[00:13] * TheMuso hasn't upgraded, so not experiencing it. [00:13] RAOF: Do you use pulse much/listen to audio much in maverick? [00:14] A certain amount. [00:14] I still haven't got around to setting up my stereo system after moving, so I don't yet do the more fancy stream-over-wifi that I did in Sydney. [00:17] Apart from volume being muted each time I start my session, everything seems to be working. [00:26] RAOF: Oh thank you. thats what I was goign to ask about. [00:26] RAOF: I get that too, and I have a theory about that, which I am not sure how to test. [00:26] What's the theory? [00:27] RAOF: But I can tell you if you were to kill pulse running as your user before you restarted, your data would be saved and your next session would have sound. [00:27] Basically, when you restart, pulse gets a SIGINT/SIGTERM, and starts to save its data, but it gets hit with a SIGKILL before it finishes what it needs to do. [00:27] This is not an ALSA problem, and as I siad, if you kill pulse before restarting, the data is saved properly. [00:28] gah typing [00:29] Heh. Took me a couple of passes to spot the mistake :) [00:29] Would you like me to confirm that behaviour for you? [00:30] RAOF: If you don't mind. Firstly confirm that you kill pulse and your volume data is saved properly first I think [00:30] RAOF: I have a bug number, let me dig it up. [00:31] I'll do it when I restart for updates. [00:31] ok cool thanks. [00:32] RAOF: bug 592016 [00:32] Launchpad bug 592016 in pulseaudio (Ubuntu) "[Maverick] Sound is always muted on startup; unmuting makes the volume at lowest level (affects: 11) (dups: 1) (heat: 104)" [Undecided,Confirmed] https://launchpad.net/bugs/592016 [00:47] Hrm no nvidia driver updates yet, so I have to either wait, or switch back to Nouveau. === oubiwann-away is now known as oubiwann [00:48] tselliot was working on an upload last night; it looked like he'd got all his ducks in a row, but it hasn't been uploaded yet. [00:51] np [01:04] * RAOF looks forward to blowing away this btrfs and replacing it with ext4 tonight, so that updates don't kill performance for quite so long. [01:04] heh [01:09] i'm seriously disappointed. i updated my laptop today, and despite the warnings, X still works! [01:09] i was expecting it to break! [01:14] chrisccoulson: You waited too long. If you'd done it 24 hours earlier there would have been a period when apt-get would want to remove xorg, ubuntu-desktop, that sort of thing :) [01:15] RAOF - ah, i shouldn't have waited then ;) [01:15] * micahg feels better for not upgrading till after X 1.9 transition :) [01:31] Where does one's personal gstetings data get stored, i.e in which subdir of the home dir? [01:37] gsettings even [01:38] .config ? [01:40] Looked there, but nothing looks obviously gsettings-ish. [01:41] Do we have anything that's using gsettings? Presumably the answer is yes. [01:42] My guess would be ~/.config/dconf [01:42] empathy does [01:42] That certainly contains a binary blob called “user” [01:44] RAOF: nothing there with that name. I just changed some settings in empathy and closed it. [01:54] TheMuso: Any idea why I appear to have two separate pulseaudio processes running? [01:54] RAOF: One for gdm? [01:54] I'm not running gdm. [01:54] Both are running as my user. [01:54] whaa [01:54] No idea. [01:55] RAOF: how are you checking? [01:55] htop. [01:55] There are two separate pulseaudio processes, with distinct PIDs and distinct nice values. [01:56] Are they both pulseaudio, or is one pulseaudio-gconf-helper? [01:56] i.e I have one pulseaudio, and run gconf-helper which is part of pulse. [01:56] Oh and seems gsettings is currently usig the memory backend. [01:56] has there been a regression that would make gnome keyring throw ioerrors again? [01:57] we're seeing that and are wondering about it :( [01:57] nessita: I know! let's switch to kwallet [01:58] No, they're both listed as /usr/bin/pulseaudio --start --log-target=syslog [01:58] RAOF: no idea then. [01:59] It could well be an artefact of the way I've worked around not being able to login via gdm. [01:59] possibly. [02:05] RAOF: press H [02:05] Sarvatt: To make gdm work? [02:06] to make it hide the other pulseaudio thread :) [02:07] Oh, no. It's not a separate thread. [02:08] (I can see the two separate threads of one of the processes separately) [02:08] TheMuso: the nvidia-current thats going in maverick is in https://edge.launchpad.net/~ubuntu-x-swat/+archive/x-updates it requires adding Option "IgnoreABI" "True" to the serverflags section in xorg.conf though to work with xserver 1.9 :( [02:09] but at least it works! [02:10] Yay. Failed upgrade due to libgirepository breakage/override issues. [02:10] * TheMuso updates his mirror [02:11] its fixed on any mirror? been broken all day here [02:11] Yeah, still broken here. [02:12] Well, unless it got fixed in the hour between me starting the update and the update actually finishing. [02:12] slightly less here now than before, only wants to remove epiphany-browser gnome-games libgirepository1.0-0 libseed0 lightsoff seed swell-foop [02:13] it was epiphany-browser gnome-games libgirepository1.0-0 libmutter-private0 libseed0 lightsoff mutter netbook-launcher seed swell-foop ubuntu-netbook unity before [02:13] heh [02:17] Oh, it didn't try to remove unity for me. [02:17] guessing that was just because of mutter [02:17] * ajmitch didn't have any problems after upgrading python-gobject [02:18] but it's not like I have much installed beyond the default desktop packages [02:18] python-gobject wants to remove all of that too [02:19] perhaps a few other things depend on libgirepository1.0-0, as it was removed & replaced with libgirepository1.0-1 for me [02:19] Alright. Let's play “kill pulseaudio before restarting” [02:19] Yeah. My problem is that both those packages own the same file, but obviously there is no conflicts/replaces etc to resolve that. [02:19] Or is there... [02:20] there's an exact match on 0.9.3-0ubuntu1 [02:20] Hrm there is conflicts/replaces. [02:20] so if you had an earlier libgirepository1.0-0 installed, it won't conflict/replace [02:21] Right [02:21] That may be it [02:21] most likely, should be an easy fix [02:22] and then the others (mutter, epiphany-browser) get to be rebuilt, yay [04:31] Is it known that dconf and dconf editor from dconf-tools crash horribly? [05:10] TheMuso, dconf problems? === awalton is now known as awalton__ [05:38] robert_ancell: to fix up gir-repository, is it as simple as changing the 1.0 references to 1.1 in the .gir files? [05:38] it's somewhat broken [05:53] robert_ancell: Trying to have a play with dconf/dconf-editor from dconf-tools and they crash. Not a biggy atm, just wanted a peak. [06:16] ajmitch, yes, I've uploaded that change [06:16] TheMuso, they work for me, are you fully upgraded? [06:16] ajmitch, (the same change was made in gir-repository git) [06:24] robert_ancell: Should be, as I said its no real bother atm. [06:25] TheMuso, that's weird though. what sort of errors are you getting? [06:25] TheMuso, if you run gcalctool, change the mode, and then restart does the mode stay the same? [06:26] robert_ancell: I am talking about the dconf and dconf-editor binaries from gconf-tools. [06:26] s/gconf-tools/dconf-tools/ [06:27] TheMuso, yes, is dconf working otherwise? [06:28] robert_ancell: Yes, empathy seems to be fine with it. [07:01] robert_ancell: sorry for bothering, I'm obviously too slow :) [07:02] ajmitch, np [07:02] :) [07:38] good morning [08:10] didrocks, hey [08:10] hey robert_ancell :) [08:10] didrocks, is mutter working well for you now? sorry about that [08:11] robert_ancell: can't test right now. I think it should. I'll have a proper test later. no worry ;) [08:12] didrocks, If I don't see seb128 before he starts, can you remind him to look at the libgirepository package he made - the two versions contain the same files [08:12] robert_ancell: sure, I will [08:24] RAOF: I see that a lot of people are complaining about the new X server and nvidia proprietary driver, known issue? [08:25] didrocks: Yeah; tselliot's got a new nvidia proprietary driver (at some advanced stage of) ready, which works against the new X server. [08:25] RAOF: there is no Break: against the old one? (the one not compatible with the new X server?) [08:26] RAOF: or are all people not reading the "partial upgrgade" thing? [08:26] The old one should be Depend-ing on something that the new X server doesn't provide. [08:27] We've been propagating some changes in the way Debian handles the dependencies, the goal of which is to make it easier for apt to upgrade in situations where there's a valid solution. [08:28] RAOF: hum, I can upgrade my X server there without any nvidia update… So, the provides: doesn't seem to work :/ [08:28] It used to offer the “solution” of removing all of X more frequently than it should be. [08:28] yes, but if I can upgrade normally and break my machine there… ;) [08:29] can I check something to see what's not going well and why it doesn't wait for the new driver? [08:30] It would appear to be a packaging bug in nvidia-current. [08:31] is it linked to Provides: xserver-xorg-video-7 ? [08:31] It *should* be Depend-ing on _something_ X related :) (preferably xserver-xorg-video-abi-8.0) [08:31] ok, I see no such depends :) [08:32] can you see that with tseliot? It's too late for now but for next time :) [08:32] I can't blame the user on the French forum for not reading the "partial upgrade" this time ;) [08:32] I just can say "it's an unstable release, do not rely on it" ;) [08:32] (and ok, I'm waiting before upgrading!) [08:33] You should be able to add "Option" "IgnoreABI" "true" to xorg.conf to make the nvidia driver work. [08:34] (At least, that's my understanding) [08:34] RAOF: yeah, people get that workaround on the forum [08:34] so, it should work [08:34] but I can wait :) [08:35] Do you want to break your graphical output? [Y/n] [08:35] Looks like fglrx is also a little messed up. [08:36] Depends: libxext6 (>= 0) [08:36] Hah. [08:36] hum :-) [08:36] That's a pretty easy dependency to satisfy :) [08:38] you never know… Did someone try foo -2.24? ;) [08:39] that can be… an interesting experience! [08:40] :) === oubiwann is now known as oubiwann-away === oubiwann-away is now known as oubiwann === almaisan-away is now known as al-maisan === al-maisan is now known as almaisan-away === almaisan-away is now known as al-maisan [08:59] hey [09:00] salut seb128 [09:00] lut didrocks [09:00] en forme? [09:00] ça va ;) OOo a planté à cause d'un symbol file pas à jour, donc je reprends le build ce matin avec dpkg-buildpackage -nc, ordi toujours à plat :) mais ça va! et toi? [09:01] lol [09:01] ca va [09:02] worked a bit late for those gir rebuilds [09:02] did that fix unity or not? [09:03] seb128: well, as I'm still in build mode, I can't easily switch sessions :/ I have to take my scratchbox to make some tests today === oubiwann is now known as oubiwann-away [09:03] didrocks, don't bother I will upgrade now to see how it goes [09:04] seb128, hey [09:04] seb128: thanks :) [09:04] robert_ancell, hello mr broke everything so the european guys have something to do during their day ;-) [09:04] seb128, :) [09:05] seb128, the package renaming you did broke upgrades, not sure what to do about that [09:05] robert_ancell, did it? how? [09:06] seb128, because both versions contain the same files [09:06] well I used a conflicts,replaces [09:06] seb128, weird, I upgraded my mini and it borked installing the -1 version [09:06] but it should conflict only with your buggy build [09:07] not with 0.6 [09:07] since the soname changed [09:07] the mini didn't have my build on it [09:07] what file conflicted? [09:07] the sonames changed from .0 to .1 [09:07] /usr/lib/libgirepository-1.0.so.1.0.0 [09:07] there should be no file conflict [09:07] you didn't have 0.9.3-0ubuntu1? [09:08] no === oubiwann-away is now known as oubiwann [09:09] keep an eye out for it anyway [09:09] robert_ancell, ok, weird, it was libgirepository-1.0.so.0.0.0 in 0.6 [09:09] can you check in /var/log/dpkg.log [09:09] what version was upgraded to what [09:09] ie grep libgir /var/log/dpkg.log [09:09] ok [09:10] booting it now.. [09:10] thanks [09:11] 0.6.14-1ubuntu1 [09:12] hum, ok [09:12] ajmitch had a similar problem - not sure if he got the buggy version first though [09:12] I will check [09:13] I might just conflicts on the previous version [09:13] you are right [09:14] the libeverything and the marshallingtest were already using .1 [09:14] doh [09:14] they like to make our lives hard! :) [09:14] hm? [09:14] oh, the conflicts/replaces [09:14] robert_ancell, I will take the debian change [09:14] breaks plox [09:14] :) [09:14] * Install Everything-1.0 and GIMarshallingTests-1.0 typelib in new [09:14] gir1.0-everything-1.0 package (Closes: #550478) [09:15] robert_ancell, ^ that will avoid the issue [09:15] seb128, heh, that Everything file likes to cause us problems - we spent ages trying to decide where it lived [09:16] I meant to backport the debian change when they did it [09:16] that would have avoided the conflict [09:17] seb128, oh, I was thinking - we're really screwed when gi does upgrades - all t he packages I looked at that install gir files have gir-1.0 hardcoded into them. So they better be happy with the current filenames. [09:17] s/filenames/binary format/ [09:18] looks like there is no need to file a bug about the libgirepository1.0-1 install issue [09:18] you people should use update-manager :p [09:18] it overwrite conflicts [09:19] robert_ancell, that's maybe why the directory didn't change with the abi version? [09:19] seb128, yes, because it is the individual packages that choose the install location. We might need to dh magic if they change it in the future [09:19] gtg, see you guys tomorrow [09:19] to use some dh magic [09:20] kiwinote: good morning! I just checked your deb-files branch, nice fixes. but it appears there is a test failing when I run test_database.py. - please re-merge and double check, but tests should be fixed [09:20] now [09:20] robert_ancell, have fun, see you tomorow [09:24] mvo: hi [09:24] mvo: is the test that is failing the self.assertEqual(len(db), 6) one? [09:25] kiwinote: I had the one about "expensive gem" failing [09:26] kiwinote: state was "NOT_FOUND" when it should be "NEEDS_PURCHASE" [09:33] mvo: thanks, the rest looks fine [09:34] kiwinote: so you have a new failure? or what do you mean with "the rest" ? [09:34] mvo: sorry, I meant that I had tested all the changes [09:37] kiwinote: ok :) thanks! [09:59] hey chrisccoulson [09:59] how are you? [09:59] hi seb128, i'm good thanks. how are you? [10:00] I'm fine thank you [10:03] seb128: did an upgrade this morning and now my login screen is pretty unresponsive [10:03] did you hear of any problems [10:03] ronoc, like no user listed? [10:03] ronoc, or...? [10:03] seb128: it basically show the name of the machine, doesn't get to listing users [10:04] the mouse works but i have to hard reset [10:04] known issue [10:04] seems to be a race [10:04] you can switch to a vt and sudo restart gdm [10:04] bug #614810 [10:04] Launchpad bug 614810 in gdm (Ubuntu) (and 1 other project) "gdm displays no user at startup in Ubuntu Maverick (affects: 5) (dups: 1) (heat: 34)" [Medium,Triaged] https://launchpad.net/bugs/614810 [10:05] ronoc, it doesn't happen every time [10:05] seb128: for me it is [10:05] seb128: I found a crash report for _usr_bin_do-release-upgrade [10:05] ronoc, that's a different bug [10:06] ronoc, did you try restarting gdm? [10:06] will try now [10:06] bbiab [10:15] mvo: sometime during the merges we started getting get_software > app_name > cat_name instead of get_s > cat_name > app_name. There is a fix for this in my branch. [10:16] heh, i'll actually be at UDS quite early this time, rather than turning up late on sunday evening [10:20] kiwinote: thanks! [10:28] ronoc, seb128: i solved this issue by fixing the permissions of /lib/dbus-1.0/dbus-daemon-launch-helper [10:29] -rwsr-xr-- 1 root messagebus 47536 2010-08-09 18:28 /lib/dbus-1.0/dbus-daemon-launch-helper [10:29] what issue? [10:30] the no user on gdm [10:32] Anyone free over here? [10:32] I need to get an empathy bug fix in [10:32] which bug? [10:33] cassidy: Yeah, Hi [10:33] cassidy: A regression bug due to the freenode default network one [10:33] cassidy: I know you commented on the bugzilla bug [10:33] ah ok this one [10:33] seb128: still no joy [10:34] cassidy: bug #615081 [10:34] Launchpad bug 615081 in empathy (Ubuntu) "log in button is not enabled when adding IRC account (affects: 1) (heat: 8)" [Medium,In progress] https://launchpad.net/bugs/615081 [10:34] I have the fix ^^ [10:34] ronoc: what are the permissions of /lib/dbus-1.0/dbus-daemon-launch-helper ? [10:35] bilalakhtar, could you submit it upstream please? [10:35] cassidy: ok [10:35] that's not an Ubuntu specific issue so there is no reason to fix it in Ubuntu rather than upstream [10:36] cassidy: Since it is caused due to the freenode bugfix. [10:36] alf__: -rwxr-xr-x [10:36] cassidy: ok, doing [10:36] alf__ for root [10:36] group and owner [10:36] cassidy: Have you accepted the DEFAULT_IRC_NETWORK patch? [10:36] ronoc: they should be -rwsr-xr-- 1 root messagebus 47536 2010-08-09 18:28 /lib/dbus-1.0/dbus-daemon-launch-helper [10:36] alf__, are you sure? [10:37] alf__: have you experienced the problem I am having [10:37] alf__, the issue started after a gdm update where dbus didn't change in months, it's a gdm race [10:37] bilalakhtar, yeah I merged it this morning [10:37] cassidy: Thanks a lot! [10:37] thanks for the patch :) [10:37] but I used gimpnet as default instead of freenode [10:38] basically get to gdm, but the login window is not displayed, the screen right before login is displayed [10:38] that still easier for you, you just have to patch one line [10:38] cassidy: Yes, I expected you to use GIMPnet, since that is preferred by GNOME [10:38] where the name of the machine is displayed with the computer large icon [10:38] seb128, ronoc: that fixed the problem for me, perhaps it is seperate to the race condition [10:38] alf__, you though it fixed it [10:38] alf__, you just maybe got lucky on next try [10:39] seb128: I can reproduce this every time [10:39] haven't been able to log in for an hour [10:39] ronoc, you are the very first to not have it racy [10:39] ronoc, install xdm or use startx or enable autologin [10:39] lucky me [10:40] okay will try that [10:40] seb128: Are you free to sponsor a rhythmbox regression? [10:40] bilalakhtar, not now but I will do some sponsoring today and it's on my list [10:40] seb128: i did notice a broken packages problem with gir and python [10:40] seb128: would this have anything to do with it ? [10:40] seb128: perhaps, although in /var/log/syslog i got complaints about "setuid helper has wrong permissions" [10:41] ronoc, that should be sorted in the next hour [10:41] grand [10:41] bbiab [10:42] cassidy: Shall submit later today [10:43] alf__, weird, I don't get that permission issue there [10:45] seb128: are your permissions -rwsr-xr-- 1 root messagebus ? [10:47] yes [10:48] seb128: for me (and apparently ronoc) it was -rwxr-xr-x root root [10:48] weird [10:48] did that start recently? [10:49] seb128: I just noticed it today after a dist upgrade [10:51] I merged the new dbus with debian this week but that should affect that [10:51] it's still fine on my machines and there was nothing about permissions in the debian dir diff [10:51] anyway the "no user listed" started with the gdm update before dbus changed [10:52] so it's a different issue [10:52] seb128: yes, it seems so [10:52] mvo, good morning, how are you going with integrating Ubuntu Single Sign-On? [10:53] mpt: you mean how its going? or what the plan is? [10:53] mvo, both :-) I know about [10:54] mpt: I played with the ubuntu-sso-client trunk yesterday, but it was not working for me, probably because of me not installing it properly [10:55] mpt: it looks straightforward though [10:59] ronoc, still no luck? [11:00] seb128: no joy still, couldn't install xdm because the repo is broken with the gir stuff, startx fails because x is running already, when I kill x it automatically restarts bringing me to the unresponsive screen [11:00] my only option is to return to the vt [11:00] ronoc, well you can sudo stop gdm [11:00] ronoc, or startx -- :1 [11:01] ronoc: did you try changing permissions and rebooting? [11:01] I didn't alf__ [11:01] just because i would rather not change permissions down /lib, but if it comes to it I will try it [11:01] ronoc: you should :) [11:02] ronoc: they are incorrect [11:02] do you guys are your lib dir on a different partition or something? [11:02] nope [11:02] no [11:02] it still doesn't make sense to me [11:03] the dbus.postinst set the permission [11:03] mvo, the reason I'm asking is that I'm wondering about the case of reinstalling purchases. When that brings up the dialog, the dialog should allow *only* sign-in, not registration, because registration is a dead-end for that feature. [11:03] ronoc, alf__: can you run [11:03] dpkg-statoverride --list /lib/dbus-1.0/dbus-daemon-launch-helper [11:04] seb128: will need to boot back into the maverick [11:05] seb128: I get no output [11:05] alf__, can you sh -x /var/lib/dpkg/info/dbus.postinst configure [11:05] mpt: good point [11:05] alf__, and tru to figure why it doesn't change the permission for you? [11:06] mpt: when this was discussed with and471 I pointed out to jonlea that for our use-case the login dialog would almost always be visibile (re-install) [11:07] mpt: iirc the ubuntu-sso one has a strong focus on registration [11:08] mvo, yeah, and that's fine if USC is remembering (or telling gnome-keyring to remember) the sign-in details, because 90% of the time you'll still want only registration. But, we still need the sign-in-only case. [11:08] mpt: I think its the right thing to have only one and doing it via dbus etc is the right way of doing it. but time is short, FF is very close [11:08] mpt: well, in USC we will never need registration currently as the agent is not capable (currently, there is a open bug about it) to get pre-exisiting credentials [11:09] mpt: you always have to login in the webkit view currently to purchase [11:09] seb128: going to try again now [11:09] mvo, is that being fixed for Maverick? [11:09] mpt: but yeah, once that is fixed/added the 90% figure works for USC as wlel [11:09] mpt: I hope, we can bring it up in the next meeting [11:09] mvo, does it make sense to have a flag in the API that says "show only the sign-in choice, not registration, please"? Then the SSO team can implement it later [11:10] I think it does [11:10] ok [11:10] I'll discuss that with Chipaca and nessita [11:10] but we need to coordindate with them, I assuem they are under pressure as well [11:10] thans [11:10] thanks [11:10] yeah [11:12] mvo, which of (1) apturl replacement, (2) gdebi replacement, (3) devildante's add-ons work, are on track for merging? [11:12] and (4) the OneConf stuff [11:12] seb128: sudo sh -x /var/lib/dpkg/info/dbus.postinst configure seems to run fine :S [11:13] alf__, does it change the permissions for the helper? [11:13] seb128: yes [11:13] alf__, dpkg -l dbus | grep dbus? [11:13] mpt: (1) and (2) are the same branch [11:13] mpt: oneconf is done [11:13] seb128: 1.2.24-3ubuntu1 [11:13] alf__, it's rather the state I'm interested in [11:14] ie is ie "ii" [11:14] or iU [11:14] mpt: its a plugin AFAIK didrocks does not need more from usc [11:14] or what? [11:14] seb128: iU [11:14] alf__, ah! [11:14] alf__, that's your issue then [11:14] alf__, sudo apt-get -f install [11:15] alf__, dbus is not configured, your upgrade probably broke [11:15] mvo: mpt: all is ok for oneconf, it's already in the archive, apart from the issue of syncing which is out of oneconf scope, should be ready [11:16] mvo: just that I will need a unregistering() function on USC side (if you uncheck the box "show other inventory"), so that it's taken into account without restarting USC, but I can do that (after my holidays, but before beta) [11:16] didrocks, cool. nzmm_ is working right now on a categorized list view for "Installed Software", and once he's got it done, you can reuse it for "items that aren't on this computer"/"other items" [11:16] didrocks: ok, that sounds like a bugfix [11:17] mpt: excellent! [11:17] mvo: right, that's why I'm not pushing too hard on that right now. I would prefer syncing working first :) [11:17] didrocks: yep :) [11:18] mvo: I just have a bug in the view for non app, which seems to be set to True even if I excplitely set "False" to the model. Don't have the time right now to have a deeper look [11:20] mpt: addons and deb-files are under review, I want them in, but I don't want to de-stablize too much [11:20] mvo, deb-files being kiwinote's work, or something else? [11:20] didrocks: if you have a test-case I can have a look [11:20] mpt: kiwinote work [11:20] ok [11:21] * mpt keeps forgetting that OneConf is a plug-in [11:21] mvo: I'll setup a mini view with it, just not that week (and on holidays then) :) Still, it's a bugfix IMHO [11:21] mvo: btw, I think you saw my dummy branch for featured app? [11:22] didrocks: yeah, will merge it today [11:22] mvo: thanks ;) [11:25] seb128: ara just pointed me to http://pastebin.ubuntu.com/476371/ - that is known, right? [11:26] mvo, yes [11:26] dpkg --force-overwrite -i /var/cache/apt/archives/libgirepository1.0-1_0.9.3-0ubuntu2_i386.deb [11:28] seb128: libgirepository1.0 is what broke my upgrade, too. Thanks! [11:28] np [11:30] seb128, thanks! [11:30]