things, nothing strange... ubuntu software center also [16:38] some other users mentioned not having ubuntuone either but I fail to see how that could be due to it [16:39] already tried reinstalling ubuntuone [16:40] pitti: good morning, mediatomb, I'm switching to mozjs185, libjdic-java if paros is unused, then drop both [16:41] hey micahg [16:41] micahg: hm, I didn't see anything related to xulrunner in mediatomb [16:42] pitti: we disabled javascript in lucid which was what made it a xul rdepend, but now that we have mozjs, we can reenable it [16:44] rodrigo_ tried your branch.. no luck [16:44] :( [16:51] tedg_, ok, figured it out. it's in dbusmenu, will work on a patch [16:54] mterry, Woot! [16:55] tedg_, mterry, all set for your appdeveloper week sessions later on? [16:56] dpm, yeah [16:56] dpm, I've changed the topic. Can I talk about flower aranging? ;-) [16:56] tedg_, only if you do a videocast :) [16:56] "Zen of developing software through appropriately placed flowers" [16:56] mterry, awesome [17:07] pitti: BTW, I have gnome-web-photo, just having a little trouble w/DEP-5 in the copyright file, will try to finish it tonight [17:07] rodrigo_: have those guys submitted any fixes to the gnome 3 stuff yet? [17:08] jcastro, no, I had a merge proposal from jbicha, but he's not from that team [17:11] rodrigo_: I'll send a follow up [17:11] jcastro, ok [17:35] tedg_, what's your deadline? I've got it half fixed, but now turns out something else is wrong with how emacs works. I have a classroom to run in a half hour [17:38] mterry, I don't specifically have one yet, kenvandine? [17:44] mterry, well just distro patch the fix tomorrow? [17:44] mterry, it would gave you time for testing without blocking anything [17:44] we don't need every fix to be in today tarball [17:44] ok [17:44] i thought today's was a more lenient push than future. not that it isn't still a bug [17:45] mterry, well no hurry as well [17:45] I'm just pointing it that backporting a revision as a distro patch tomorrow is an option [17:46] but getting in today upload is fine as well ;-) [17:59] launchpad says unity 3.8.8 was released 22 minutes ago. Does it fix bug 732412? There is no change to the bug yet although it is assigned to the milestone. [17:59] Launchpad bug 732412 in unity "LibreOffice Writer has no icon in Unity and is labeled "LibreOffice Small Device Format Importer"" [High,Confirmed] https://launchpad.net/bugs/732412 [18:01] no it doesn't [18:01] DBO, ^ [18:04] yikes [18:05] should I work around it by not shipping the "Small Device Format Importer" desktop file? LO still has one release ... [18:07] Sweetshark, it's basically your call, I would not bother [18:07] but for the record I don't see this bug on a stack beta2 install done today [18:07] libreoffice-writer is on the default launcher and starting it matches the icon [18:08] running it from the application place as well [18:09] if the binary which makes it bug is not installed by default it seems somewhat low priority and easier to fix later in an unity upload [18:21] tedg_, no big rush === hallyn is now known as hallyn_afk [18:51] tedg_, I'm less enthusiastic about fixing this short term. Maybe let's plan for a distro-patch or blacklisting [18:52] tedg_, the basic problem is that emacs is adding new submenus that are empty (I fixed a bug here where we didn't watch such menus for new items). But having fixed that, emacs never seems to add more items and my attempts to force it haven't quite worked. So I need to do some deeper digging [18:57] mterry, Hmm, okay. Fixing the empty items is good though. [18:57] mterry, I'm sure someone else does that. [18:57] tedg_, it's actually a bit worse than that. I don't believe we watch any submenus added in after the fact [19:03] mterry, Hmm... yeah, that is worse. [19:06] I got gnome-session 3 to work with the Ubuntu session but I'm having trouble with Ubuntu Classic [19:07] I can get either gnome-panel or compiz to start, but not both [19:11] jbicha, any issues with empathy other than it not working with the me-menu yet [19:12] bcurtiswx: I don't really use empathy since it doesn't do IRC that well & I don't have need for other protocols these days [19:13] jbicha, OK. np. if you get a change to play around tho, please do and lemme knwo of any issues [19:13] chance* === AndrewMC is now known as Guest80546 === Guest80546 is now known as AndrewMC [19:25] pitti: ping? [19:26] Sweetshark: (on the phone) === hallyn_afk is now known as hallyn === tkamppeter_ is now known as tkamppeter [20:48] hi all [20:49] all? [20:53] kenvandine: hey Ken [20:53] hey pitti [20:53] kenvandine: would you be able to sit in tomorrow's release team meeting? [20:53] kenvandine: it will probably be a very quick one [20:53] pitti, sure [20:53] kenvandine: but I'll be in a train tomorrow, with poor or absent connectivity [20:53] kenvandine: thanks; I'll try to update the wiki page in the morning [20:53] great [20:53] (still busy doing b2 stuff) [20:54] pitti, 1500 UTC? [20:55] yes [20:55] ok === tedg_ is now known as tedg [21:07] pitti - i postponed 2 of the remaining WI's on the firefox-4 blueprint now [21:08] (the eclipse one and the translation one) [21:08] we're not goint to get eclipse done, and we'll fix translations next cycle anyway [21:08] **going [21:08] chrisccoulson: ok, thanks [21:09] * micahg would like to disable the internal browser in eclipse which would solve the problem [21:11] * pitti yays the software-center upload, tackling 3 RC bugs [21:11] tremolux: bug 723911 also has an aptdaemon task, that's also still RC? [21:11] Launchpad bug 723911 in aptdaemon (Ubuntu Natty) "deauthorize does not remove sources.list" [Medium,Triaged] https://launchpad.net/bugs/723911 [21:12] kenvandine: I updated https://wiki.ubuntu.com/DesktopTeam/ReleaseStatus [21:12] kenvandine: might get some more bug fixes by tomorrow, of course [21:12] pitti, thx [21:12] kenvandine: but I guess it'll only be a quick meeting tomorrow anyway [21:12] * pitti waves good night, have to get up at 6 [21:12] good night pitti [21:13] chrisccoulson, kenvandine: in case someone asks, I'll be mostly offline tomorrow, I by and large spend the day in a train [21:13] good night pitti [21:18] good night pitti [21:33] * didrocks waves good night as well [21:33] good night didrocks [21:33] good evening kenvandine! [21:37] hey ronoc [22:16] gnome3 ppa question: noticed that LANG and GDM_LANG are no longer getting set upon login from GDM. temporarily placed "export LANG="en_US.UTF-8"" in /etc/profile, but seems like that shouldn't be the right place to set it? (no more lang chooser in gdm greeter looks like?) [23:28] pitti was beta 2 like a birthday present you gave yourself?