[06:44] good morning [07:54] could someone lend me their un modified directory of /usr/share/icons/hicolor/22x22/apps [08:55] good monring seb128 and njpatel [08:55] morning mvo [08:55] mvo! [08:56] hey Hobbsee [09:10] mvo: so, why do a lot of people not have linux-$arch packages installed when using the dist upgrader? [09:11] Hobbsee: do you have a example bugnumber for me please? there is code that deals specifically with the kernel, but it might be not dealing with every situation [09:12] Hobbsee: (sorry, I was a couple of days away and haven't looked at the latest reports yet) [09:12] mvo: no, just a relatively high # of people in #ubuntu+1 not having their kernel modules, and complaining that their wifi doesn't work now [09:12] mvo: no problem. How was your holiday? [09:12] seb128: WDYT about disabling the gphoto gvfs backend? it's pretty buggy ATM [09:13] seb128: f-spot-import is still not called by default, but we should fix that instead IMHO [09:13] hey mvo, welcome back! [09:13] hey pitti! [09:13] * Hobbsee hugs pitti - morning! [09:13] pitti: we got no bug about it and we didn't disable gphoto gvfs automounting upstream, you need to click on the camera icon to mount it [09:13] * mvo hugs pitti [09:13] s/didn't/did [09:13] hey mvo!!! [09:13] Hobbsee: very good, was good to refresh a bit [09:14] seb128!!! [09:14] * mvo hugs seb128 [09:14] but its also good to be back again :) [09:15] Hobbsee: ok, thanks for the info! I will watch out for this in the bugreports and hang out in #ubuntu+1 a bit more :) [09:17] seb128: hey, what about bug 185387, there's a new patch upstream for GIO/GVFS [09:17] Launchpad bug 185387 in nautilus "can not drag a file to a nautilus folder which uses list mode" [Unknown,Confirmed] https://launchpad.net/bugs/185387 [09:17] tjaalton: excellent, thanks (@ key stuck bug) [09:17] Hobbsee: works? nice :) [09:17] hello [09:17] tjaalton: haven't upgraded yet [09:18] but i'm getting it with less freq than before, too [09:18] tjaalton: out of curiosity, what happend to the hw overlay patches for i965? did they made it into hardy or are they defered ? [09:18] Hobbsee: ah ok, it should be fixed now [09:19] mvo: they have been in hardy for a couple of weeks now [09:19] * tjaalton checks [09:19] tjaalton: cool [09:20] yep, since Fri 14 Mar [09:22] tjaalton: no opinion about this one, that's not a priority and the patch is not trivial [09:22] seb128: bummer :) [09:23] tjaalton: and I don't use the list mode so it doesn't annoy me when using nautilus ;-) [09:23] icon view is confusing :) [09:23] seb128: right, but (1) it's not automounted, (2) clicking in places menu brings an error message, (3) opening in nautilus can't open the files, and (4) it prevents you from using f-spot/gthumb since gvfs locks cam access [09:26] pitti: (1) it's not supposed to be automounted otherwise there is a lock on the camera and you can't use a software to do photo import (2) seems a gnome-panel bug (3) that's orthogonal issue to f-spot not being ran when the camera is plugged [09:26] seb128: right, my (2) seems easy to fix, but (3) makes it pretty useless... [09:27] maybe we should bring that up in the desktop team meeting [09:28] but IMHO browsing files is not the first thing *I* want to do on my camera [09:28] (assuming we can fix it to work in the first place) [09:32] pitti: right, I don't really care about it right now, we have a zillion of bugs but almost none about this backend, so I was in "just ignore it" mode [09:32] but I've no strong objection against dropping it for hardy [09:32] I'm primarily interested in your opinion and the plan how it should be like in hardy final [09:33] I think upstream thinks it's cool because you can browse your camera as any disk and dnd photos to copy those [09:33] well, the plan is what we have now [09:33] right, technically that is cool indeed :) [09:33] the backend is available and not used [09:33] people who want to use it can use it from the gnome-panel or the computer location [09:33] and f-spot should be opened when plugging a camera [09:33] so that will stay, ok; then we shuold at least get it fixed to work [09:34] seb128: ok, I agree [09:34] thanks [09:34] I don't intend to spend any work fixing this one, no [09:34] if it's too buggy drop it [09:34] does it work for you? I just get errors when I click on a picture [09:34] we have too many bugs right now and can afford spending efforts fixing things we don't use [09:34] (that was my initial guess why I asked whether we should drop it) [09:35] morning everyone [09:36] hey huats [09:37] pitti: no that doesn't work because eog doesn't use gvfs so it doesn't understand gphoto locations, but I can browse and dnd photos to my desktop [09:37] ah, so that was it [09:38] pitti: I'm fine disabling it if it comes confusing === asac_ is now known as asac [09:45] pitti: that seems to be a gvfs or nautilus bug, selecting the actions to open a folder or ask what to do doesn't work either [09:49] pitti: btw do you plan to look at this g-v-m mount errors soon? Would be nice to make sure they come from g-v-m and that changing it will not create side effects on gvfs [09:49] seb128: which mount errors? [09:50] I'm just aware of bug 202931 [09:50] Launchpad bug 202931 in gnome-volume-manager "ejecting a CD/DVD with drive button gives always an error message but still works" [Medium,In progress] https://launchpad.net/bugs/202931 [09:50] which I'll fix soon [09:50] pitti: bug #202931 [09:50] exactly :) [09:50] seb128: I'll try to do it today [09:50] pitti: I get a similar error when putting a CD in the drive [09:51] oh, weird [09:51] I don't [09:51] and g-v-m does not automount any more [09:51] but that's only on my desktop [09:51] could be an user setting [09:51] maybe similar to bug 95368 ? [09:51] Launchpad bug 95368 in hal ""Cannot remove directory" on unmount due to stale .hal-mtab entries" [High,In progress] https://launchpad.net/bugs/95368 [09:51] i. e. stale /media/.hal-mtab? [09:51] or conflicting options in fstab, broken cdrom symlinks, etc.? [09:52] /dev/hdc /media/cdrom0 udf,iso9660 user,noauto 0 0 [09:52] hum [09:52] noauto there [09:52] that's ok [09:52] auto == mount at boot time [09:53] WM_CLASS(STRING) = "gnome-mount", "Gnome-mount" [09:53] the dialog is a gnome-mount one [09:53] "gnome-mount -b -d /dev/hdc" is the corresponding command and it has been started by nautilus [09:54] but something has been quicker at mounting it apparently [09:54] what does it say? [09:54] Couldn't mount the volume [09:54] but it's already mounted so that's normal [09:55] in any case, worth taking a look at /usr/lib/gnome-volume-manager/gnome-volume-manager -n [09:55] if it tries to do something [09:56] manager.c/2940: Device added: /org/freedesktop/Hal/devices/volume_label_G71_VA31015 [09:56] manager.c/2723: Changed: /dev/hdc [09:56] manager.c/2157: mounting /org/freedesktop/Hal/devices/volume_label_G71_VA31015... [09:56] manager.c/821: executing command: /usr/bin/gnome-mount --hal-udi=/org/freedesktop/Hal/devices/volume_label_G71_VA31015 [09:56] manager.c/3099: Mounted: /org/freedesktop/Hal/devices/volume_label_G71_VA31015 [09:56] [09:56] yeah, it's mounting it [09:56] a-haa [09:57] * pitti just notes that autoipod is still enabled; is that supposed to stay in gvm, or move into nautilus, too? [09:58] seb128: that should be covered by /desktop/gnome/volume_manager/automount_media [09:58] but the gconf keys are entirely ignored in the code, hmm [09:59] $ gconftool-2 -g /desktop/gnome/volume_manager/automount_media [09:59] true [09:59] right [09:59] that's what I though, the redhat patch is supposed to disable those actions [10:00] seb128: ok, I guess I'll get the same if I enable the gconf key; is there a bug about it? [10:00] It should be easy to fix [10:00] I guess filter_out_media_handling() needs to be applied at more places or so [10:01] pitti: no bug about that apparently, I just noticed on my desktop [10:01] should I open a bug? [10:02] please; assign it to me and set it to 'in progress' [10:02] ok [10:02] still busy with fighting postgresql, but I'll poke at g-v-m in the afternoon [10:02] * pitti hugs seb128 [10:02] * seb128 hugs pitti [10:04] pitti: bug #209586 [10:04] Launchpad bug 209586 in gnome-volume-manager "still automounts cdroms in hardy but should not" [Low,In progress] https://launchpad.net/bugs/209586 [10:06] merci [10:16] hey MacSlow [10:16] seb128, Salut [10:16] MacSlow: could you look to bug #207693? [10:16] Launchpad bug 207693 in libwnck "gnome-panel crashed with SIGSEGV" [High,Triaged] https://launchpad.net/bugs/207693 [10:17] MacSlow: gnome-panel crashing due to your tooltip patch, it has already 30 duplicates [10:21] seb128, the steps to force the crash don't lead to a crash here... I tried it before the weekend too [10:21] MacSlow: just mouseover out and in the tooltip a few times [10:22] MacSlow: anyway seeing the number of duplicates there is a bug in the code, could you have a look to the issue? [10:23] seb128, I will... but just moused-over (in and out) 30 times and nothing happened :) [10:23] I'll try on the laptop again [10:24] MacSlow: could be a race and not easy to trigger on your box, maybe try to run it in valgrind, it's likely to display errors [10:36] seb128: btw the gvfs update corrected the bug I was facing... [10:36] thanks [10:36] which one was yours? [10:36] timeout errors? [10:37] yep [10:37] I was facing it, every time... [10:37] and none since this morning :) [10:40] huats: good ;-) [10:45] shw [10:45] sorry, EFOCUS [11:12] pitti: here is the latest change for gdm [11:12] pitti: http://sinecera.de/ubuntu-gdm-themes_0.28.dsc, http://sinecera.de/ubuntu-gdm-themes_0.28.tar.gz, http://sinecera.de/ubuntu-gdm-themes_0.28_source.build, http://sinecera.de/ubuntu-gdm-themes_0.28_source.changes [11:15] could someone with french skilz translate http://paste.ubuntu.com/6255 for me please? [11:15] (only the error message) [11:15] mvo: can create symlink, no such file or directory [11:15] aha, thanks! [11:15] can't [11:15] you are welcome [11:16] cracking on the upgrade issues list? ;-) [11:16] yes, its long [11:16] and seems to be mostly non-duplicates - this is both good and bad [11:16] good = no obivous bugs / bad = many different ones [11:20] hey [11:21] seb128, would you be interested with tango icons for nm-applet ? [11:21] I have uuencoded the files [11:21] crevette: ask asac, I don't work on network-manager [11:21] ah sorry [11:21] asac, would you be interested with tango icons for nm-applet ? [11:21] :) [11:22] upstream author doesn't seems interested with [11:27] kwwii: thanks, I'll upload it [11:33] do we have python-support experts here? bug #208961 looks like a race in python-support [11:33] Launchpad bug 208961 in mesa "Update manager fails - From Gutsy to Hardy Beta" [High,New] https://launchpad.net/bugs/208961 [11:33] kwwii: hm, it only got marginally smaller; 3048270 -> 3036892 [11:34] pitti: the point was not to make it smaller :-) [11:34] * Removing extra unused pics from the HumanList theme to reduce the file [11:34] size. [11:34] kwwii: 0.26 grew by quite a bit, and I thought you said back then that this was an error? [11:35] mvo: open a bug about the issue in the bts? [11:35] pitti: one of the other themes had extra pics in it, I thought it would help to remove them [11:35] apparently it did not [11:35] * pitti dputs for now [11:36] cool, thanks [11:38] anyone know how to set the Xephyr size (gdmthemetester uses it and it comes up too big on my screen) [11:55] crevette: how do they look like? maybe the icons should be made themable? [11:56] asac, original :http://bmm80.free.fr/GNOME/shots/nm-fedora.png [11:56] tango style [11:56] or New GNOME either: http://bmm80.free.fr/GNOME/shots/nm-gnome.png [11:57] the gnome stuff looks much nicer [11:57] and it fits with the throbber in firefox [11:57] or nautilus [11:58] yep [11:58] asac: from the looks, we should be using the new gnome stuff [11:58] I could send a debdiff if I find how to uudecode in the package :) [11:59] kwwii, and the feodra idonc are blurry [11:59] oups [11:59] kwwii: what does "New GNOME" mean? will that be integrated in upstream network-manager? [11:59] crevette: right [11:59] asac: no idea [11:59] asac, New GNOME; means for me, GNOME updated with tango style [11:59] that happend few release ago [12:00] asac: new gnome is the new icons for gnome which are drawn in tango style [12:00] so they will be official soon? [12:00] crevette: the question is to know why those are not used upstream yet? [12:00] but I doubt that nm icons will go in the gnome icon theme [12:00] or still an alternative theme? [12:00] the current fad is to put the icons with apps (again) [12:00] seb128, Novell guys never pushed to get them upstream [12:01] crevette: maybe you can suggest them on the network-manager list? [12:01] i am subscribed and will see what the devs have to say [12:01] seb128, I've opena bug and discussed on NM ML, but the developpers were not okay to replace their fedora icons (the auther is a RH guy) with GNOME icons [12:02] asac, already done [12:02] crevette: aren't those themable? [12:02] asac, how ? [12:02] crevette: really? have an archive link? [12:02] yep [12:02] letme find it [12:02] asac, http://bugzilla.gnome.org/show_bug.cgi?id=504822 [12:02] Gnome bug 504822 in nm-applet "Use GNOME updated style for icons" [Enhancement,Unconfirmed] [12:02] no idea ;) ... i don't know nothing about themes. seb128, how is that done technically? [12:02] i know about stock icons, but what to do if you want more? [12:03] http://mail.gnome.org/archives/networkmanager-list/2008-January/msg00018.html [12:03] asac, I don't now either about theme [12:03] hate hate hate hate hate epiphany-browser in hardy [12:03] seb128: what else except the cert thing? [12:03] for which we are ---><---- that close to a fix [12:04] asac: crash on closing and it often there is still a process running but nothing displayed (which could be the same issue) [12:04] -it [12:04] asac, http://bugzilla.gnome.org/show_bug.cgi?id=504822#c8 you 've some concerns [12:04] which means opening doesn't work, it's stucked [12:05] asac: http://paste.ubuntu.com/6256/ [12:05] seb128: do you have a steps instruction to reproduce reliably? [12:06] asac: no, it happen just every hour [12:06] so I get it a lot, every day, but I've no special step, I just notice it while trying to open something and that doesn't work [12:06] but might be because I close and open a lot [12:07] so could be similar to the crash on shutdown [12:07] asac: the stacktrace seems to indicate that's a lock on shutdown [12:07] I guess most people keep an instance open somewhere and don't notice [12:07] seb128: do you have nspr dbgsym installed? [12:08] no, but I can install those [12:08] please do and install the nss ones as well [12:08] ok [12:08] I've closed it now to copy the stacktrace on paste.ubuntu.com but I'll give you a new backtrace next time I get the bug [12:12] asac, I'll try to provide you a patch tonight [12:28] thanks [12:30] good morning [12:58] hello pochu [13:01] hey seb128 [13:03] seb128: is the archive backend working, or does it need to be enabled somewhere? (if I try to open a .zip it's opened in file-roller) [13:04] pochu: it's working but not user visible yet, will upload the nautilus change this afternoon [13:12] seb128: can you look into the human icon theme package and tell me if I did the .links file correctly? [13:12] seb128: http://sinecera.de/human-icon-theme_0.26.dsc, http://sinecera.de/human-icon-theme_0.26.tar.gz, http://sinecera.de/human-icon-theme_0.26_source.build, http://sinecera.de/human-icon-theme_0.26_source.changes [13:12] kwwii: will do [13:12] kwwii: did you try to build it? [13:12] seb128: no, how should I do that? [13:13] you don't try your package updates? [13:13] I though you had some packaging training ;-) [13:13] seb128: usually, there is no reason as it is only a png or such [13:14] sudo apt-get build-dep human-icon-theme [13:14] dpkg-buildpackage [13:14] seb128: hehe, I know enough to get myself in trouble and anger you :p [13:19] w00t - it builds and works great :-) [13:19] http://sinecera.de/human-icon-theme_0.26_all.deb [13:22] asac: are you uploading a liferea linked to xulrunner's sqlite? I thought you were, but from your comments in the bug I'm not sure now [13:33] MacSlow: btw, bug #199380 has a valgrind log of the issue, could be due to libcompizconfig [13:33] Launchpad bug 199380 in libcompizconfig "gnome-panel crashed with SIGSEGV in g_object_get() (dup-of: 207693)" [High,New] https://launchpad.net/bugs/199380 [13:33] Launchpad bug 207693 in libwnck "gnome-panel crashed with SIGSEGV" [High,Triaged] https://launchpad.net/bugs/207693 [13:36] pochu: i haven't prepared the fix. I just prepared the xulrunner part [13:36] pochu: we can build as now ... and start with proper LD_LIBRARY_PATH [13:47] mvo: could you sponsor bug #199402 since you sponsored the recent uploads? ;-) The guy who attached the change is upstream so that should be correct [13:47] Launchpad bug 199402 in metacity "gnome-appearance-properties crashed with SIGSEGV in meta_theme_get_frame_style()" [Medium,Confirmed] https://launchpad.net/bugs/199402 [13:53] seb128 that debdiff only have the patch [13:54] thomas commited that on upstream [13:54] can i update gt with detach and profile fixed from svn? [13:55] ember: next tarballs are due next week, there is no hurry to use svn changes [13:55] cool [14:02] asac: http://paste.ubuntu.com/6257/ [14:03] asac: new stacktrace, same issue again [14:26] seb128, pitti: can one of you upload the human-icon-theme package? it works fine everywhere I test it [14:26] I'll have a look soon [14:27] kwwii: which? did I miss a ping? [14:27] excellent, thanks [14:27] pitti: nope, until now I only asked seb128 to take a look at it [14:27] http://sinecera.de/human-icon-theme_0.26.dsc, http://sinecera.de/human-icon-theme_0.26.tar.gz, http://sinecera.de/human-icon-theme_0.26_source.build, http://sinecera.de/human-icon-theme_0.26_source.changes [14:27] pitti: you are welcome to sponsor if you want, I'm just finish some gvfs changes and I'll look at it next [14:28] but if you want to do it I'll not complain, I've a busy todolist ;-) [14:44] asac: don't we need to build it linking to xulrunner's sqlite.h? [14:47] pochu: i don't think so. if we get to know that we do, we should fix that as well [14:49] someone asked me about some tango icons being CC licensed. is that the case? can they be distributed with a GPL application at all? [14:53] asac: according to http://tango.freedesktop.org/Frequently_Asked_Questions#Terms_of_Use.3F they are cc-by-sa 2.5 which is not DFSG free (3.0 is though) [14:53] not sure if it's ok for Ubuntu though [14:53] yes. thats why i wonder [14:54] who encourages artists to use this kind of license? [14:55] pochu: does cc-by-sa have a "or any later version" clause by default? [14:55] seb128: hm, bug 102495 -- AFAIR nautilus opens f-spot now, right? so this would actually be WONTFIX? does nautilus have similar dialogs now? [14:55] Launchpad bug 102495 in gnome-volume-manager "'Always perform this action': Ignore for photo card option doesn't work." [Low,In progress] https://launchpad.net/bugs/102495 [14:55] seb128: I guess it hasn't, but I want to verify before I close the bug [14:56] kwwii: do we distribute cc-by-sa 2.5 artwork (tango)? [14:56] kwwii: read a few lines above [14:57] pitti: nautilus by default just run the action so the bug is deprecated I would say [14:57] right; so it's replaced by the application chooser in the prefs [14:58] yes [14:58] seb128: at least, let's say it's *supposed* to run f-spot :) [14:58] right ;-) [14:58] I tried both 'f-spot' entries in prefs, neither works ATM [14:58] seb128: btw, CD eject error message bug fix committed [14:59] ah, excellent [15:02] asac: don't think so: [15:02] asac: don't think so: [15:02] A new version of this license is available. You should use it for new works, and you may want to relicense existing works under it. No works are automatically put under the new license, however. [15:02] http://creativecommons.org/licenses/by-sa/2.5/ [15:03] pitti: is by-sa 2.5 ok for ubuntu (aka tango) ^^ ? [15:04] yes [15:05] pitti: ok ... an upstream asked me how he can use tango icons in his GPL program (he is copyright holder) ... i said he should add an exception for artwork, did I tell the truth? [15:05] asac: hm, I recommend to ask elmo for that; CC-BY-SA on its own is DFSG-free, but I don't know about the swamp of GPL compatibility [15:06] ok [15:06] pitti: pochu said that by-sa 2.5 is not dfsg free [15:06] only 3.0 [15:07] we have always considered it free enough for Ubuntu at least [15:07] k [15:07] Debian didn't, right [15:07] maybe we should start: UFSG :) [15:08] ATM, UFSG = DFSG + GFDL + CC-BY-SA [15:08] yay, I formulated a sentence with *only* acronyms :) [15:26] mvo: we need to revert the human-theme changes (the murrine and clearlooks engine stuff), think you could find time to help me with that? [15:29] kwwii: yeah, that should be possible [15:33] mvo: cool, thanks...I am heading out to the store now, I'll ping you later or tomorrow, depending [15:38] Can someone confirm that Visual Effects == "none" is Ubuntu-speak for "metacity is enabled", and "normal" and "extra" are Compiz? [15:40] kwwii: ok, I will not be around too late today, but tomorrow works for me as well [15:40] marnanel: confirmed (my fingeres wanted to type configure instead ;) [15:41] mvo: :) [15:41] mvo: thanks [15:53] though one day we hope compiz will come with a non-compositing backend ;) [15:54] Keybuk_++ === Keybuk_ is now known as Keybuk [16:06] i can't run compiz with the nvidia drivers i downloaded [16:06] is there any know issue on that? [16:08] nxvl: the only issue i know of is jockey not adding the right options to xorg.conf for nvidia-glx [16:08] if you use nvidia-glx-new or install the non-legacy driver yourself it should work fine [16:12] Amaranth: mmm, maybe is my fault [16:12] * nxvl tries [16:19] Amaranth: mm now it works [16:19] Amaranth: but i don't see the window manager borders [16:20] sudo nvidia-xconfig --arg-argb-glx-visuals -d 24 [16:20] err [16:20] sudo nvidia-xconfig --add-argb-glx-visuals -d 24 [16:20] arg describes my mood :P [16:25] seb128: hmm; my bug fix for bug 209586 incidentally also fixed autophoto [16:26] seb128: IOW, when I plug in my camera, f-spot opens now [16:26] seb128: I wonder whether I should disable that explicitly now, or leave it enabled until we fix nautilus? [16:27] Amaranth: https://bugs.edge.launchpad.net/ubuntu/+source/compiz/+bug/145360 [16:27] Amaranth: apport has just make me jump into that bug [16:28] yet another generic "it died in eventLoop" crash that gets duped to one of three reports [16:28] Amaranth: apport fault :P [16:29] i'm fine with it, really [16:29] since we have no hope of fixing them based on what apport reports anyway and most of them are not reproducable either [16:29] Amaranth: are you using nvidia own driver or ubuntu nvidia one? [16:29] ubuntu [16:29] that must be the why [16:30] bug #209586 [16:30] only thing on my system that isn't from the repos is savage (cool game) [16:30] pitti: ah, excellent [16:30] pitti: fix nautilus to do what? [16:30] seb128: it was a "break the world" gconf bug :) [16:31] seb128: open f-spot for cameras [16:31] as it indicates in the media prefs [16:31] well, you said that your change fixed it? [16:31] seb128: it fixed it in g-v-m [16:31] Amaranth: what do you have under "Driver" on xorg.conf [16:31] but AFAICS we should disable it in gvm and leave it to nautilus? [16:31] ahh, ok [16:31] nxvl: nvidia... [16:31] mm [16:31] same here [16:31] I though -g-v-m was taking over and breaking the nautilus action [16:32] hum [16:33] pitti: disable it I would say, that will give us the motivation to fix nautilus quickly [16:33] seb128: heh, ok [16:35] Amaranth! [16:36] Amaranth: would you mind to have a look at bug 201330 ? andrea_c7a in #ubuntu-bugs pointed me to it [16:37] the 'fix' in there is wrong [16:37] it whitelists every card someone says works fine [16:37] even though i know a couple of those have issues [16:38] to be safe i want to just blacklist all the ones fglrx works with [16:39] so if you have a laptop and fglrx supports your card then you have to use fglrx because that means you have an r300 or newer and the mobile versions of those have problems [16:50] pochu: liferea uploaded. fix will be effective once the next xulrunner rolls as well [16:51] asac: thanks [16:51] asac: did you see the new stracktrace? is that anything useful for a bug report? [16:51] seb128: attach that to the same bug? [16:51] ill look in code later tonight [16:51] without that i cannot tell [16:51] (and will try to reproduce harder) [16:52] asac: the crash on shutdown bug? [16:52] yes i think thats the one you are currently seeing, right? [16:52] well, the stacktrace from today are a "some process is hanging in background and blocking new instances" [16:53] asac: which is different of the "crash on shutdown" [16:53] but could have the same cause [16:53] yeah... sounds related [16:53] attach to the same for now [16:53] ok [16:53] (until we know more) [16:53] asac: bug 209640, it has the "No symbol table info available." too, but it's with -0ubuntu2 [16:53] hmmm ubotu is dead :) [16:54] seb128: have you considered shipping shares-admin in a different package? so that foobuntu can decide whether to ship it or not [16:54] asac: https://bugs.edge.launchpad.net/ubuntu/+source/liferea/+bug/209640 [16:54] pochu: could you reproduce? [16:54] maybe he didn't restart liferea :) [16:55] i usually ask them to restart X, because reporters often don't get that they didn't restart after upgrade [16:55] not sure if liferea users are better :) [16:55] and liferea is in the tray ... so likelyhood if it still running is high :) [16:56] heh [16:56] otoh how could it run if it crashes ;) [16:57] but if you're still unsure, he's jwendell on #ubuntu-{bugs,devel} [16:57] pitti: does restricted-manager run anything that might lock the apt cache (this requires root)? bug #208112 looks like there might be a interessting race condition [16:57] asac: added to https://bugs.edge.launchpad.net/ubuntu/+source/xulrunner-1.9/+bug/191052 [16:59] ok ... i soft-milestoned until we know more [16:59] ok [17:00] it's annoying but epiphany is not the default browser so not really an hardy target [17:00] hardy-updates or 8.04.1 maybe ;-) [17:01] i hate timer driven crashes :( [17:02] seb128: well ... i need milestones (which i can move forward as required) ... to schedule my bug work efficiently so close before the release [17:02] you mean you don't enjoy a crash that happens well after the thing that caused the crash thus making it impossible to debug? [17:02] Amaranth: right [17:02] stay away from compiz then :P [17:02] i backported about 100 security fixes for that in the past three years :) [17:03] really annoying [17:03] asac: right, me too, I asked a 8.04.1 milestone today, waiting for it ;-) [17:05] seb128: you remember if that happens in session that pop up a http password dialogß [17:06] ß [17:06] damn keyboard ;9 [17:06] arghhh [17:06] asac: not sure if they pop up the password, but I might log in bugzilla.gnome.org during those sessions, the login, password are on the page itself though [17:07] ok ... good pointer. thanks [17:07] you are welcome [17:08] i think we had some leakage in the password management callbacks at some point. not sure if thats gone in latest patch. but if thats the case it might be that it holds an indirect ref on the timer and so on [17:08] but ill see [17:09] that one I'm not sure how I get it, it's there often enough though [17:09] the crash on closing is really easy to trigger on my desktop and laptop hardy installations [17:09] in any case let me know if you need extra informations [17:12] mvo: not in --check mode, only if someone installs a driver [17:29] mvo: not sure if you read that before, but could you do the bug #199402 sponsoring? [17:30] yo seb working on iso image icon [17:30] hey lapo, cool ;-) [17:31] seb128: I saw a comment of yours on #525153, are you involved in the development there? [17:32] lapo: let's say I'm rather doing active testing and small changes, I'm not one of the upstream maintainers for gvfs [17:32] lapo: what icon name do you suggest as fallback? [17:36] seb128: the media-encrypted approach is broken IMHO, since you can encrypt every possible media and you'll need to duplicate every possible media icon (bad, unmaintainable) [17:37] seb128: for the fallback just the unencrypted media icon it would use [17:38] seb128: I have serious dubts on own the whole gvfs icon handling is done [17:39] what do you mean? [17:39] this icon is used for luks mounts [17:40] indeed [17:40] the icon proposed shows an hd with a lock, but you could also encrypt a cd for example [17:41] so having an hd+lock is wrong in that case [17:41] the best approach imho would be to handle encryption with an emblem [17:43] tedg: are you planning on do the GPM 2.22.1 update? [17:44] pochu: Yes. Probably to 2.22.1-1 [17:44] seb128: is an "luks mount icon" needs to be provided better to have something really generic like a floder+lock with the folder as a fallback probably, not sure tho [17:46] seb128: sorry, missed that. let me have a look [17:47] lapo: no idea what luks is and if a cd can use that, but right an emblem seems to be a good approch, I don't think gvfs can set emblems though [17:47] lapo: anyway what icon should we use in hardy for now then? [17:47] seb128: will do that when I'm back, I take a break now [17:47] mvo: ok, enjoy! [17:48] seb128, luks is an encrypted FS [17:48] seb128: not sure how it is supposed to work, luks can be used on volumes I believe, so a folder is probably the best fallback [17:49] crevette: since you know what luks is, which STANDARD icon it should use? [17:50] lapo, I never use it, but I was planning to give it a try (since 1 year at least) [17:50] :) [17:50] eheh, same here more or less :-) [17:50] I'm not sure it could be applied to a folder [17:50] only [17:50] I think it handle block devices [17:51] yeah [17:51] so partitions, and loopback files [17:51] appart of a lockdonw, I down see [17:51] if (strcmp (hal_device_get_property_string (volume, "volume.fsusage"), "crypto") == 0) [17:51] is_crypto = TRUE; [17:51] don't see [17:52] lapo: that applies to filesystems apparently [17:52] so not to a directory [17:52] block devices then [17:52] yeah, not on specific folder [17:52] lapo: right [17:52] anyway using the hd icon to describe a luks mount looks broken to me [17:52] no? [17:54] that's why I was thinking about the folder that is the most generic thing possible to describe a mount point [17:54] I don't think a folder icon would be nice to list a partition in computer [17:54] I would take the volume icon + encrypted icon [17:55] now what should look like the encrypted icon [17:56] crevette: that's what I proposed, but looks like it's not possible atm [17:57] seb128: true tho [17:57] what icon was used previous cycle for those? [17:57] seb128: no one, I don't think we have it in the previous cycle [17:58] s/have/had/ [17:58] we had it I think [17:58] no idea then [17:58] pitti: did we had icons for luks devices previous cycles? [17:59] seb128: yes, that's a gvfs-age regression [17:59] apparently it wants to use separate icons for encrypted devices [17:59] but fails to fallback to the standard ones [17:59] pitti: do you remember what icon was used? [17:59] (human-icon-them does not provide encrypted icons) [17:59] pitti: right, we are trying to figure what the "standard one" is [17:59] pitti: neither does gnome-icon-theme [18:00] pitti: that's not a fallback issue, they just use a name which didn't exist [18:00] they implemented a new naming but the icon theme didn't catch up yet [18:00] seb128: I'm not sure the icon theme will ever catch it up eh :-) [18:01] seb128: there can't be a sane fallback tho, the less broken is probably media-harddisk :-/ [18:01] lapo: that will have to do for this cycle [18:01] lapo, perhaps a vaulted disk ? [18:02] seb128: it's not media-harddisk it's drive-harddisk, sorry [18:02] crevette: vaulted? [18:03] http://www.intelliot.com/blog/wp-content/uploads/2007/08/vault.jpg [18:04] okay time to back to home [18:04] see you there [18:37] slomo_: err, we still have taglib-sharp 2.0.2.0 :( [18:38] slomo_: I prepared the paperwork some time ago, but I forgot about it. do you want me to file a FFe? if so, I hope you can argue for it :) [18:45] yes [18:45] definitely [18:46] ok, on it [18:47] slomo_: have you tested it on hardy? (if not, we will need to do it for the motu-release team to ACK it) [18:48] pochu: no, test it if you want, i can guarantuee that it works fine ;) [18:48] heh, ok [18:52] hi there [18:53] if I need to uudecode a file in a debian package where should I put this file? under debian or patches ? [18:58] slomo_: I'll test it using Banshee (guess that's fine) [18:59] pochu: of course, if you can import music in banshee and the tags are right it works [18:59] bbl now [19:00] crevette: gnome-session has them in debian/ [19:01] slomo_: I guess banshee 1.0 is a no-go for Hardy, right? [19:01] pochu: there's no 1.0 before hardy release [19:02] pochu: just alphas until then... and they miss features, etc which stll need to be ported to the new code [19:02] hardy+1 and backports :) === fta_ is now known as fta [19:03] ok [19:03] * pochu tests taglib-sharp [19:06] crap [19:06] banshee crashed :/ [19:07] hmm, weird [19:07] it crashes on quit [19:07] slomo_: can you open banshee and close it without doing anything? :) [19:16] pochu: bug 199496 [19:17] blah, anyways that's the banshee/f-spot/tomboy/etc crash at exit, probably the one you're having with banshee too [19:18] wow, 127 dups [19:18] pedro_, slomo_: i guess it should be milestoned as we ship tomboy and f-spot by default ;) [19:20] mm i thought it was, ok it's now [19:23] thanks [19:25] hmm, now I've found that Rhythmbox won't update the database automatically... [19:25] so it still has the old metadata [19:25] (I've looked at it with decibel-audio-player too and it's fine there) [19:27] hmm, moving out .gnome2/rhythmbox and .gconf/rhythmbox and reloading the library didn't change anything... [19:27] thanks pochu [19:28] can I put several pixmap in one uue file ? [19:28] * pochu checks with easytag [19:28] I'm not used to that kind of file [19:28] neither am I :) [19:30] easytag shows the files have changed, so I must be doing something wrong with rhythmbox... [19:30] slomo_: anyway banshee/taglib-sharp changed 14 files in a row, so I'm more comfortable now ;) [19:33] How stable is this new tool to configure several screens? [19:33] I just tried it and it failed [19:35] gnome-display-properties? [19:36] dennda: did you file a bug report btw? [19:37] pochu: not yet. I am here with some of my fellow students doing some work. We are still busy. Just wanted to know how well the tool works on an intel X3100 because I was unable to adjust the external screen properly. Just out of interest [19:37] pochu: I will examine that later [19:37] (How well it is supposed to work, that is) [19:38] just changed the orientation here and it worked fine [19:39] X3100 too [19:39] pochu: with an external screen? [19:39] nope, an vga monitor [19:39] (this is not a laptop) [19:40] With gutsy I have a somewhat strange behaviour: My laptops screen size gets adjusted to the size of the external one. The external one fits just fine but the laptops display is just cut off [19:41] I havn't had enough time to check how it behaves with hardy, though [19:41] I hope it has become better. That's something really important [19:49] does pbuilder produces a build log ? [19:51] crevette: not by default. there's an option though [19:51] okay [19:51] thanks [19:52] pbuilder build --logfile /some/path *.dsc [19:52] pochu: I assume you don't have any additional screen standing around? [19:52] okay, you're great [19:52] thanks [19:53] dennda: I don't. perhaps the folks at #ubuntu-x know about any issues with it [20:00] asac: you said in jwendell's bug report you have uploaded liferea -0ubuntu4... did you meant -0ubuntu3? I haven't seen ubuntu4 in hardy-changes yet :) [20:00] yes u3 [20:00] typo [20:00] pochu: ^^ [20:03] ok, thanks [20:04] slomo_: http://pastebin.com/m1be2eaa6, fine with you? (specially the part where I mention you) [20:16] pochu: sounds good [20:16] and segv on quit is a known mono or gtk# bug [20:16] ok, sending [20:16] good work :) [20:17] now I'll expect you to add some convincing comments to it ;) [20:17] * crevette can't have find and uuencode working together [20:17] and sorry for forgetting about it :( [20:26] slomo_: https://bugs.edge.launchpad.net/ubuntu/+source/taglib-sharp/+bug/209874 [21:17] seb128: what are you guys doing about this issue? http://bugzilla.gnome.org/show_bug.cgi?id=482354 [21:23] is a set of icons can be provided in a tar.gz instead as a uue set of files ? [21:28] walters: nothing yet [21:28] walters: too many things to do, and I don't use firefox so I'm not annoyed by it, dunno if you got lot of demands for a change [21:28] seb128: ok, keep me updated please on what you decide; I've been trying for months to fix this aspect of the linux desktop, and I'd really like to see it happen [21:29] asac: ^ do you know about that? [21:29] crevette: no, you can't diff a binary format and tar.gz is one [21:29] okay :/ [21:38] walters: there's https://bugs.edge.launchpad.net/ubuntu/+source/firefox-3.0/+bug/175904 (which links to the metacity bug), which is milestoned for Hardy. asac is aware of it but I don't know what he has decided to do [21:39] pochu: thanks, i've subscribed [21:40] seb128: its a issue metacity vs. compiz [21:40] asac: it's not [21:40] compiz doesn't raise as aggressively [21:40] metacity does [21:40] it's something which is not clear documented by the spec [21:41] and that the wm implement as they want [21:41] technically that's not a bug [21:41] yes ... thats why i say metacity vs. compiz [21:41] :) [21:41] both compete to set a standard [21:41] well, it's not a bug, nothing say they have to be consistent [21:41] but given that metacity had the compiz behaviour before, i'd say that metacity breached a de facto standar [21:41] it's almost like having multiple free implementations of the same thing is a bad idea... [21:42] where did i say its differnt [21:42] differnt [21:42] ouch [21:42] I think the previous behaviour (the window bar in the window list highlighted) was good, but this one is highly annoying [21:42] i just say: once the behaviour changes in a way that without adding tweaks to the app people feel annoyed its a problem [21:42] and maybe an indicator that it might have required multiple steps :) [21:43] but then i see how firefox could behave better. [21:43] and if it does, the metacity behaviour would be superior [21:44] i haven't tried how the metacity patch feels [21:45] maybe its even a gtk issue that it doesn't support more fine grained control on what gtk_window_present does [21:45] is present an atomic WM operation? [21:45] I would argue that's a firefox bug [21:46] it should add tabs if there is some firefox windows on the current workspace or viewport [21:46] and open a new one if there is none there [21:46] i don't think its a real bug anywhere [21:47] it just happens that everything together is now annoying :) [21:47] last time i looked there was no simple api support for this kind of logic [21:47] in gtk/gdk [21:48] which indicates that there is definitly room for improvement for other parts of the stack too [21:48] but i agree that in the end firefox should handle it [21:48] preferably through a simple api [21:48] right, gtk doesn't allow to do that easily [21:48] seb128: well, if I have "open links in a new tab", I'd expect firefox to do that even if I'm on a different workspace [21:48] which is due to the fact that gtk didn't know what metacity would do at some point imo [21:49] so I don't think opening windows if there's no window in the current workspace is good behaviour [21:49] pochu: well, in this case firefox should not call the present function then and just silently add the tab [21:49] right [21:49] seb128: but still it hsould raise on other desktop [21:49] why? [21:49] and blink in window taskbar [21:50] if you decide to add tabs you likely know what you are doing and where the thing is [21:50] or isn't it possible in gnome to show all windows in the window taskbar? [21:50] that's possible [21:50] i think somewhere i had a blinking window task and could click on that to zoom to that desktop [21:50] right click on the task list corner and properties, you have an option there [21:50] i would like that behaviour if open in new tab is used [21:50] not sure if that can be done without present [21:52] you see what i mean by this "present on other desktop" use-case? [21:52] yes [21:52] yeah. [21:52] what pidgin does when you receive an im message [21:52] hard to chew on all this [21:52] or xchat [21:52] they blink [21:52] but you don't need to present the dialog for that [21:53] present raise it to the foreground too [21:53] which is different of asking attention [21:53] the present api is give me this on screen now so I can use it [21:54] anyway ... my short term solution is to turn of "open new links in tab" .. and use "open new links in new window" instead [21:54] which is probably more comprehensible for the normal ubuntu user anyway [21:54] ehrm [21:54] no [21:55] I don't want to hold firefox please. [21:55] Nafallo: those are options, why would you hold firefox rather than changing the settingµ? [21:56] hold firefox? what does that mean? [21:56] seb128: ah. I read "turn of" in a totally different way :-) [21:56] asac: he means not upgrade to not get the behaviour you describe [21:56] I thought it was hide the option or so :-P [21:56] * Nafallo totally relies on tabs at work :-) [21:56] ok i added ubufox and milestoned it for that setting change [21:58] maybe we get the a better solution for the .1 release [21:58] as long as I have the choice to switch... ;-) [21:58] Nafallo: you have :) [21:59] unless you add a lockPref on your own :) [21:59] hhee [21:59] s/he/eh/ [22:00] you can lock prefs ... e.g. if an admin doesn't want students or kiosk systems to be changed pref-wise [22:36] asac, one more reason for me to not use ubufox. browser.tabs.loadDivertedInBackground is not perfect but it's best workaround we have if the metacity patch is rejected. [22:38] tabmix+ or another tab extension should do to restore the full expected behavior [22:38] the patch is not rejected [22:38] I don't really care about using it or not but I would prefer if it was special casing firefox in the behaviour change [22:38] well, above, you didn't seem to want it [22:38] I don't have a strong opinion [22:38] but I think firefox is buggy in what it tries to do