=== asac_ is now known as asac [04:14] Good morning [04:14] dobey: yup, saw the MP, will do this morning === pitti_ is now known as pitti === ejat- is now known as ejat === ejat- is now known as ejat [06:07] pitti: Good morning! Do you know how access control to sane/cups/etc devices is handled? [06:07] Do we have udev rules + an ACL, or something different? [06:07] hey everyone...having a heck of a day with compiz...anyone else having issues? [06:08] RAOF: cupsd runs as root, so we don't do much of ACL there [06:08] RAOF: for sane and gphoto cams we do use ACLs, see /lib/udev/rules.d/70-udev-acl.rules [06:08] ENV{ID_GPHOTO2}=="*?", TAG+="udev-acl" [06:08] ENV{libsane_matched}=="yes", TAG+="udev-acl" [06:09] RAOF: see /lib/udev/rules.d/40-libsane.rules and 40-libgphoto2-2.rules [06:09] Aaah! I saw /lib/udev/rules.d/libsane, but that applied an ACL for a non-existent group. [06:09] yeah, we don't use the scanner group [06:10] hey jasoncwarner_ -- it got a bit better with the most recent compiz, but I still get crashes [06:10] I'm running unity trunk as of last Friday evening, and that's doing pretty well. [06:10] pitti: how should I handle gnome-tweak-tool's missing m4 directory? [06:10] pitti: I keep getting window decorator crashes [06:11] Although it's still leaking gem_objects, so the OOM killer starts kicking in after a while. [06:11] jbicha: hm, I'm not sure [06:12] I thought I could just do it in the rules file as it seems to me a bit weird to do a patch for that [06:12] jbicha: does it just require an mkdir/, or also files in it? [06:13] it doesn't actually need anything in the directory, but I believe autoconf chokes if that directory doesn't exist [06:15] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565663 [06:15] Debian bug 565663 in automake "autoconf: autoreconf fails if macro dir does not exist" [Normal,Open] [06:17] pitti: Hm. And what does udev-acl actually do? I'm trying to ensure that colord gets access to the appropriate devices (scanners & printers) without root. [06:22] RAOF: ACLs aren't meant for system daemons [06:22] RAOF: they are meant for giving device access to the user who is currently on the foreground console [06:23] RAOF: udev-acl and consolekit cause an extra rw ACL to get added for ^ [06:23] Aha. [06:24] jbicha: hm, then mkdir'ing in rules before calling configure or autoreconf ought to do it indeed [06:24] pitti: So I need to add some explicit magic so the colord user has access to the appropriate devices? [06:25] And by "magic" I presumably mean "another udev rule, applying a rw ACL for the colord user". [06:25] RAOF: is colord a system daemon or running in the user session? [06:26] System daemon. [06:26] (dbus activated, but system wide) === tkamppeter_ is now known as tkamppeter [06:27] RAOF: and it needs to talk to scanners? [06:27] for that it could ensure that the scanner system group exists and put itself in that [06:28] It is, by upstream default, run as root, but kees was concerned that it's vector for exploits; it allows the user to pass arbitrary data to lcms, and lcms has apparently had security holes in the past. [06:28] for printers, it should talk to them through cups, otherwise cupsd and colord would race for device access [06:28] Well, really it wants to use sane to talk to scanners. I _think_ it already talks to printers through cups, tkamppeter might know better. [06:29] libsane presumably [06:29] Indeed. [06:29] libsane. [06:29] so, put it into "scanner" and "lp" [06:29] (groups) [06:29] I still don't think that it's such a good idea to have this as a separate daemon, as it will provoke device locking race conditions, but that's an upstream design problem, not an "access permission" one [06:30] well, for scanning there is no daemon, so it has little other chance [06:30] Right. [06:32] My understanding is that it's not going to fight for the scanner very often, and only in response to a user request (ie: I've stuck a standard colour-profile image on my scanner, kindly calibrate it for me). [06:36] It also wants to be able to enumerate scanners, as they're the sort of thing you want to calibrate :) [06:40] pitti: I did that in my rules file, but that didn't work when you tried to build it? [06:41] apparently not; I need to try again [06:45] RAOF, AFAIK colord only talks with CUPS and not with the printers directly, but I do not really know about the internals of colord. [06:51] Hey, there we go. Now colord sees the hpaio scanner. [07:28] jbicha: aah [07:28] jbicha: your debian/rules patch has the mkdir commented out [07:28] jbicha: how about I comment that in, and move it into post-patches::? [07:29] oops sorry about that, thank you [07:29] +post-patches:: [07:29] + [ -d m4 ] || mkdir m4 [07:29] somethign like this? [07:30] jbicha: btw, can I ask you to keep the changelog target as "UNRELEASED" as long as this isn't really uploaded? [07:30] yes, post-patches sounds like a better place [07:30] ok, builds now [07:30] pitti: ok, I'll use UNRELEASED from now on [07:30] thanks [07:31] * pitti pushes and uploads [07:32] hrm. did we get rid of the settings dialog for setting the font sizes in oneiric? :-) [07:33] Nafallo: You can fiddle with them in gnome-tweak-tool, but there's no by-default way of doing that, no :( [07:34] hehe. seems there was a lot of settings that disappeared. hopefully they'll come back on whatever p will be. [07:34] Nafallo: you can also try System Settings>Universal Access and then pick a Text size, but it's not very precise [07:35] okay. at least I've got options now. thanks guys. [08:07] morning [08:09] hey rodrigo_, how are you? [08:10] bonjour seb128 [08:11] hey rodrigo_ pitti, how are you? [08:11] pitti, I'm great thanks ;-) [08:13] hi pitti, seb [08:13] morning all [08:13] hey Sweetshark, had a nice weekend? [08:16] seb128: hm, just noticed the "things to sync on Debian" section [08:16] seb128: there are no names who requested them; how do you handle this usually, -b seb128? [08:16] pitti, well that was rather for the early in the cycle before the freezes [08:17] pitti, I just noticed that people put things there recently, 2 of 3 listed have sync request bugs [08:17] ah, will look there, thanks [08:18] pitti, danke [08:18] * pitti syncs gtkmm2.4 himself then, this is good and harmless [08:20] all syncs done [08:21] pitti, but yeah, I tend to sync with my name when I do those early in the cycle if people don't put their name next to the sync [08:21] and updated the pad accordingly [08:46] good morning everyone [08:46] hi chrisccoulson [08:47] hey chrisccoulson [08:47] hi pitti, rodrigo_. how are you? [08:56] hey chrisccoulson, how are you? [08:57] chrisccoulson: quite fine, thanks! had a nice weekend [08:57] hi seb128. i'm good, but a bit tired thanks [08:57] my daughter kept me awake for most of last night [08:57] we did quite a lot of garden and house work on Saturday, and went swimming/idling yesterday :) [08:57] oh :-( [08:57] is the ok? [08:57] the->she [08:57] chrisccoulson: teeth again or did she catch a flu? [08:58] i'm not too sure. she seems fine this morning. i'm not sure why she doesn't feel tired like me ;) [09:09] oh, i already had gnome-screensaver to go [09:09] yeah, not sure why jbicha did it while it was assigned to you on the etherpad, I guess it was an overlook :-( [09:09] it seems like i shut my eyes for a few hours at the weekend and everyone takes stuff i was working on ;) [09:10] mdeslaur also reverted the changes that give it the gnome-shell look [09:10] chrisccoulson: yes, sorry about that, didn't see until after I had already finished it [09:10] but he's reverted it for all sessions [09:10] chrisccoulson, did he? where? [09:10] seb128, he pinged me on IRC yesterday [09:10] oh ok [09:11] rodrigo_, oh, you have up on fighting to keep the 2s pad delay? ;-) [09:13] huats, lut [09:13] huats, how is the gktsourceview update going? [09:14] asking here as haven't found an answer on #ubuntu: how do I recover from a black screen, where all programs are still running normally, just screen is black, without rebooting, and while keeping programs open? [09:16] Shambler[Bishop, hi, try #ubuntu for user questions [09:17] ok. is there a more dev-oriented channel I can ask in? solving my problem without losing open programs, seems like it might require dev smarts [09:18] Shambler[Bishop, not really, dev channels are not support ones [09:18] seb128, I should be able to finish it tomorrow (today I have a big deadline) [09:18] fair enough [09:18] Shambler[Bishop, but I would try to stop the screensaver and to restart compiz if I were you [09:18] seb128, I'll let you know [09:18] huats, ok [09:19] yep screensaver is already off [09:19] Shambler[Bishop, without any description of what version of ubuntu you use, what desktop, what you were doing etc it's almost impossible to help you there [09:20] chvt and xrandr might be of some utility [09:20] seb128, oh, the new thunderbird theme does work with the light radiance theme btw [09:20] latest, in virtualbox, the new desktop with everything moved to side of screen, no screensaver, wasn't doing anything at time (it was left to its own devices, encoding a dvd in the background, as it is still doing now) [09:20] chrisccoulson, why not? [09:20] i needed to restart thunderbird when i tried it though ;) [09:20] I think gfx driver has crashed or something [09:20] seb128, last time i tried it, it didn't work [09:20] but after i restarted it, i realized that it did actually work [09:21] the icons get their colours from the theme [09:21] oh ok, I read "doesn't work", dunno why ;-) [09:21] chrisccoulson, great [09:21] so it looks pretty good with radiance too [09:22] is the default in the default install now btw? [09:22] seb128, no, but it will be later today [09:22] excellent ;-) [09:28] ok, let's finish with the updates [09:29] chrisccoulson, want to claim one or two updates? [09:29] seb128, yeah, can do [09:29] \o/ [09:29] hm, firefox seems to have forgotten to scroll down when marking text and pushing below the lower edge [09:29] seb128, not gave up, but I thought we'll try and see if this introduces more bugs than it fixes [09:30] rodrigo_, why would it introduce bugs? [09:30] it was the default for several cycles [09:30] seb128, that's what I've been told [09:30] seb128, 2.0s was the default [09:30] 0.5s just tends to be a bit short and you still get palm clicking sometimes [09:31] rodrigo_, what bastien said is that 0.5s leads to palm clicks because it's a bit short [09:31] seb128, but asI said, we'll see, if nobody complains we'll be ok :) [09:31] but we lived with it without complain for some years [09:31] yeah [09:31] the 2s just feels buggy [09:31] you get a good 1s where you try to use the pad and where it doesn't work [09:31] yeah [09:32] pitti, want to do the gvfs update? [09:32] pitti - that's a known bug [09:32] seb128: yes; I still need about an hour or two for jockey, then I can get to this [09:32] we had a distro patch to fix that in firefox 6, but this code got completely rewritten in firefox 7 [09:32] pitti, no hurry, thanks ;-) [09:32] the new -updates fglrx/nvidia driver packages hit around FF, and it was not so trivial to show them [09:32] but I got the underlying bug fixed now [09:33] ok, I will do updates later, I need to pilot first ;-) [09:33] chrisccoulson: ugh, that's still not fixed upstream? weird that so few people c&p from firefox, I seem to do that rather often [09:33] pitti - yeah, me too. it's always been a problem, but it was masked by the presence of the statusbar in old versions [09:33] so we carried a patch since firefox 4 to fix that [09:34] but the patch needs a big rethink now [09:37] pitti - https://bugzilla.mozilla.org/show_bug.cgi?id=644621 [09:38] Mozilla bug 644621 in Selection "Drag selection scrolling does not work properly in fullscreen and maximized mode" [Normal,Reopened: ] [09:39] chrisccoulson: ah, so if there was an one-pixel border at the bottom, it would work? [09:40] pitti - yeah [09:40] jbicha, pitti: do you have any clue about [09:40] https://code.launchpad.net/~ubuntu-branches/ubuntu/oneiric/gnome-tweak-tool/oneiric-201108220805/+merge/72388 [09:40] I will never understand why those happen I think [09:41] I actually used ubuntu:gnome-tweak-tool for sponsoring this [09:41] cyphermox: By the way, bug #828556 is available for your delectation. [09:41] Launchpad bug 828556 in network-manager "When wireless is enabled, anything touching the network deadlocks in the kernel" [Undecided,New] https://launchpad.net/bugs/828556 [09:41] it happens when I push the current upload to bzr, but then the package importer sees that the actual .dsc doesn't quite match what's in the branch [09:42] pitti, why would the Categories be different? [09:42] I have no idea :/ [09:42] ok, thanks [09:45] bleh, DSL reconnect [09:45] if the archive version is right and the branch is wrong, the merge should just be rejected [09:45] otherwise we need to reupload; perhaps the patch application got wrong at some point [09:45] seb128: ^ [09:46] pitti, ok thanks [09:48] Bah. Are there any good tools for working out what's dbus-activating colord about 15 seconds into boot? [09:49] RAOF: hm, not that I know of; you could perhaps launch dbus-monitor early in the boot and have it log into a file? [09:50] RAOF: but perhaps it's easier to grep /usr/bin/, /usr/sbin etc. for the d-bus name of colord [09:50] pitti: I did unapply the .desktop patch before I did my merge proposal, maybe that's a bad idea [09:50] pitti: Hah! When all else fails, grep the binaries? :) [09:50] jbicha: for this reason I really hate the UDD branches; preapplied patches are EBW [09:50] and there's no way to not break them, unless you tear apart a chicken [09:51] jbicha: so seems we need another upload there? [09:51] Ah. Hello cups! [09:51] jbicha: which category is right, "settings" or "utility"? [09:52] who can I talk to about user groups in ubuntu? [09:52] settings [09:52] unix user groups, that is [09:53] and I'll make sure I have reapplied patches before uploading [09:53] rodrigo_: me [09:53] pitti, ok :) [09:53] the ubuntu-desktop branches are so much easier to do the right thing with [09:53] pitti, so, my question is about the wheel group [09:54] rodrigo_: I'll be right back with you, need to restart session [09:54] keyring went AWOL [09:54] pitti, ok [09:56] rodrigo_: so, wheel group? [09:57] pitti, yes, so upstream accountsservice uses the wheel group to add admin users [09:57] on ubuntu that just adds the user to 'adm', iirc [09:57] we use the "admin" group for admins [09:58] so, if I'm right,since wheel group is recognized, it should do the correct thing, which is to add the user to the groups in https://bugs.launchpad.net/ubuntu/+source/gnome-control-center/+bug/810907 [09:58] Ubuntu bug 810907 in gnome-control-center "User Accounts profiles need to be updated for Ubuntu" [High,Confirmed] [09:58] pitti, yes, but wheel is a pretty standard thing, isn't it? [09:58] rodrigo_: or rather, it was thirty years ago :) [09:58] rodrigo_: we don't really use it, as we use sudo [09:59] my question is if we still have that group, should we patch apps or just make it do the correct thing? [09:59] pitti - if you want to work around the scrolling issue btw, just use ctrl+f to open the find bar ;) [09:59] that is, if you use the wheel group, create an admin user with the correct groups in [09:59] or ctrl+/ to turn on the addon bar [09:59] chrisccoulson: ah, nice [09:59] uhm... hi. is nspluginwrapper in oneiric broken just for me or for everyone? [10:00] rodrigo_: the concept of wheel is rather incompatible with the concept of sudo [10:00] broken = uninstallable [10:00] rodrigo_: the consistent Ubuntu thing is to put admins into "admin", and that's it [10:00] adm is fine, too [10:00] pitti, patching apps = upload a patch I have for accountsservice to add users to all needed groups in ubuntu [10:00] BigWhale: you need to enable multiarch [10:00] humm [10:00] how do I do that? [10:00] rodrigo_: "wheel" is for users which are supposed to mess around in /usr/ without any extra passworrd/su/sudo [10:01] BigWhale: slangasek sent a mini-howto to u-devel@ recently [10:01] pitti, hmm, ok [10:01] pitti, I'll try to find it [10:01] thanks [10:01] rodrigo_: normal users should be put into no group at all (except their own), or at most "sambashare" [10:01] Hi, everyone. I want to change the default Korean font in ubuntu-desktop. How can I do it? [10:02] rodrigo_: admins should be put into "admin", "adm", and "lpadmin" [10:02] rodrigo_: .. and "sambashare" [10:04] jincreator: you mean for Ubuntu in general, or just for your local box? [10:05] jincreator: there are two places for it to change: (1) the set of default packages in the seeds (ttf-whatever), which will define the "ubuntu-desktop" metapackage [10:05] jincreator: and (2) the fontconfig configuration, which is mostly in the fontconfig-config package (fontconfig source), plus some tweaks in language-selector [10:19] I come back from vacation and everything seems different in Oneiric. :> [10:20] seb128: do you know whether we still actually need those 99_ltmain_as-needed.patch patches? [10:20] pitti, we don't [10:20] BigWhale: welcome to the feature freeze rush :) [10:20] ubuntu start button is now in the launcher? [10:20] for good? [10:21] I wouldn't call it "good" [10:21] pitti, they are not working anyway for things using dh-autoreconf since the patches copy is overwritten by dh-autoreconf [10:21] it broke corner activation [10:21] I can't activate the dash with the mouse any more [10:21] I don't like it there... :/ [10:21] yeah, it's also quite pointless [10:22] pitti, dh_autoreconf got a --as-needed which is equivalent though [10:22] " * dh_autoreconf: Add --as-needed for automatic patching of ltmain.sh" [10:22] but I didn't try it [10:32] so the desktop dailies grew again by 2 MB or so [10:33] dpm, Hi, can we start reporting bugs about untranslated strings in ubiquity or there is work in progress ? [10:36] jibel, string freeze is not until 3 days time IIRC. However, people have already started translating it. So as long as people have been keeping up with untranslated strings and ev has been exporting and committing translations, it should be fine to report bugs. Better double-check the ubiquity logs/commits to see if there have been recent translation commits first, though [10:39] dpm: hey David, how are you? [10:39] dpm: we need a new -base langpack refresh for beta; can we get an export towards the end of the week? [10:39] I'd like to upload fresh langpacks on Friday, so that they can build over the weekend [10:39] hey pitti, very well, thanks :) [10:39] let me have a look... [10:41] tkamppeter: How familiar are you with hplip's sane integration for multifunction printers? What are the prerequisites for that to work? colord currently gets started early in the boot process (I think by cupsd), and it doesn't pick up my hp multifunction scanner until I kill the daemon and start it again. [10:41] tkamppeter: So I suspect that maybe the sane device isn't available until cups is fully up, and colord doesn't notice the hotplug. [10:44] tkamppeter: For more information, this is a wireless printer, so there's no physical connection to my system. [10:46] pitti, sounds good, would Thursday work well for you? We've got an export scheduled already there (https://dev.launchpad.net/Translations/LanguagePackSchedule). The only caveat is that lately full exports have started taking up to 24 hours (bug filed in LP) and might not be ready until Friday afternoon. If that works for you, then we only have to request a full export on Thursday. If you think it's too tight, we can ask the LP folks to start the expor [10:46] t earlier. [10:49] dpm: what is afternoon there? [10:49] anything until 1400 UTC would work for me, but not afterwards (as it needs some 2 hours to prep them) [10:50] pitti, do you know what added the 2mb? [10:51] pitti, yeah it'd be 14:00 UTC (assuming it takes 24h), but I cannot tell for sure. To be on the safe side, I'll ask the LP folks to start the export Thursday morning instead of Thursday 14:00 UTC [10:55] dpm, okay. thanks! [10:56] jibel, yw :) [10:56] mvo, hey [10:56] mvo, could you review https://code.launchpad.net/~evfool/update-notifier/fixwording/+merge/69992 [10:56] mvo, trivial one ;-) === MacSlow is now known as MacSlow|lunch [11:22] pitti - what caused the dailies to grow in size? [11:23] it's not me this time is it? [11:23] chrisccoulson: I don't know yet, we don't have an earlier one any more to compare [11:23] (the ones which were 703 MB) [11:23] I could compare them to alpha-3, and just check what grew [11:24] but this probably makes more sense after a -base langpack refresh [11:24] * pitti -> lunch, bbl [11:24] actually, it looks like firefox shrank a bit [11:24] (~200kB) [11:24] nice! [11:24] seb128: want me to update gnome-common after lunch? [11:25] can also do simple-scan, I can test it [11:25] that probably means that thunderbird will shrink by a similar amount [11:25] although I had expected Robert to do that one :) [11:28] bug #830377 needs some consideration for those that use NFS for $HOME. [11:28] Launchpad bug 830377 in lightdm "autofs5 needs to start before lightdm" [Undecided,New] https://launchpad.net/bugs/830377 [11:34] pitti, if you want to do those please do, thanks [11:34] pitti, seems like robert_ancell decided to stay away from packaging this cycle ;-) [11:43] seb128, gedit is blocked on the new pygobject btw [11:43] is it? [11:44] seb128, http://git.gnome.org/browse/gedit/commit/?id=9d5cd427873809f91bbef305033b87246718fb43 [11:45] chrisccoulson, it somewhat seems upstream being overzealous on versions [11:45] seb128, yeah, it doesn't look like there's anything which actually requires it [11:46] chrisccoulson, revert the commit? ;-) [11:47] seb128, yeah, will try that [11:52] anyone has a debian system at hand? [11:52] oh, I have one! my backup server, sorry for the noise :-) [11:53] rodrigo_, ;-) [11:56] pitti: thanks! [11:57] seb128: I have a look at this [11:57] mvo, danke [11:59] hey seb128 ...dbarth caught me up on compiz and them backing out window decorator etc [11:59] hey jasoncwarner_ [11:59] allo! [12:00] they do back what out? === zyga is now known as zyga-siesta [12:20] what was the name of that gnome tool to import music from a cd, I completely forgot [12:20] ? [12:20] ah, sound juicer! [12:20] * rodrigo_ shouldn't ask questions so quick :) [12:21] right [12:21] I'll go for lunch now, to recover my brain :) [12:21] bbl [12:21] chrisccoulson, \o/, desktop updates! ;-) [12:24] pitti: ping? hi, could i get a priority bump on those test builds please? https://launchpad.net/~unity-team/+archive/compiz-testing/+builds?build_state=pending [12:26] dbarth: sure, done [12:27] pitti: thanks :) [12:45] pitti: btw, are you going to do a new release with that change? or should i make an upload with that patch for ubuntu? [12:46] dobey: already happened :) [12:46] pitti: great! [12:46] can fix my ftbfs then [12:53] seb128: done [12:54] seb128: I guess at this point I should concentrate my efforts on porting the U1 control panel to GTK3, to unblock the pygobject update? [12:54] pitti, is that blocking the update? [12:55] pitti, if I had to pick something that needs to be done it would be "fix the retracers" [12:55] we need those back === MacSlow|lunch is now known as MacSlow [12:57] seb128: yes, u1-c-p and python-ubuntuone-client mix static and GI, and break with new pygobject [12:57] I have the library fixed (that's easy), and the control-panel half done; but they do some rather complicated low-level hacks which take a bit to port [12:57] ah, right, retracers.. [12:57] ok [12:58] bbiab, supermarket [13:16] * pitti 's brain melts down.. [13:19] * mdeslaur cools pitti's brain with dry ice and liquid nitrogen [13:20] aaah -- although I would have preferred an ice cream, but oh well :) [13:20] but getting up at 6 does have its benefits these days -- you have most of your day over when it starts getting hot === hggdh_ is now known as hggdh [13:34] seb128: hey so, I had a hard time finding folder sharing in 11.10, and I (think) we're supposed to be using nautilus-share right? It appears to be an extension or something. [13:36] RAOF, the scanner driver of HPLIP recognizes the availability of scanners via the existence of CUPS queues with the hp: backend. So seems CUPS must be listening so tha the scanners get listed by SANE. [13:38] RAOF, perhaps you should add a wait loop which checks for CUPS listening every second and that until a timeout of 10 seconds. Then list the scanners. === ara_ is now known as ara [13:41] jcastro, it's broken [13:50] RAOF, file/device permission settings are done by the UDEV rules shipped with the HPLIP package. [13:57] seb128: Hi Sebastien, I have noticed that you often monitor bugs to ensure that they are properly dealt with. Could you please take a look at bug 770091? I really think that we shouldn't ship that bug with Oneiric too. [13:57] Launchpad bug 770091 in ubiquity "Unsolicited attempt to import files and settings" [Undecided,Confirmed] https://launchpad.net/bugs/770091 [13:57] pitti: Hi Martin, this is a simple example of my problems in Oneiric as regards compliling glibc stuff: [13:57] http://gunnar.cc/ubuntu/compile-failure-oneiric-example.txt [13:57] In Natty that code compiles and runs fine. [13:58] GunnarHj, hey, ok [13:58] GunnarHj, the build failure is due to --as-needed [13:58] GunnarHj: try swapping it: [13:58] GunnarHj, try putting the "test.c -o test" part between gcc and pkg-config [13:58] gcc test.c `pkg-config --cflags --libs glib-2.0 gobject-2.0` [14:00] pitti, seb128: Thanks for the suggestions, I'll try those variants later. [14:00] GunnarHj: or gcc `pkg-config --cflags glib-2.0 gobject-2.0` test.c -o test `pkg-config --libs glib-2.0 gobject-2.0` [14:01] the important part is the the test.c (or test.o if you compile it seperately) comes before the libs it needs (pkg-config --libs) [14:02] seb128: "Due to --as-needed", can you please elaborate? [14:05] GunnarHj: oneiric's ld uses "--as-needed" by default which results in libs getting only linked if object files (or other libs) *before* the -llib parameter in the ld call uses any symbols of that lib [14:10] GunnarHj: see also https://wiki.ubuntu.com/NattyNarwhal/ToolchainTransition#How_to_Fix_a_Problem [14:16] geser: Thanks a lot for those explanations. (I didn't read that remark by seb128 too carefully, and I thought it was related to the bug report I asked him to look at.) [14:18] I'm getting a lot of dbus disconnects and timeouts with a bunch of different applications in oneiric...is that a known issue? [14:19] uhm...maybe gconf and not necessarily dbus [14:20] mdeslaur, no, how does it manifest? [14:20] seb128: Evo shows an error dialog, a python app I use triggers apport, etc. [14:21] seb128: it's intermittent...and jdstrand is seeing it too [14:27] mdeslaur, does dbus-daemon use a lot of cpu when it happens? [14:27] chrisccoulson: good question...I'll check next time it happens [14:28] It probably happens to me about twice a day [14:28] mdeslaur, i get an intermittent issue where indicator-datetime-service keeps spamming the session bus, causing everything else on the bus to grind to a halt [14:28] (including compiz) [14:28] it basically results in everything using dbus to lag horribly [14:28] perhaps you see a similar issue ;) [14:28] chrisccoulson: hmm...could be related...do I just check top? [14:29] mdeslaur, yeah [14:29] chrisccoulson: ok, will do, thanks. [14:30] guys is it me or does nothing display in dash-lenses unless you type something into the search bar, too use to everything being there and getting refined by the search feature [14:30] davmor2, that's known already [14:31] chrisccoulson: oh great just checking it wasn't an issue here :) [14:32] rodrigo_: Hi Rodrigo, have you seen my latest comment on https://code.launchpad.net/~gunnarhj/gnome-control-center/langfix/+merge/71197 ? Any chance that we can process that MP today? [14:33] GunnarHj, no, didn't see it, looking now [14:33] pitti: care to also sponsor https://code.launchpad.net/~dobey/ubuntu/oneiric/ubuntuone-installer/fix-ftbfs/+merge/72429 please, as it needs the new distutils-extra? :) [14:36] re [14:36] dobey: can do [14:36] bah, got a load around 9 and xorg closed... [14:36] pitti: great, thanks! [14:36] re seb128 [14:36] hey dobey [14:37] hate oneiric, the load on my box is most of the time around 1, cpu fan spin and nothing is showing up in top, the process list, iotop, dbus [14:38] seb128: well t hose extra 5 watts have to go somewhere i guess :) [14:38] if only I could figure what is making it [14:39] yeah, i know how you feel :) [14:39] dobey: upped [14:40] I'm close of reinstalling natty [14:40] pitti: cheers [14:40] it's hard to get any work done with a sluggish box [14:41] seb128: do you have a lot of IO in the background? [14:41] pitti, no, it's an idle box [14:41]