[00:49] done [01:03] bryceh: xdiagnose is still a recommends so wouldn't it still be pulled in automatically? [01:04] jbicha, it was a recommends previously. We moved it to depends because it wasn't getting pulled in for people. Bug 814346 said to revert that change which is what I've done. [01:04] Launchpad bug 814346 in xorg "xorg-common depends on xdiagnose which pulls in all of gtk3" [Undecided,Fix released] https://launchpad.net/bugs/814346 [01:06] oh ok [01:07] I'm assuming it's not a problem, since we had it as recommends for a quite a while before this [01:12] I don't run pure kubuntu so I don't know [01:13] bryceh: Did you add xdiagnose to the actual seeds, or just to ubuntu-meta? [01:15] TheMuso, just to ubuntu-meta [01:15] TheMuso, how do you add it to the seeds? [01:15] bryceh: Right, it needs to be done at the seed level, otherwise that change will be lost next time the meta is updated. [01:16] TheMuso, ok, would you be willing to make that change? I've not dealt with the seeds before [01:16] bryceh: You need to branch lp:~ubuntu-core-dev/ubuntu-seeds/ubuntu.oneiric and add the package to the desktop file in that branch. If it needs to be a recommends, you need to surround the package name in brackets. [01:17] bryceh: Sure. I'll make the change, and you can branch the branch I referred to above and see what I did. [01:17] bryceh: So is it a recommends of ubuntu-desktop, or a depends? [01:18] TheMuso, I believe it should be a depends. We want it always to be installed [01:18] bryceh: Ok then. [01:21] bryceh: Ok, now if you branch lp:~ubuntu-core-dev/ubuntu-seeds/ubuntu.oneiric and look at the last revision, it should show you how a package gets added to the desktop seed. If you wanted to make it a recommends instead of a depends, you enclose it in brackets. [01:22] And now, once your ubuntu-meta upload is published, we will upload a new revision of ubuntu-meta, which uses an update script to automate the updating of the metapackages. [01:23] RAOF: You may be interested in this too should you ever need to add something to the seeds, that is if you haven't done so already. [01:24] I'm pretty sure I've touched the seeds before :) [01:25] TheMuso, cool thanks [01:27] bryceh: np [01:27] RAOF: Ok, just making sure. [01:41] hey bryceh and RAOF, why would I be getting on 2d right now instead of 3d? shouldn't nouveau be able to handle 3d? did something change recently? [01:41] I had been running nouveau for a while and switched to nvidia earlier this week [01:43] jasoncwarner_, are you sure it is falling back to 2d? are you using the unity-greeter? [01:43] it seems to default to the top session, not the last one [01:43] so if i forget to click it, i get unity-2d :) [01:46] jasoncwarner_, a common reason for getting 2d instead of 3d is a problem with glx libraries [01:46] I'm using unity-greeter, but I am explictly choosing my session...I choose 'ubuntu' and I get 2d [01:46] when I choose 'gnome', I get a message 'can't load session 'gnome'" [01:46] jasoncwarner_, did you file a bug report on this yet? If not please do so, so we can see the logs and see what's going on [01:46] jasoncwarner_, if you want to diagnose yourself, here's what I'd look at... [01:47] check glxinfo to see if you're booting with software rendering [01:47] look in /var/log/Xorg.0.log to see if the glx libraries are loading, and if not see if there's an error message [01:47] check dmesg to make sure the kernel modules are getting loaded properly [01:49] jasoncwarner_, I looked at the nvidia-graphics-drivers package and there hasn't been an update to it recently [01:50] https://launchpad.net/ubuntu/+source/nvidia-graphics-drivers shows 3 weeks since last change [01:50] jasoncwarner_, so next look in /var/log/dpkg.log to see what updated. [01:51] jasoncwarner_, I've also looked through the bug tracker for other recent reports of nvidia problems but didn't spot one... [01:52] so if you could file a bug report on this one it'd be a big help; even if it ends up not being an X issue, I'd still like to take a look. [01:53] bryceh: reporting now...I dont' see anything obvious in logs, but i'm also not hte best person to look at 'em. [02:01] bryceh: FYI - https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/814384 [02:01] Ubuntu bug 814384 in xorg "Unity won't load, nvidia won't install" [Undecided,New] [02:03] thanks, looking [02:04] jasoncwarner_, what error message do you get when you try installing nvidia? [02:04] jasoncwarner_: Fun! You've got the nvidia glx installed, breaking nouveau, but the nvidia kernel module isn't built for your current kernel. But is for -5 [02:04] one sec...rebooting...will get in a minute [02:05] bryceh: Didn't we attach dkms.log to these bug reports previously? [02:05] RAOF, DpkgLog.txt [02:06] RAOF, I'm examining it now :-) [02:06] No - the *dkms* log. Which might explain why nvidia isn't built for the -6 kernel. [02:06] http://paste.ubuntu.com/649621/ [02:06] bryceh: "sorry, installation of this driver failed. Please have a look at the log file for details: /var/log/jockey.log" [02:06] and then, in the log files...(one more sec) [02:07] RAOF, sorry (old man eyes) yeah it's at https://launchpadlibrarian.net/75721742/DpkgLog.txt [02:07] Someone's playing with gtk themes ;) [02:07] dah nevermind [02:07] :D [02:07] https://launchpadlibrarian.net/75721741/DkmsStatus.txt [02:07] there [02:08] dpkg and dkms are too similar :) [02:08] there should be another dkms log file hmm [02:08] bryceh: Yeah, that's the status. But dkms also writes a log file. [02:08] if it's not attached I'd guess it's not present [02:08] bryceh: jockey log [02:08] http://pastebin.ubuntu.com/649625/ [02:08] That's kinda annoying :) [02:09] /var/lib/dkms/*/*/build/make.log [02:09] jasoncwarner_, do you have any make.log files under /var/lib/dkms? [02:09] ooh jason found one bug at least [02:09] 2011-07-22 11:35:15,160 DEBUG: Could not instantiate Handler subclass __builtin__.NvidiaDriverBase from name NvidiaDriverBase [02:09] Traceback (most recent call last): [02:09] File "/usr/lib/python2.7/dist-packages/jockey/detection.py", line 957, in get_handlers [02:09] inst = obj(backend) [02:09] TypeError: __init__() takes exactly 3 arguments (2 given) [02:10] ah, and here's the second: [02:10] 2011-07-22 11:35:30,856 DEBUG: NVidia(nvidia_current).enabled(): target_alt /usr/lib/nvidia-current/ld.so.conf current_alt /usr/lib/nvidia-current/ld.so.conf other target alt /usr/lib/nvidia-current/alt_ld.so.conf other current alt /usr/lib/nvidia-current/alt_ld.so.conf [02:10] 2011-07-22 11:35:30,856 DEBUG: KMH enabled: False [02:10] 2011-07-22 11:35:34,096 WARNING: modinfo for module nvidia_current failed: ERROR: modinfo: could not find module nvidia_current [02:11] well, that says the nvidia kernel module isn't installed but doesn't say why [02:12] bryceh: no make.log fikes in /var/log/dkms, but I do have an 'nvidiaa-current' directory under there if that means anything [02:12] jasoncwarner_, it looks like the issue is that the nvidia kernel module didn't build. [02:13] likely because something changed with the -6 kernel [02:13] but we need a nvidia build log to know why it's not building [02:14] bryceh: I guess that is one problem, but another question I have is why isn't nouveau allowing me to run 3d when it was earlier this week? [02:14] right, and that's the second problem (or third, counting the jockey bug mentioned above) [02:15] bryceh: :) so many problems! [02:15] Well, for the boot in your logs the reason why nouveau wasn't allowing you to run 3d was that you had nvidia-current installed (but not usable) :) [02:15] The Xorg.0.log shows this: (==) Using system config directory "/usr/share/X11/xorg.conf.d" [02:15] FATAL: Module nvidia not found. [02:15] (EE) NVIDIA: Failed to load the NVIDIA kernel module. Please check your [02:15] (EE) NVIDIA: system's kernel log for additional error messages. [02:15] (EE) Failed to load module "nvidia" (module-specific error, 0) [02:15] (EE) Failed to load module "nv" (module does not exist, 0) [02:15] (EE) Failed to initialize GLX extension (Compatible NVIDIA X driver not found) [02:16] RAOF: should I just remove nvidia for now and revert to nouveau? [02:16] is X trying to load nvidia even though the kernel module is broken? [02:16] yep [02:16] jasoncwarner_: Yes. It's not possible for nouveau to work while nvidia is installed. [02:16] jasoncwarner_, yeah nouveau isn't loading because nvidia's still installed [02:17] so just sudo apt-get remove nvidia-current would suffice? [02:17] bryceh: X can't know the kernel module is broken until the nvidia DDX has failed to load ;) [02:17] [ 5.438] (EE) Failed to initialize GLX extension (Compatible NVIDIA X driver not found) [02:17] jasoncwarner_: Yup, that'd do it. [02:17] RAOF, right and by then it's too late [02:17] RAOF: doing it...will reboot in a minute [02:18] bryceh: Well, not quite. It falls back to nouveau quite happily. *But* because the alternatives are set for nvidia it's loading nvidia's libglx, which only works with the nvidia driver. [02:18] actually... is it too late? Couldn't X just note the missing kernel module and try harder? [02:18] ah, right the alternatives. dah [02:19] This is why Jason gets a display *at all* :) [02:19] so... next question is why is there no make.log [02:20] RAOF / bryceh : removing nvidia like that gets me back in to 3d..thank you! [02:30] jasoncwarner_, could you re-try installing nvidia once more and see if a make.log file gets generated? [02:31] yeah...give me a few minutes...talking to ted [02:46] grabbing tea... bbiab [03:26] bryceh: FYI...I just did a reinstall of nvidia and it worked...weirdly [03:29] jasoncwarner_, ok [03:29] jasoncwarner_, a lot of these kinds of nvidia loading bugs get resolved by uninstalling and reinstalling [03:30] it forces it to rebuild against the new kernel. In theory it should be doing that automatically already. [03:33] yo, bryceh and RAOF ... kernel question...i can't remember if we said anywhere if we were shooting for 3.0 in oneiric? I saw Linus just pushed 3.0 [03:34] jasoncwarner_, yes [03:34] easy enough ;) [03:34] jasoncwarner_, the kernel team's policy in oneiric is basically to follow upstream and ship what is stable, which in this case will be 3.0.x [03:35] bryceh: ok, very cool...thanks. [03:36] Unless 3.1 comes out remarkably quickly :) [03:41] Which I doubt, we are still in the rc stage. [03:44] TheMuso: Not as of 45 minutes ago :) http://www.omgubuntu.co.uk/2011/07/linux-kernel-3-0-released/ [03:45] humphreybc: Oh thanks for that, seems the latest linux kernel versions RSS from kernel.org is not yet updated. [03:46] =) [03:51] hi! I'm using 11.04. In the last month i keep being ask to enter a password before i am connected to the local network at home even though the laptop is set to automatically login on start-up anyone know why his started and how to fix? [03:55] You'd be seeing bug #744929 [03:55] Launchpad bug 744929 in gnome-keyring "After auto-login, prompted to unlock keyring multiple times" [Medium,Fix released] https://launchpad.net/bugs/744929 [03:59] Good morning [04:01] thanks [04:02] well, it needs to ask you at least once [04:02] if you see the question several times, that'd be above bug [04:07] Good morning pitti! [04:08] hey RAOF, how are you? [04:08] Looking forward to the weekend :) [04:09] Also frustrated at how bad my writing has become. There was once a time when I could have whipped up an EGL packaging howto in an hour or so, and been happy with it. [04:11] RAOF: writing C/Python has now become better than writing English? :-) [04:11] Pretty much :( [05:19] good morning [05:36] bonjour didrocks [05:36] guten morgen pitti [05:36] how are you? [05:36] I'm great, thanks! my CD build script is making progress [05:37] nice :) [05:58] * didrocks scratches his head to make a conditional qmake include for armel… [05:58] seems you can only specify unix/win/mac [05:58] ah, only platform, not architecture? [05:59] didrocks: could you perhaps add an armel check into the include file itself? [05:59] so that it's basically an empty file on !armel? [06:00] pitti: it's a whole plugin subdirectory that needs to be ignored on armel [06:00] ah :( [06:01] another way is to apply a patch to the upstream qmake file only for armel… [06:01] which I don't really like [06:06] hi .. didrocks [06:06] is this true : http://forum.parallels.com/showthread.php?p=451706#post451706 [06:08] ejat: not sure what "uniform location" means [06:09] didrocks: IIUC that's got to do with GLSL uniforms. [06:09] Which are… things. [06:09] oh, that would be a vertex question for jay then :) [06:09] * RAOF would need to look up glsl spec to accurately describe them. [06:11] thats a thing .. http://imageshack.us/photo/my-images/171/screenshot20110719at115.png/ [06:11] is it bcoz of the unity .. and if i try to login .. to normal desktop with effect .. its flickering .. [06:12] before this it doesnt happend like that [06:12] ejat: before what? last release? [06:12] which version is installed? [06:19] unity: [06:19] Installed: 3.8.16-0ubuntu1~natty1 [06:19] Candidate: 3.8.16-0ubuntu1~natty1 [06:19] compiz: [06:19] Installed: 1:0.9.4+bzr20110606-0ubuntu1~natty2 [06:19] Candidate: 1:0.9.4+bzr20110606-0ubuntu1~natty2 [06:21] how to debug / trace [06:24] didrocks: good morning [06:24] ara: good morning [06:24] morning desrt [06:24] hey ara, how are you? [06:24] desrt: heya [06:24] pitti: hihi :) [06:24] hey pitti, didrocks :) [06:25] ara: he's ignoring us [06:25] he has his makefiles [06:25] hey desrt! [06:25] ;) [06:25] hey ara :) [06:28] didrocks : or is there a way to reset/restore unity environment [06:29] ejat: I think for this gl stuff, it's better to directly see with our opengl master, jaytaoko on #ayatana (freenode) [06:29] ejat: he's on US time though [06:30] owh .. its related to ayatana ? [06:31] owh .. thanks . ill try to check with the master .. [06:31] ejat: #ayatana are were the unity devs are :) [06:32] owh really .. didnt notice about it .. ok ill check with them .. thanks … btw .. will compiz 0.9.5 will backported to natty ? [06:37] ejat: no, there are too many changes to backport it [06:37] and yw :) [06:38] yw? [06:40] ejat: you're welcome [06:40] ow sorry [06:41] :) [06:41] pitti: are you aware of a debian/.install.linux ? I can't find any hack in debian/rules and no documentation? [06:41] ejat: no worry :) [06:45] didrocks: never heard about that one [06:46] didrocks: it's not done on e. g. debian/kfreebsd? [06:46] pitti: no I see nothing like that in the debian/ directory === tkamppeter_ is now known as tkamppeter [06:57] mvo: word. [06:59] desrt: good morning! the only word I can currently say is "a cup of tea, fast^Wplease" [06:59] (ok, that was more than one word) [06:59] and many more outside of the quotes [06:59] heh :) [06:59] you should switch to coffee for the mornings. works faster. :) [07:00] hey mvo [07:00] maybe, but you know what they say about old dogs and new tricks [07:00] hey pitti [07:00] * desrt woke up this morning to didrocks saying 'coffee is ready in the kitchen'. score. [07:00] desrt: four words: coffe baah! tea yummy! [07:00] desrt: woah! didrocks \o/ [07:00] srsly. [07:00] desrt: ... a treat you never get at home? :-( [07:00] achievement \o/ [07:01] hey mvo ;) [07:01] pitti: often i get it at home as well, actually [07:02] good morning didrocks [07:31] didrocks: Check out man debhelper. [07:32] didrocks: dh_install will look for .install.$ARCH, .install.$OS, .install, and pick the first one that matches. [07:33] oh, in DEBHELPER CONFIG FILES [07:33] I just looked an man dh_install [07:33] RAOF: exactly what I wanted! excellent ;) [07:33] thanks a lot! [07:49] hey [07:49] salut seb128 [07:49] lut didrocks, ca va ? [07:50] seb128: ça va, et toi? [07:50] très bien ;-) [07:50] it's friday apparently [07:50] didrocks, can I upgrade my unity? ;-) [07:50] seb128: seems some people on the french forum dared to do so, and it seems to even work! :-) [07:50] \o/ [07:51] Ah, I see ctrl-tab works now! [07:51] indeed :-) [07:51] seb128: so, you read it like me, RAOF told it's working for him. So if it fails, it will be entirely his fault, isn't it? :) [07:52] indeed [07:52] And this is on my troublesome sandybridge laptop. It can't possibly fail to work for you! :) [07:52] the issue is that he will declare it's end of week and just ignore us :p [07:52] RAOF: heh, test-proof hw! :-) [07:52] hum, right… [07:53] I'm pretty sure we've previously established that X is perfect. Any bugs which may appear to be in X should actually be referred to the kernel team :P [07:55] RAOF: yeah, and every glitches in your keyboard or your screen is because you don't take enough care of your computer and should clean it more :-) [07:56] I'm glad we have dx [07:56] before their time people were just blaming all the desktop bugs on gtk :p [07:56] didrocks: Or possibly bake it again :) [07:56] RAOF: heh, more than true! :-) [07:56] seb128: indeed, it's a nice diversion ;-) [07:57] hey seb128 [07:57] hey pitti, how are? is your desk clean yet? ;-) [07:57] I'm still blaming GNOME, not DX, for making it so excruciatingly difficult to restore my keybinding shortcuts now [07:58] morning pitti. I was looking at the natty langpacks on the server last night, to build and test the Catalan one as you asked me, but I wasn't sure what the best way would be. Build all langpacks on the server without uploading them and then test the -ca ones locally? or just fetch the sources for the -ca ones and build and test them locally? What do you do in these cases when you're testing langpacks? [07:58] good morning everyone [07:58] seb128: getting there :) 16 mails in my ubuntu folder to go, and ubuntu-defaults-image builder is working now (writing manpage now) [07:58] hey chrisccoulson, how are you? [07:58] pitti, \o/ [07:58] pitti - oh, i thought everyone blames firefox for their problems :) [07:58] seb128 does ;) [07:58] I do blame firefox! [07:58] hi seb128, i'm good thanks. how are you? [07:58] dpm: I just rsync the source directories for the language I'm interested in, and build/install them locally on my desktop [07:58] heh :) [07:58] chrisccoulson, I'm great thanks [07:58] need some coffee though ;-) [07:58] ok, thanks pitti [07:59] seb128: but seriously, there is no way at all to set up keybindings any more? [07:59] pitti, ccsm [07:59] lol [07:59] for switching desktops, opening terminals, sending a window to the background, and all the other useful things [07:59] seb128: didn't find it in ccsm [07:59] the only place where I found it at all (and which works) is gconf-editor now [07:59] it's in the same categories that the number of ws [07:59] pitti: you have [general] and [gnome compat] [08:00] the tab before [08:00] seb128: gnome compat doesn't have useful stuff [08:00] TheMuso: Say I had a Dell 6420 here in front of me, and I found that almost the entire dynamic range of the volume control is covered by 0%~5%, would ubuntu-bug grab all the information that you needed straight up? [08:01] seb128: ah, the general tab has the focus stuff [08:01] seb128: erm, I mean the windowo handling stuff [08:01] seb128: but I didn't find focus-follows-mouse [08:01] pitti, viewport switched has the go to ws ones [08:01] that only was in gconf [08:01] you need to track options yeah [08:01] seb128: and switching to desktop like ctrl+alt+2 for ws2 isn't there either [08:02] it's in viewport switcher [08:02] yeah, annoying to go through [08:02] "go to specific viewport" tab [08:02] aah [08:02] I recommend doing a script that restore your standard gconf keys [08:02] if you often unity --reset [08:02] (and soon gsettings ;)) [08:02] I just did unity --reset yesterday [08:02] not often, just yesterday [08:03] I thought everything crashed around me, but turned out that it was "just" Qt going mad when opening calibre [08:03] pitti, and another question: I've been working with skaet to gather a set of contacts to test the localized images, and I've created this page for that: https://wiki.ubuntu.com/OneiricOcelot/LocalizedImagesRollout - I've tried to anticipate the questions the people creating and testing the images will have. There are some things that are still not clear to me and I was wondering if you or cjwatson could give me a hand. For example: (roughly) what will [08:03] in my debian sid chroot, calibre works fine, so I'm using that now [08:03] the workflow be to create a localized ISO? [08:03] dpm: I'm just starting to write documentation how to build localized images [08:04] currently typing the manpage for the scripts, and later I'll also write a wiki page about the process [08:05] pitti, oh, excellent. So basically, someone creating an ISO will have to install the scripts, run them, obtain the ISO and upload it somewhere, I guess? [08:06] dpm: right [08:06] dpm: well, and do the actual customizatinos to the configuration files of the defaults package [08:07] pitti, ah, I see. So will there be a defaults package for every language, or a master one with all of the configurations? [08:08] dpm: there will be an ubuntu-defaults-zh-cn in Ubuntu, as this is the one we'll "officially" build ourselves [08:08] dpm: others, like the French loco's, can/should stay in a PPA for now [08:09] pitti, oh, so a requirement for each team wanting to build a localized ISO will be for them to provide their own defaults package [08:09] dpm: correct [08:10] dpm: we have a script ubuntu-defaults-template which will create a package for you, with commented out examples and documentation in comments, etc. [08:10] morning [08:10] plus the manpages [08:10] hey rodrigo_ [08:10] hi pitti [08:12] great, thanks pitti. And just to make sure, the installer (d-i + ubiquity) will remain translated in any case, right? I.e. we're not removing their translations from the main ISO, so that people not aware of the LoCo versions can still do the installation process in their language as usual [08:12] dpm: right [08:12] cool, thanks [08:17] pitti: for the default applications at a system level, there is two solutions: either diverting /usr/share/applications/defaults.list or ship an additional file in /etc/xdg/. There is no way AFAIK to do that at the admin level (/etc) yet and so, I'll have to run that to the freedesktop ML [08:23] didrocks: sorry, I don't understand -- so additional file in /etc/xdg will work, or not? [08:23] didrocks: but diverting defaults.list seems ok for now [08:23] pitti: it won't for now, it will need to be discussed in freedesktop. I think there is still a value for admins though [08:23] but ok, will divert for now [08:23] right [08:24] didrocks: I just finished typing the manpage, and my last image build test is about to finish [08:24] didrocks: I was going to upload 0.9 after that, but can wait a bit if you want to commit this first [08:25] pitti: excellent! I'll need to go to the train station soon, should take an hour, but then, I'll work on that, depends if you want do to an early release [08:27] didrocks: I guess I'll just upload it then; feel free to do a 0.10 then [08:27] didrocks: we aren't short on version numbers, after all :) [08:29] pitti: heh :-) [08:30] pitti, what's the name of the script to do the actual localized image build? [08:30] dpm: ubuntu-defaults-image [08:30] great, thanks [08:30] dpm: in the ubuntu-defaults-builder package (not in oneiric yet, but will be in an hour) [08:31] pitti, sounds good, thanks [08:31] didrocks: FYI, support for a hook/chroot script exists and works now, so you can do the additional mangling ther [08:32] e [08:32] pitti: oh, excellent, yeah, I'll certainly use that [08:32] didrocks: I don't have a hook/image script yet (for mangling the iso parts outside of the squashfs chroot), but adding that should be easy [08:33] uploaded 0.9 [08:34] in other news, thanks compiz or unity for finally fixing the "blurry text in panel" bug [08:34] ? [08:35] indeed [08:35] RAOF: reading a window title or menu or indicator title now doesn't make my eyes bleed any more [08:35] Hm. Maybe I'm an outlier with my Hinting:none ways. [08:35] pitti: what do you think about renaming unity/ directory in the template to defaults/ and have defaults/default-application.txt and defaults/launchers.txt ? [08:35] it now has a rather 1980-style black/white "engraved" effect, but this now looks deliberate (and thus fixable) [08:36] didrocks: I'd use desktop/default-applications.txt [08:36] didrocks: and keep unity/, as the launchers are really unity specific; they don't apply to shell or gnome-panel [08:36] the new switcher looks nice ;-) [08:37] pitti: ok, let's do that then. (I was thinking desktop would only be for .desktop files but ok :)) [08:37] seb128: think i could get a one-time postinst 'killall dconf-service'? [08:37] How do you switch to a window, rather than an application, with the new switcher, though? [08:37] you don't yet [08:37] RAOF: not possible [08:37] Which is why it's hidden away under ctrl-tab, I take it? ☺ [08:38] desrt, we could do that in the postinst, the service would respawn and the session keep working right? [08:38] correct. [08:38] RAOF: that and the fact that you can't easily transitionate people to a new plugin [08:38] it's fine as long as there is no write in progress at the time [08:38] which is likely [08:38] desrt, what happens if we don't restart the service? [08:38] it's not really critical to be honest -- i'll think about it more [08:39] seb128: the dconf you ship now doesn't support storing maybe types in the database [08:39] didrocks: Boo. Incidentally, I like “transitionate”. It's got a nice sound to it ;) [08:39] due to dbus limitations [08:39] RAOF: heh [08:39] i added a workaround on the client and service side to support this [08:39] desrt, what we usually do is teach to the service to do whatever is needed, like reloading, on sighup and send a sighup [08:39] but the updated client library doesn't make any difference unless the service is also restarted [08:39] the upshot is that you will have strange values appear in the dconf database... [08:39] desrt, that avoid catching it at the wrong time while it's doing something [08:39] not corrupt -- just strange. GSettings will correctly ignore them. [08:39] didrocks: Generally in English we'd use ‘transition’ as a verb, but ‘transitionate’ sounds much cooler! [08:39] seb128: ya. was just saying this to didrocks [08:40] it's actually an impossible problem with dbus [08:40] services cannot safely exit without the possibility of a race [08:40] old old old bug [08:40] right, argument for years with debian as well [08:40] " [08:40] RAOF: time to update the english dict :) [08:40] restart the dbus service on dbus package update [08:40] oh. that's an even older bug [08:40] i remember talking about that in the basement of a hotel in montreal :) [08:41] this is a different bug [08:41] ok [08:41] there is no way for a dbus service to safely release a name and process any remaining incoming requests for that name without risking a new copy of the service being activated and processing new requests at the same time [08:41] well anyway if you do an update where the service should be restarted please state it in the NEWS [08:42] so you have a choice: risk the concurrency, or risk lost messages [08:42] seb128: sure. [08:42] chrisccoulson, you didn't do the evo-couchdb splitting we talked about, right? [08:42] rodrigo_, hey [08:42] rodrigo_, new gsd today? ;-) [08:43] seb128, yes, probably [08:48] didrocks, seb128, if you know anyone who'd like to volunteer for creating and testing the French image, do you think you could ask them to add themselves to https://wiki.ubuntu.com/OneiricOcelot/LocalizedImagesRollout ? [08:52] desrt, thanks for the glib email [08:54] dpm: I'll do it as I was already in charge of our hackish respin :) [08:55] didrocks, excellent, thanks ;) [08:55] yw ;) [08:57] rodrigo_, not yet [08:57] chrisccoulson, ok, I'll do it then [08:57] rodrigo_, cool, thanks! [09:00] seb128: no prob. [09:00] mvo, dunno if that's you or gtk3 which fixed it by now typing names in the update-manager list works just great without having the cursor jumping backward \o/ [09:00] seb128: nothing you didn't already know. more for the benefits of the others, i guess [09:00] desrt, indeed [09:00] desrt, oh, and stop rolling new glib tarballs before I've time to properly test the previous ones :p [09:00] seb128: :) [09:01] seb128: it was necessary, actually [09:01] you can never upload .12 [09:01] because the gtk that doesn't crash with it already depends on .14 :) [09:01] well i did backport the commit to gtk yesterday [09:01] ah [09:01] okay then :) [09:02] would anyone else be up for creating and testing the localized ISO for their language? Just add yourselves to the table on https://wiki.ubuntu.com/OneiricOcelot/LocalizedImagesRollout ! [09:02] desrt, btw I dropped the eventfd revert and the build failed after 3 hours on amd64 again [09:02] damnit [09:02] and i guess it didn't apply cleanly anymore [09:02] desrt, blocked in some of the gdbus tests [09:02] desrt, I didn't try, you asked me to drop it [09:02] let me check [09:03] i'm fairly sure it will break against gcancellable [09:03] it's probably reasonably easy for you to make it apply again [09:03] but if not i can do it on the train [09:04] desrt, indeed, one hunks failing, but should be easy to update [09:05] desrt, I will try to upload .14 to the ppa without the revert still to have another build round just to see how often it blocks still [09:05] seb128: i'd prefer you didn't do that, in fact [09:05] well, actually, PPA is not so bad [09:06] desrt, why? [09:06] just don't release the eventfd stuff into the distro [09:06] ok [09:06] because if it builds, it will only be by chance [09:06] and then everything will be fucked up again [09:06] i'm going to need to get a hardy VM with a pbuilder to trace the problem [09:06] don't expect that to happen before next week [09:06] until then, just carry the patch [09:07] desrt, ok [09:08] gonna go catch a train now. laters. [09:08] desrt, see you [09:09] ensuring the canadian is not lost to catch his train, bbiab :) [09:09] ;-) [09:25] mpt - when thunderbird adds folders with the new messages to the message menu, should they appear before or after the shortcut entries (ie, compose/contacts)? or does it not matter? [09:25] i'm currently fixing a bug where it gets them all mixed up, and wondering where the entries should actually go [09:27] chrisccoulson, spec suggests application item, then shortcut items, then message sources (e.g. folders) [09:27] but I see it's unclear [09:27] mpt - ok, i can make it do that [09:28] they currently end up before and after the shortcut entries [09:29] thanks chrisccoulson. I tweaked the spec to make it clearer: https://wiki.ubuntu.com/MessagingMenu?action=diff&rev2=78&rev1=77 [09:30] The "Evolution" section in the wireframe also shows that order [09:30] mpt - thanks [09:34] mpt - ok, this seems to fix it - https://code.launchpad.net/~chrisccoulson/indicator-messages/fix-menu-ordering/+merge/68811 [10:27] rebooting, brb === MacSlow is now known as MacSlow|lunch [10:52] hm, Alt+F10 not working any more, is that a bug or feature? [10:53] ah, seems it only stopped working for un-maximizing; that works with alt+f5 now [10:53] (if only I could tell, the title bar menu is gone) [10:54] njpatel: not sure about that one, didn't see any commit which was willing doing that AFAIK ^ [10:55] it could also have been a custom setting of mine, not sure [10:55] (which went away with unity --reset) [10:55] weird [10:55] F10 isn't working either [10:55] wtf [10:55] it just seems to be in my muscle memory to maximize and unmaximize with Alt-F10 [10:55] njpatel: that might be the same problem as with the Win+Number shortcut? [10:55] pitti, I think so [10:56] curiously Windows alone works to trigger the dash, just not Win+something [10:56] so compiz, not unity if so [10:56] alt+f1 and alt+f2 work fine [10:56] yes, it stopped working before yesterday's compiz update [10:56] wth, middle click on launcher icons stopped working :-( [10:56] can't open a second firefox on another workspace? [10:57] pitti, didrocks if I change "show first menu" to Ctrl+Alt+F, then it works, but none of the F-keys work alone from what I can tell [10:57] seb128, yes, we're going through an event sub-system change at the moment, bear with us, will be fixed for next week's release promise [10:58] ah, so "known" [10:58] njpatel: that's ok, as long as it's (sort of) expected breakage [10:58] thanks [10:58] pitti, okay, did some tests [10:59] pitti, didrocks Alt+F10 works for me for maximising [10:59] but F10 alone doens't work to show menu [10:59] njpatel: yes, but not for unmaximizing again [10:59] confirming F10 bug [10:59] right, not for unmaximising [10:59] indeed [11:00] oh, and now f10 works :S [11:00] wtf [11:00] that worked until yesterday; it could be that this was a custom setting of mine [11:00] huh ,that's werid [11:00] weird* [11:00] didrocks, pitti can you confirm: a window that has global menu support: F10 woks, but a window that doesn't (say CCSM): it doens't work? [11:01] njpatel: indeed, you seem to be right [11:01] njpatel: empathy has global menu support, doesn't it? F10 doesn't work there [11:02] confirm that it doesn't work in ccsm [11:02] doesn't work in firefox or empathy either, though [11:04] pitti, works for me in empathy [11:19] mvo: ping [11:20] mvo: transaction role that corresponds to 'an upgrade/update in progress' == 'role-upgrade-system' ? [11:22] hello ronoc [11:22] hey, mvo [11:23] ronoc: hold on a second, that is the one, there is another one [11:23] cool [11:23] oh, i can't make the dash go fullscreen now [11:23] is that known? [11:23] ronoc: role-commit-packages [11:23] mvo, perfect thanks, also what software were you referring to you in your email about needed a certain version before ican start querying restart required state [11:24] ronoc: ups, sorry. the package is ubuntu-system-service [11:24] ronoc: just add it to the recommends or depends of your package :) [11:24] mvo, I will do [11:24] ronoc: but its uploaded in oneirirc, so it should be fine for most people [11:26] mvo, should have a release out today with the beginnings of the apt communication. I use my own transaction with upgradeSystem call at start up and when I see an ActiveTransactionChanged signal to determine the state of the upgrade system [11:26] ronoc: sounds good, so the python dbus stuff I send you helped with getting that? [11:27] ronoc, what are you doing with package management? [11:27] mvo, definitely just needed to rewrite the callback in C [11:27] seb128, for the new session indicator, needs to be smarter around upgrades/updates [11:27] ronoc, oh ok, nice ;-) [11:27] mvo, thanks for the help [11:27] yw [11:28] seb128, hopefully :~| [11:28] njpatel, can I add the "you should fix the indicator-session user switching menu" to the next week list? ;-) [11:29] seb128, thats fixed [11:29] njpatel, I want my guest session back! [11:29] ronoc, where? [11:29] seb128, i will release it today [11:29] ronoc, didn't you say it was an unity bug? I have the new unity [11:29] i need to merge in njpatel's branch later before i merge [11:29] oh ok [11:29] njpatel, ronoc: thanks ;-) [11:30] i just want to get more apt progress updates into the menu before releasing [11:30] seb128, see, we do fix bugs! [11:30] hum, alt-f2 is broken in some cases as well [11:30] like you can't click on icons [11:30] seb128, yep, there is something funky with keybindings right now [11:30] ronoc, yeah, no hurry [11:30] njpatel, well in this case it's "hit a key and delete the char you typed" [11:30] njpatel, it leads to "can't click on any of the history icons" [11:31] njpatel, is it worth opening bugs or will the issue just be solved because they are known anyway? [11:32] seb128, known [11:33] event stuff, if it's not fixed by next week then feel free to open a bug [11:33] ok [11:33] njpatel, were the count emblem on the launcher supposed to be moved to the top right corner? [11:33] seb128, yep [11:33] ok ;-) [11:34] mvo, [11:34] UpdateManager.py", line 1157, in initCache [11:34] hello [11:34] dialog.vbox.set_spacing(6) [11:34] AttributeError: 'MessageDialog' object has no attribute 'vbox' [11:34] mvo, do you know if that's known? [11:34] hey miber45 [11:35] is it the right place to ask a question about screen resolution with ubuntu natty? [11:36] seb128: yeah, the incredible mterry fixed it already :) [11:36] mvo, ok, seems it's bug #813778 [11:36] Launchpad bug 813778 in update-manager "update-manager crashed with AttributeError in initCache(): 'MessageDialog' object has no attribute 'vbox'" [Undecided,New] https://launchpad.net/bugs/813778 [11:36] seb128: are you using the lastest version? [11:36] mvo, closing the bug [11:36] mvo, no, I didn't update for 2 days [11:37] mvo, updating now ;-) [11:38] miber45, the correct place for support is #ubuntu. this channel is for developers, but feel free to hang around [11:38] hi all [11:38] bah, I give such dialogs several times a day [11:38] ok thanks [11:38] "GConf error: Configuration server couldn't be contacted: D-BUS error: Method "Set" with signature "s(ib)" on interface "org.gnome.GConf.Database" doesn't exist" [11:39] so, the latest version of Unity as received in 11.04 latest update is in one word: excellent and breath taking. [11:39] can it be that the resolution improves as well? new drivers for intel accl. hw? [11:39] and it is better than Mac in several orders of magnitutde [11:40] also quite amazing how it suits at the same time, both mouse interface and touch interface on my ideapad [11:40] e.g. no change required (since mouse cursor movements are interpreted as gestures anyways) [11:41] and firefox became faster than chrome, so since 11.04 last update, I stopped using chrome :) [11:41] * sivang notices this is FF5 [11:43] * sivang thanks the desktop team and bows. [12:26] njpatel, the window close button in the panel on my secondary monitor doesn't work (it opens the dash on the primary monitor rather than closing the window on the secondary) [12:26] do you know about that already? [12:26] or should i report a bug? [12:28] chrisccoulson, yep, we know about it [12:28] chrisccoulson, will fix for next week [12:28] njpatel, cool, thanks [12:28] * rodrigo_ -> lunch [12:38] pitti: when you get a chance: https://code.launchpad.net/~didrocks/ubuntu/oneiric/ubuntu-defaults-builder/add_default_applications/+merge/68829 (please, bear with my perl) === MacSlow|lunch is now known as MacSlow [12:57] mvo: ping [12:57] ronoc: pong [12:57] mvo, i dont see transaction finished signal being sent ? [12:58] I get the propertyChanged signal but not the transaction finished [12:58] ronoc: thats ok, its just running in Simulate() mode [12:59] ronoc: or do you mean that when a real upgrade is running you don't get it? [12:59] mvo, well I use simulate mode to determine the state of the system all the time [12:59] i suppose i should not do that ? [13:00] ronoc: no sure, what do you mean with "all the time"? [13:01] mvo, when i get the activetransactions changed signal, I kick start an upgradesystem which then gives me my own transaction on which i then call simulate [13:02] I suppose I need to use the id given in the activetransactions changed signal to see how that 'real' transaction is doing [13:03] ronoc: right, just ignore the simluation one [13:04] mvo, ok simulation then just for start up to see if we have updates ? [13:04] pitti, seb128: is there a list of all the packages covered by upload privs for ~ubuntu-desktop membership? [13:08] didrocks: looking [13:09] dobey: yes, there is [13:09] edit_acl.py -S oneiric -P ubuntu-desktop query -> http://paste.ubuntu.com/649949/ [13:10] pitti: i found a components page on the wiki, but it didn't seem to be up to date [13:10] dobey: edit_acl.py is in lp:ubuntu-archive-tools [13:10] ah ok [13:11] pitti: hrmm, i don't see some things in that pastebin that should probably be there [13:12] dobey, like? [13:13] seb128: ubuntu-sso-client, ubuntuone-storage-protocol, ubuntuone-client [13:14] seb128: and bindwood and configglue [13:15] dobey, the list is incomplete indeed, we usually fix it when we hit issues [13:15] well when I say "we" it's people who have desktop set only upload rights [13:15] but I guess we don't do a lot of u1 uploads so nobody hit those [13:15] right [13:16] can easily be fixed though ;-) [13:16] somebody needs to drop an email to cjwatson with a list of things that need to be added to the set [13:16] i was looking for a list to see what all u1 stuff was there, for when i apply for ~ubuntu-desktop :) [13:17] well it could make sense to do an u1 set [13:17] nessita or you would probably easily be accepted for an u1 set [13:17] not sure for desktop since you don't do a lot of desktop updates out of u1 [13:18] i guess. i was going to apply for motu too, but that doesn't help with our stuff that is in main, really [13:19] you already have ppu for ubuntuone-client right? [13:19] yeah [13:19] you could request some extra source to be added but I think in practice it would make sense to have an u1 set [13:20] i don't disagree with that [13:20] what's the bset way to get a new set created? [13:21] pitti, ^ is that a tb thing? === m_conley_away is now known as m_conley [13:23] chrisccoulson, why did you add an about:blank line to the "new window" firefox,unity command? [13:23] it does "firefox -new-window about:blank" [13:24] the about:blank part is annoying :p [13:24] seb128: I see liferea on that list but it doesn't seem to be installed, although I'm just running updates to double check [13:24] what list? [13:25] seb128, i'm not sure if there's a reason for that [13:25] kenvandine, hey, still no gwibber love with the new version :-( [13:25] didrocks: MP updated, there's some small things to fix, but looks good otherwise! [13:25] seb128, still acting weird? [13:26] seb128: ignore me mis-read the purpose of the list it's the acl not what the desktop meta pulls in my bag [13:26] seb128, can you describe the behavior more? [13:26] dobey, seb128: yes, creation of a new package set starts with mailing TB [13:26] kenvandine, yeah, for one thing the indicator-message count is to 79 and if I open gwibber I will get like 3 new entries in the list [13:26] davmor2: right, it's a list of privs, not a list of stuff on the CD :) [13:26] ups [13:26] pitti: ok. did you see my question re: apport yesterday? [13:26] wrong focus [13:27] seb128, i [13:27] kenvandine, so count is wrong, then I still get the flicker behaviour [13:27] seb128, i'll look into the message count thing [13:27] kenvandine, I will make you a screencast [13:27] dobey: I don't think I did [13:27] seb128, http://bazaar.launchpad.net/~mozillateam/firefox/firefox-trunk.head/revision/932 ;) [13:27] kenvandine, how does it get the count? is there any way I can query that count manually for debug purpose? [13:27] dobey: yeah I just saw the ubuntu-desktop query bit not the acl bit at the beginning :) [13:27] dobey: I got several X crashes yesterday, probably it was caught by that then [13:27] chrisccoulson, thanks! [13:27] pitti: was wondering if there's a way to do apport magic for plug-ins. like if nautilus crashes, will it go through all the installed plug-ins and run their apport collection script as well? [13:27] seb128, not really... it determines the number of new rows inserted in the db [13:27] pitti: looking :) [13:27] chrisccoulson, I never noticed because usually I middle click but the unity guys broke that [13:27] for specific streams [13:28] then adds them to the previous count [13:28] dobey: yes, that's a matter of writing a nautilus package hook [13:28] and focusing the stream in the client should 0 that count [13:28] kenvandine, well is that number a "items since you opened the ui" then or, or a "history count"? [13:29] "unseen" [13:29] ok, so that's wrong [13:29] since you last focused the stream [13:29] or does it count all the "love that" and replies as items? [13:29] pitti: hrmm, so i have to add special stuff to my apport script? [13:29] seb128, the "love that" would be included [13:30] pitti: is there a good example somewhere? :) [13:30] and actually for facebook, replies would too [13:30] since those aren't separate [13:30] kenvandine, that could explain part of the count but that still seems high [13:30] and i could probably make that smarter [13:30] facebook kind of skews the logic [13:31] since they update the original post and set a new timestamp [13:31] dobey: you need to figure out which other package you want to collect data for, and call apport.hookutils.attach_related_packages(report, [pkg1, pkg2, ...]) [13:31] dobey: /usr/share/apport/package-hooks/ has some examples for this function [13:32] ok [13:36] didrocks, around [13:36] since today i cant log into oneirick anymore [13:36] i had to install kubuntu [13:37] kenvandine, http://ubuntuone.com/p/15p4/ [13:38] seb128, interesting... it's like it can't decide if it should show or hide the last tile [13:39] jasoncwarner_ has the opposite problem, it never quites fills up the view [13:40] kenvandine: that is the issue I see too there are normally a couple of tiles missing to fill up the whole page until you scroll [13:40] didrocks, how do i enable the unity-greeter per default [13:40] ? [13:40] pitti: new proposal pushed and tests still run, thanks for the hint on next . I was looking for a "next 2" (like continue 2 in python), but didn't find anything :) [13:40] seif: hey [13:40] seb128, so the flicker starts when you begin to scroll right? [13:40] kenvandine, ok, so do you want bug report about any of the issues (count, flicker, empty bottom)? [13:40] seif: ok, first not logging, is there the greeter not showing? [13:40] or you end on a blank session? [13:40] kenvandine, it does flicker when I click on the page down thumb [13:40] the greeter shows [13:41] seb128, let me try scrolling that way... [13:41] but as soon as i click a user it goes away and comes back and asks me to pick a user [13:41] kenvandine, it does flicker a few times and then scroll and stop flickering [13:42] ah! [13:42] i reproduced [13:42] davmor2, is that one of the bugs you filed? [13:42] seif: the crasher is known [13:42] kenvandine: indeed [13:42] kenvandine, ;-) [13:42] seif: /etc/lightdm/lightdm.conf for choosing the greeter [13:42] seb128, i always scroll with my wheel... it looks like moving one tile at a time goes a little nuts [13:43] seb128, file the other 2 bugs please [13:43] didrocks, so i should stay with kdm [13:43] ? [13:43] kenvandine, ok [13:43] thx [13:43] seif, didrocks: what is not working on the login? [13:43] seb128, what's weird is keyboard nav doesn't do it [13:43] insertin gpassword [13:43] as soon as i choose a user [13:44] it restarts [13:44] in that case the gdm [13:44] kenvandine, well I usually scroll with the wheel as well, I just tried the scrollbar thing because I had issues with the wheel at first when the list was almost empty [13:44] seif, do you use a protected user directory? [13:44] seb128, yes sir [13:44] but it work with kdm [13:45] seif, ok, known bug, you can click on "other" and type your username and then password [13:45] seb128, ok will try [13:45] seif, it has an issue with protected user dirs, it tries to read the .dmrc for the user you select and hit a bug [13:45] seif, others should workaround it [13:45] it's easier than changing the greeter [13:45] seb128, ok cool [13:46] pitti: hrmm, so the source_nautilus.py there would need to do the add_related_packages() thing, and that call would result in the source_$package.py being run for each package, right? [13:46] seb128, what is the path to the unity-greeter [13:46] ? [13:46] dobey: that's the idea, yes [13:46] seif, dpkg -L unity-greeter [13:46] if you have it installed [13:46] i do [13:47] pitti: ok, it doesn't look like nautilus is doing that in natty at least. was that added in oneiric? [13:47] dobey, doing what? [13:48] desrt, wb [13:48] seb128: attach_related_packages() for all the installed extensions [13:48] seb128, and how do i set unity-greeter for lightdm [13:48] although the pydoc suggests it doesn't run the hooks [13:49] "In the future, this might also run their hooks." [13:49] dobey, it never did, we just reassign using report.add_package_info [13:49] seb128: thanks :) [13:49] desrt, question for you [13:49] seb128: very long lunch break [13:49] desrt, rather than reverting the eventfb patches, would force the eventfd to 0 in the configure.ac work? [13:49] eventfd [13:49] seb128: it's possible. [13:50] seb128: but i think the problem actually lies with the fallback case [13:50] desrt, looking at the ifdef in the code it seems it would to me [13:50] since we can't get an eventfd on the builders, the fallback code runs [13:50] seb128: it's quite possible. give it a shot. [13:50] desrt, well I did upload with it force to off and it built, but that's one build on one arch so it might just be luck [13:50] I'm waiting for i386 to build [13:50] the ppa builders are a bit busy today [13:51] should have had pitti do the upload :) [13:51] seb128: if it works, it's another helpful hint [13:51] well he can bump the score of other people upload if asked ;-) [13:51] desrt, it seems to work [13:51] seb128: pitti was just suggesting it did [13:51] desrt, I will keep you updated [13:51] dobey, well I'm telling you it didn't, I wrote that hook, it just use report.add_package_info() to reassign [13:52] well not sure if report.add_package_info() is supposed to get you the infos for the package you reassign to [13:52] but that's the call we use [13:52] seb128: it doesn't even do that [13:52] report['RelatedPackageVersions'] = package_versions(*packages) [13:53] what? [13:53] nautilus? [13:53] no, attach_related_packages() [13:53] oh ok [13:53] i know what nautilus does [13:54] it just does the same thing, by looping over the .so files in the libdir and getting what packages own those files, and their versions, and attaching them in an attachment :) [13:54] it doesn't reassign anything [13:54] didrocks: thanks! three more small fixes, then this is gold [13:55] pitti: looking :) [13:56] meh, v4l-utils FTBFSed again, can't just drop the build dep [13:56] dobey, that's not true, calling report.add_package_info() leads to the bug to be filed against the package specified in that call [13:56] worked locally because I have ia32-libs installed [13:56] we need to actually drop lib32v4l* stuff, drop the dependency from ia32-libs, and add the i386 deb there [13:56] dobey, that's why we did it, to stop receiving nautilus bug when it was segfaulting in the u1 code :p [13:56] * pitti leaves that for someone else to worry about, though; out of time [13:58] seb128: well, that only happens if there is a stack trace, and it has that info, it seems. :) [13:58] dobey, right, which in practice work most of the time [13:58] dobey, it just need to have the filename, not the symbols [13:58] which in practice is always right if you dump is valid [13:59] your [13:59] it doesn't even rely on the retracer [13:59] seb128, did not work [13:59] when i choose others it also restart [14:00] well, it could still be wrong, since it stops at the first occurance. but ok [14:00] lightdm does not even start [14:00] seif, ol, weird [14:00] dobey, well in practice I've not seen cases where it was wrong [14:00] dobey, which is "good enough" for me over several cycles we had the hack [14:01] well, we don't crash it any more, so it's all good :) [14:01] ;-) [14:03] my brain is lost now in the apport code though. i can't tell if that is running the report for the package passed in add_package_info() [14:07] argh, no rodrigo any more [14:07] desrt: I am testing the list support in dconf-qt [14:07] seb128: do you happen to know what's left for bug 787694? the control-center task is still open [14:07] Launchpad bug 787694 in ubuntuone-control-panel/trunk "GNOME 3 System Settings: Provide stub icons for launching Ubuntu specific capplets" [Medium,Fix committed] https://launchpad.net/bugs/787694 [14:08] desrt: I just tried and got a crash, I am now doing a quick debug [14:08] seb128: argh, ignore me [14:08] pitti, ok ;-) [14:08] seb128: u1-control-panel != g-control-center [14:08] TGIF/holiday etc. [14:09] pitti: hrmm, we might just need to make a release of u1-control-panel for that [14:10] dobey: yes, that's fine; I wrote that into my status now [14:10] pitti: oh interesting, it seems that lintian is looking at debian/copyright Maintainer field to tell that I'm trying to do an NMU :) [14:10] so that's by and large "done" [14:10] didrocks: yeah, it always complains about that one; just ignore it [14:10] pitti: yeah, rodrigo's fix for it is in our trunk already, i remember reviewing it :)( [14:10] Kaleo: cheers. [14:10] Kaleo: i never tested it :) [14:10] err just the ) there [14:11] desrt: darn [14:11] pitti: ok, then fixed (enhanced as well some description) and uploaded! Thanks for the review and for exercising my rusty perl :) [14:12] heh, ain't it fun [14:14] yeah, I always enjoy it, I was trying to convince desrt as well :-) [14:15] o/ [14:16] mvo, shouldn't update-manager handle removing something when needed? [14:16] kenvandine: http://paste.ubuntu.com/649995/ [14:16] dist-upgrade is holding back updates for cjohnston because indicator-me needs to be removed [14:16] dist-upgrade wants to do the right thing [14:17] kenvandine, you probably indicator-session to conflicts,replaces,provides it for easier update [14:17] upgrades rather [14:18] kenvandine, that would tell the resolver that indicator-session is providing the feature rather than breaking it [14:18] ah... so add a provides [14:19] kenvandine, c,r,p is a special combo in some way, I though you should use them together [14:19] but get mvo to confirm to be sure [14:19] i added a replaces and breaks [14:19] desrt: ok, I fixed it [14:20] Kaleo: great :) [14:20] Kaleo: i knew you'd be able to do it faster :) [14:20] desrt: tssss [14:20] mvo, ^^ ping me when you can [14:20] desrt: lazy! [14:20] Kaleo: i hate c++ :) [14:20] and i had no way to test the patch [14:20] i tried to make didrocks do it, but he was lazy too :) [14:20] desrt: lacking computers :) [14:20] desrt: so I am not so strong in GLib things [14:21] desrt: I think I spotted a leak but not sure [14:21] Kaleo: i'm sure the GLib part of the patch is right. now you fix the Qt part :) [14:21] desrt: I add the fix for that to my patch [14:21] desrt: you used a g_variant_iter_next without freeing the variable [14:22] Kaleo: do you see "&s"? [14:22] desrt: yeah, what does it mean? [14:23] it means that you don't have to free it :) [14:23] you get a pointer directly to the serialised data [14:23] so there's no copy, and no free [14:23] oki [14:23] cool [14:23] kenvandine: i think you need the Conflicts/Replaces to have update-manager remove it [14:23] desrt: so the bug I think is that it's lacking the & before the str [14:23] desrt: with the & it does not crash anymore :) [14:24] desrt: that makes sense? [14:24] it's quite possible. [14:24] let me look [14:24] oh yes. indeed. [14:24] how embarassing [14:25] that mistake was copy/pasted [14:25] so you have to fix it in both spots [14:25] desrt: I suppose it's the same issue in both places [14:25] desrt: don't you want to just do that upstream? [14:25] Kaleo: oh. indeed. [14:25] :) [14:25] if the patch is fine after the & then i'll modify and push [14:25] desrt: the patch is fine for reading lists [14:25] desrt: let me try writing [14:26] good. that will test change notification as a side effect, as well [14:26] since that's a different code path [14:26] need to make a QVariant for that case [14:27] desrt: change notification for list is supported? [14:27] desrt: (I don't need it but it's a great to have) [14:28] Kaleo: yes. it's just the same as for any other value [14:28] desrt: yeah, but I mean, if the content of the list changes, that is for example if an item is removed [14:29] Kaleo: you just get told "the key has a new value" [14:29] not that an item was removed or added [14:31] desrt: right [14:32] desrt: I have some errors when writing: [14:32] unity-2d-launcher: [CRITICAL] GLib: g_variant_builder_add_value: assertion `is_valid_builder (builder)' failed [14:32] unity-2d-launcher: [CRITICAL] GLib: g_variant_builder_end: assertion `is_valid_builder (builder)' failed [14:34] looks like i missed part of the patch [14:36] didrocks: ping [14:36] seb128, i fixed that segfault when accounts are added in the new gwibber [14:36] m4n1sh: hey [14:37] didrocks: since RainCT is now available [14:37] Kaleo: i see the problem. missing line. [14:37] didrocks: just wanted to know is there any chance of fixing that raptor issue before feature freeze [14:37] desrt: g_variant_builder_init(&builder, G_VARIANT_TYPE_ARRAY); [14:37] Kaleo: ya. exactly. [14:37] desrt: and then it works :) [14:37] that will fix it [14:37] kenvandine, the one when picking a color for which I got you debug infos yesterday? [14:37] Kaleo: cool. pushing upstream. [14:37] desrt: ok then, patch is good to go! [14:38] thanks for doing my dirty work :) [14:38] didrocks: you can do your magic! :) [14:38] desrt: you owe me a (belgian) beer [14:38] m4n1sh: would be nice to really fix it, I saw RainCT commented on it, if you fix it now, there is enough time for the next zg release to go in of course :) [14:38] seb128, no... the one you hadn't reproduced :) [14:38] desrt: pushing everything to trunk? [14:38] didrocks: already done [14:38] or maybe you had... segfault one though [14:38] kenvandine, oh, I somewhat though it was the same bug :p [14:38] Kaleo: excellent, will make a release as soon as desrt… [14:39] didrocks: afaik there is an upstream release to raptor, the changes can be backported [14:39] Kaleo: ok, doing now then :) [14:39] mildly related [14:39] kenvandine, that the one you fixed yesterday, since I didn't get the segfault since you fixed the account adding [14:39] didrocks: fantastic, I will push that into u2d's ppa right after you [14:39] it got the updated signal for an account that didn't exist which triggered the crash [14:39] kenvandine, but great! ;-) [14:39] didrocks: and then we can start merging u2d migration branches [14:39] didrocks: I am interested in this because I want zeitgeist-datasources in oneiric archive. only after zeitgeist fixes, then datasources can work properly [14:39] m4n1sh: if RainCT is pushing it on Debian, it will be even easier :) [14:39] didrocks: datasources need to brought in oneiric via Ubuntu NEW, getting is via Debian is a pain [14:39] Kaleo: indeed [14:40] Kaleo: dconf-qt needs to go in main first [14:40] m4n1sh: I meant, for zg [14:40] (1) less time left (2) debian unstable moves too fast [14:40] didrocks, m4n1sh: the raptor bug got fixed upstream I think, somebody needs to backport their fix I guess [14:40] didrocks: well, merging the branches does not depend on that; pushing the new release of u2d into Ubuntu does [14:40] seb128: yes, I was hoping RainCT would fix it as he knows the internals [14:40] seb128: they did a release apparently from what m4n1sh said, so better to keep in sync [14:40] didrocks: (which we will do on Monday?) [14:41] Kaleo: indeed [14:42] didrocks: seb128 : Does it make sense to include the new released raptor version in archives? [14:42] instead of backporting? [14:42] m4n1sh, I don't know enough about raptor to say [14:42] m4n1sh: ? I was thinking the fix was in zg rather? [14:42] dunno what is in the new version [14:43] didrocks: no, if raptor gets fixed, zeitgeist would build fine [14:43] and work fine [14:43] that is what RainCT said [14:43] we either need somebody who knows raptor to do the update [14:43] I only saw comments from RainCT and kamstrup [14:43] m4n1sh: so, maybe better to let him checking that? [14:43] seb128: doko [14:43] or to backport the fix (easier for somebody who doesn't want to break raptor or deciding on updating it) [14:43] doko handles raptor AFAIK [14:43] m4n1sh, ok, ping doko I guess then ;-) [14:44] i cant find him too [14:44] he might be away to debconf [14:44] m4n1sh, try writing him an email? [14:44] as he's the one pushing for raptor2 and who made the change in zg build-dep, waiting on him makes more sense [14:44] as we all know didrocks is the man for anything related to zeitgeist [14:44] so I pinged him [14:44] m4n1sh: zg yeah, not raptor, I just merged the change which was already done :) [14:44] well I would backport the upstream fix and let doko decide on a new version [14:44] and fix the components which were crashing because of zg [14:45] but doko decided to take the new version for a build-dep [14:45] better him to check that [14:45] I wanted to get datasources included in archives [14:45] didrocks: did you have time to look at https://code.launchpad.net/~fboucault/unity-2d/dconf_migration_super_key/+merge/67971 ? [14:45] but without a working zeitgeist it is tough to know if the datasources packages is working fine [14:46] s/is/are [14:46] Kaleo: the tab is already opened [14:46] :) [14:46] m4n1sh, can't you backport the raptor fix and test if it works? [14:46] m4n1sh, didrocks or I can sponsor a fix backport if it works [14:46] seb128: I need to try. Never done anything like that [14:46] hum [14:46] m4n1sh, do you have a vcs url or something to the upstream fix? [14:47] guys [14:47] the crash in zeitgeist that was being reported is not a zeitgeist problem [14:47] but rather a rdflib problem [14:47] that got fixed upstream [14:48] if this is rolled out then we have no real issues [14:48] seif: well all I know about this is what I saw RainCT doing [14:48] seb128: didrocks this is the bug report [14:48] https://bugs.launchpad.net/zeitgeist/+bug/807076 [14:48] seif, right, we are discussing how that will be rolled out, backport of the fix or update [14:48] Ubuntu bug 807076 in zeitgeist "raptor2 not supported" [High,In progress] [14:49] seif, is that the fix? https://github.com/dajobe/raptor/commit/37c5abe588d1ad63dc600cd9b9fd4d362f973a23 [14:50] this is the upstream bug [14:50] http://bugs.librdf.org/mantis/view.php?id=451 [14:51] seb128: seems there is 3 commits [14:51] didrocks, the other ones seems tests though? [14:52] seb128: indeed, but the testsuite is runned at build-time IIRC (long time I didn't touch it) [14:53] *grumpf* compiz freezing *grumpf* [14:53] didrocks, I would recommend assigning that bug to doko and rolling back zg to raptor instead of raptor2 until raptor2 is fixed [14:53] didrocks, he clearly didn't test is change [14:53] his [14:54] +1 [14:54] didrocks, the comments on the bug suggest backporting the fix will be annoying since it requires running autogen and some files needed are not in the tarball [14:55] Kaleo: the debian/control relaxed build-dep you did is wrong, but not very important right now [14:55] kenvandine: sorry, I got a hard freeze, what did you ask earlier? [14:55] Kaleo: I guess you did it because the natty version was earlier? (that's why I added a ~ at the end, so that you can take it in your ppa and append ~natty1 to the oneiric version) [14:56] mvo, to get update-manager to remove something on update, does it need a conflicts, replaces and provides? [14:56] mvo, dist-upgrade is wanting to do the right thing for removing indicator-me, but update-manager is holding it back [14:56] i used a breaks and replaces [14:58] kenvandine: update-manager will never remove, it will only ask about partial upgrades when it needs to remove something. so if apt-get is doing the right thing, then u-m should as well :) i.e. it should ask about a partial upgrade and then remove [14:58] ah.. so the user should do the partial upgrade [14:58] yeah [14:58] mvo, good enough for me :) [14:58] thx [14:59] well, u-m is really designed for the stable distro where you never need to remove on a regular update [14:59] yw [14:59] yeah [14:59] desrt: can you look/take as well those patches, please: http://bazaar.launchpad.net/~fboucault/+junk/dconf-qt/view/8/debian/patches/01_qmldir_so_name.patch and http://bazaar.launchpad.net/~fboucault/+junk/dconf-qt/view/8/debian/patches/02_install_headers.patch ? [14:59] makes sense [14:59] cjohnston, ^^ [14:59] didrocks: are they tested? [14:59] desrt: I think Kaleo are using them :) [14:59] i trust you :) [15:00] desrt: we have a .pc file as well, do you want it upstream? [15:00] (I think it needs to be dep on qt though) [15:00] thanks [15:01] hum no, the QtDbus transition dep should do it [15:01] transiant* [15:01] desrt: http://bazaar.launchpad.net/~fboucault/+junk/dconf-qt/view/head:/debian/dconf-qt.pc, and I can cook a git format patch to install it :) [15:02] didrocks: patches welcome :) [15:02] didrocks: i just pushed the first two [15:02] desrt: awesome, one sec, doing the pc file [15:03] seb128: can you check with Robert about getting unity-greeter into main and on the CDs next wek? [15:03] pitti, sure [15:03] would be nice to get that into a3 [15:03] pitti, https://bugs.launchpad.net/ubuntu/+source/lightdm/+bug/809710 [15:04] Ubuntu bug 809710 in lightdm "[MIR] unity-greeter" [Wishlist,New] [15:04] ah [15:04] pitti, it's blocked on MIR for now though [15:04] that should be an easy MIR [15:04] mterry, ^ [15:04] could you help there? [15:06] seb128, can you comment that on our bug [15:06] the only reason we went with raptor 2 is because you guys went with it [15:06] seb128, pitti, on it [15:07] Actually, kees better do that one for a security perspective [15:07] mterry, thanks [15:08] mterry, kees has mirs assigned to him for over a month sitting there though [15:08] mterry, well I'm not sure the greeter is a security issue, it's running with standard user rights [15:09] I whish glade would not reindent the entire file when it encounters a old file [15:09] though the current version load indicators without restricting what they can do, which is probably not something you want on your greeter [15:10] seb128, I'll review from a MIR perspective and open a bug assigned to the security team for a pass there [15:11] mterry, ok, thanks [15:11] pitti, mterry: I will need to check with robert_ancell anyway, I guess he doesn't want the current version to be default [15:11] he's trying to get a new lightdm version out for a few days [15:12] seb128, yeah, even if unity-greeter is installed, I don't think it's automatically enabled, so he can toss that switch when he's ready [15:12] so I think he will switch back to the greeter and get it in shape next once he manages to get the new version done [15:12] ah, so if the greeter is not ready yet, then let's stay with the buggy example one for now [15:12] the indicators in the unity one are certainly an issue [15:12] the [Broken UTF-8] one is weird for sure [15:12] well that's just cosmetic [15:12] seb128 is the the lightdm gtk greeter fuba currently? [15:12] what is an issue is that you can run capplets or softwares from the indicator menus [15:13] egon, mvo? [15:13] egon, "fuba" in which sense? [15:13] seb128, ups, sorry === egon is now known as mvo [15:13] egon, the trunk version has been ported to gtk3 [15:13] wrong machine [15:13] not sure how much work has been done as well on it [15:13] well, the oneiric version, segfaults for me when I click on any user [15:13] mvo, that's because you use ecryptfs [15:13] it has a known bug with it [15:14] you should be able to use "other" as a workaround [15:14] I will make sure robert_ancell get that fixed in the update if it's not fixed yet [15:15] do you happen to have the bugnumber? [15:15] seb128: I'm getting the lighdm segfault, and I'm not using ecryptfs [15:15] bug 811909 [15:16] Launchpad bug 811909 in lightdm "lightdm-example-gtk-greeter crashed with SIGSEGV in _start()" [High,Confirmed] https://launchpad.net/bugs/811909 [15:16] thanks [15:16] or bug 811524 [15:16] Launchpad bug 811524 in lightdm "Lightdm don't ask password if select user by mouse click or pressing Enter key." [Critical,New] https://launchpad.net/bugs/811524 [15:16] mdeslaur, I don't know about this one, the one I know about is "breaks when selecting an user which has an ecryptfs dir" [15:17] seb128: it's pretty much broken for everyone right now I think [15:17] desrt: here is for adding the pc file and adding it to cmake install files http://paste.ubuntu.com/650039/ [15:18] mdeslaur, works for me for my non ecryptfs users [15:18] mdeslaur, and it works on my netbook which is a stock a2 install upgraded [15:18] seb128: did you change the default greeter theme? [15:18] I did because I use ecryptfs on my laptop [15:19] seb128: I just updated my vms, they all are now broken, and I've installed a new one with yesterday's live iso and it's broken [15:19] mdeslaur, I confirm this one, .dmrc is broken in some way that makes lightdm crash. [15:19] but the default greeter was working for my non ecryptfs users as is the netbook [15:19] mdeslaur, I replaced it with a very basic .dmrc and that /fixes/ it [15:19] mdeslaur, when did that start? [15:19] dmrc does not make a difference for me [15:19] jibel, mdeslaur: .dmrc is user config [15:20] seb128: I did not update my oneiric vms in a while, so I had not noticed...it started today when I did a dist-upgrade [15:20] mvo, you use ecryptfsthough? [15:20] mdeslaur, jibel: could you try with a trunk lightdm build? [15:20] seb128, yeah [15:20] mvo, right, that's known broken [15:21] seb128, ok, do you have any more details? [15:21] seb128, I never created one. I think that's something from gdm [15:21] jibel: gdm or language-selector [15:21] gdm [15:21] seb128: are there daily packages being built somewhere I can try? [15:22] mdeslaur, no [15:22] well, wait for the next update maybe [15:22] as I said robert_ancell is trying to get a new version out for a few days, he ran into a few issues [15:22] seb128, this is what my .dmrc was http://paste.ubuntu.com/650044/ [15:22] ok, I'll wait a day to two [15:22] I will email him now and keep you updated mon monday [15:22] on [15:23] the workaround is to click on "Other" and type a username...that works for me [15:27] I guess it should at least auto-respawn segfaulting greeters [15:28] yeah [15:29] b'ah, i've got an invisible window on my screen [15:29] i thought i was going crazy! [15:29] the emperor's new windows! [15:30] lol [15:33] pitti, hum [15:33] pitti, the amd64 is having an issue [15:33] chroot.py", line 27, in setup_fakeroot_env [15:33] '%s not found; please set APPORT_LIBFAKEROOT correctly' % libfakeroot [15:33] AssertionError: /usr/lib/libfakeroot/libfakeroot-sysv.so not found; please set APPORT_LIBFAKEROOT correctly [15:33] retracer? [15:34] pitti, yes, sorry [15:34] pitti, have you seen that before? [15:34] no, I didn't [15:34] I recently changed that part of the code to fix the multiarch breakage [15:34] pitti, hum, sorry, was an error on how I ran it it seems [15:35] but anyway, this is just applied locally; I think that stuff is released now [15:35] the chroot update is failing on that in fact [15:35] The following packages have unmet dependencies: [15:35] libc6 : Depends: libc-bin (= 2.13-9ubuntu2) but 2.13-9ubuntu3 is installed [15:35] ah, no, I failed to merge trunk from the last upload [15:35] seb128: ah, ok, *phew* [15:35] pitti, sorry ;-) [15:36] no problem at all [15:36] the upgrade failure is still weird [15:36] pitti, yeah, I'm logging in to see [15:37] dpkg: error processing /var/cache/apt/archives/libc6_2.13-9ubuntu3_amd64.deb (--unpack): [15:37] subprocess new pre-installation script returned error exit status 1 [15:37] hum?! [15:37] oh [15:37] "Can't exec "/tmp/libc6.config.116121": Exec format error at /usr/share/perl/5.12/IPC/Open3.pm line 168. [15:37] open2: exec of /tmp/libc6.config.116121 configure 2.13-9ubuntu2 failed at /usr/share/perl5/Debconf/ConfModule.pm line 59 [15:37] " [15:38] oh, that again [15:38] what is it? [15:38] seb128: I got around that by temporarily replacing /usr/bin/perl with a symlink to /bin/true, IIRC [15:38] I have no idea [15:38] lol [15:38] fakechroot FTL [15:38] .config files are ridiculously hard to evade [15:39] pitti, thanks [15:39] that worked ;-) [15:42] hacks'R'us [15:42] ;-) [15:44] pitti: on what package did you apply the magic to replace the horrible C++ symbol name extraction with nice and enjoyable symbol names to read? [15:44] didrocks: sunpinyin, but I dropped the patch again as even with that it still didn't work [15:45] http://launchpadlibrarian.net/74407776/sunpinyin_2.0.3-2_2.0.3-2ubuntu1.diff.gz [15:45] pitti: oh :( so you are still generating symbols or just back to a shlibs? [15:45] just shlibs [15:45] hum, I think I'll do the same then [15:45] pitti: thanks for the info :) [15:45] de rien [15:47] tremolux, with latest software-center, ratings don't show in carousels or lists, only when i click into an item [15:47] tremolux, is that known or should I file? [15:49] mterry: we added new functionality to show top-rated, and it uses data from the server that we didn't have before. so, we clear the cache and redownload, I expect that's what you are seeing [15:49] didrocks: sorry about the mistake [15:49] tremolux, ok, so I just wait and maybe next run it will be awesome [15:49] mterry: in other words, it would be just for this one upgrade [15:50] Kaleo: no worry, I just want to know if relaxing the dep is needed for your ppa or you have a high qt version enough? (~ at the end should enable you for natty backporting) [15:50] tremolux, yup, rerunning fixed it, cool :) [15:50] mterry: yes, it should be good after this, so if it's not, please do file :) [15:50] kenvandine, gwibber thinks I've 177 messages pending now [15:50] mterry: ok cool, thanks MT! [15:50] kenvandine, I feel like I've lot of friends, I don't want to open the ui I will be disappointed :p [15:51] hehe [15:51] tremolux, now I'm seeing different items in top rated carousel vs top rated list [15:51] seb128, i'll review how that number is getting calculated [15:51] but not right now :) [15:51] mterry: haha, variety is the spice of life, no? [15:51] kenvandine, yeah no hurry ;-) [15:51] mterry: that sounds like a bug [15:52] mterry: I'll check it, you feel free to file if you like [15:52] tremolux, sure [15:52] mterry: it's brand new stuff, got some kinks it seems, thanks for helping! [15:54] didrocks: sorry, let me have a look at what I did [15:54] didrocks: ah, I remember [15:55] didrocks: I don't have a qt in my ppa [15:55] didrocks: I use the one from Natty [15:55] didrocks: which contains the needed patch :) [15:55] didrocks: hence the relaxing [15:56] Kaleo: ok, it's not totally correct then, but fair enough, let's keep it that way. Cleaning done, will push as soon as my pbuilder is happy :) [15:56] didrocks: i think paste.ubuntu.com might be the worst paste site on earth [15:56] desrt: sorry, do you want me to copy it somwhere else? [15:56] i curl on the 'plain text' version and i get a html page that says "openid transaction in progress"? [15:56] didrocks: nah. i can copy paste fine :) [15:56] didrocks: fantastic [15:57] desrt: yeah, we discussed that yesterday and you answered "wget can do it" [15:57] desrt: i wanted you to exercice this answer :-) [15:57] but seems you took the easy way, what a disappointment! :-) [15:57] wget fails :/ [15:58] yeah, that part of paste.u.c. is ridiculous [15:58] +1 [15:58] desrt: really, if you don't want to copy and paste, I can scp it somewhere else [15:58] anyway. it's pushed. [15:58] desrt: thanks! [16:00] and dconf-qt uploaded, it will need some binNEW for the -dev package [16:00] mterry: hmm, yes, so on my machine I'm seeing gparted as top rated in the carousel, but that's missing in the list. after that it seems to match. is that what you are seeing? [16:00] Kaleo: pushed to the vcs as well in case you want to backport to your ppa [16:01] tremolux, i'm seeing different apps, but same idea. only a few are different [16:02] didrocks: yep, I do [16:02] didrocks: doing it now [16:02] mterry: interesting [16:03] didrocks: silly question: where do I get the tarball from? [16:03] Kaleo: it's not that silly :) [16:04] Kaleo: http://people.canonical.com/~didrocks/dconf-qt_0.0.0.110722.orig.tar.bz2 [16:04] Kaleo: your merge proposal is for the natty though [16:05] Kaleo: I already have some comments, so let's put some there [16:05] (indicator version not being the right one) [16:05] didrocks: we have trouble attempting to read what is written in the book :) [16:05] desrt: you mean, the french or my handwriting? :) [16:06] didrocks: yeah it's for natty branch but it should work on both [16:06] the handwriting :) [16:06] Kaleo: no, because it checks for indicator.pc [16:06] didrocks: I mean [16:06] even french speakers have difficulties [16:06] didrocks: merge the branch into lp:unity-2d/4.0 [16:06] desrt: oh, that's normal, how do you think I majored at school? :-) [16:06] Kaleo: ok, doing :) [16:13] kenvandine, would you mind to check bug 728203 , seems like with the latest release (in Oneiric) is even worst :( .. thank you [16:13] Launchpad bug 728203 in gwibber "gwibber-accounts crashed with IOError in put_in_keyring()" [Medium,Confirmed] https://launchpad.net/bugs/728203 [16:14] didrocks: pushed into ppa :) [16:16] kamusin, i've looked at it... but i can't see what could be causingit [16:18] it's failing to get the default keyring... not sure what we can do in that case [16:18] it's weird, seems like he can't finish the validation process or something [16:18] i can keep it from tracebacking [16:18] gnomekeyring.get_default_keyring_sync(), [16:19] is what fails [16:19] i need to run to lunch now though, late :) [16:19] bbiab [16:19] if you need any other information just ping me ;) [16:20] kamusin, you can reliably reproduce it? [16:21] good bye everyone, see you in two weeks! [16:22] pitti: enjoy your holidays! [16:22] merci! [16:27] didrocks: https://bugs.launchpad.net/unity-2d/+bug/814709 [16:27] Ubuntu bug 814709 in unity-2d "libunity-2d-private-dev package needs to be built and published" [Medium,Confirmed] [16:27] :) [16:29] Kaleo: nice! You have a comment on your merge req. [16:29] Kaleo: and I have a fix for the .pc file to do :) [16:30] yes kenvandine, I was trying to add a facebok or twitter account and always the same behaviour [16:32] didrocks: awesome [16:35] didrocks: answered with one question for you [16:37] answered [16:47] kenvandine, bug #814717 [16:47] Launchpad bug 814717 in gwibber "scrolling can lead to empty view" [Undecided,New] https://launchpad.net/bugs/814717 [16:47] kenvandine, the description sucks but I was not sure how to describe it properly [16:48] seb128, good enough :) [16:48] thx [16:50] kenvandine, bug #814718 as well [16:50] Launchpad bug 814718 in gwibber "the indicator messages count is wrong on fb" [Undecided,New] https://launchpad.net/bugs/814718 [16:53] kamusin, what is the ownership and permissions on ~/.gnome2/keyrings ? [16:54] kamusin, and are their files in that dir? [16:54] seb128, are you aware of gnome-keyring bugs where it fails to get the default keyring? [16:54] with an IOError? [16:55] there is an identical traceback in a bug against gtkvncviewer too [16:56] kenvandine, recent bugs on oneiric both? [16:56] kenvandine, there is bug #813755 [16:56] may 5 [16:56] Launchpad bug 813755 in gnome-keyring "gnome-keyring-daemon fails to start as it can't get capabilities" [Undecided,Confirmed] https://launchpad.net/bugs/813755 [16:56] that sounds like it [16:56] kamusin said the daemon isn't running [16:56] and he doesn't have a keyrings dir [16:56] fresh install of the daily [16:57] kenvandine, not sure about the may 5 issue but the bug I just pointed might be what kamusin gets [16:57] workaround in the bug [16:57] well take the workaround from comment #5 [16:57] seb128, checking.. [16:57] not the setuid one [16:59] seb128, sweet.. work around works fine here [16:59] great [16:59] cjohnston, i bet that is why your google talk is failing too... can you try the same workaround? [17:00] seb128 thx! [17:00] thank you seb128 :) [17:01] seb128 is way better than google :) [17:01] yw ;-) [17:01] lol [17:03] kenvandine, is there any way to see the replies to comments in gwibber? [17:03] not for facebook, yet [17:04] that is a regression from the old client [17:04] * micahg would love an in-gwibber in-context button [17:04] facebook gives the comments inline in the post [17:04] micahg, there sort of is... it takes you to the web now [17:04] right, I don't like that :) [17:04] but we plan to display that inline [17:05] kenvandine, ok, another stupid questions, but is gwibber filtering what to display compared to the web ui? [17:05] sort of... it splits it up [17:05] desrt: words on the street tells that really cool people are building dconf-qt with this patch: http://people.canonical.com/~didrocks/0001-Don-t-rely-on-compiz-cmake-magic.patch :) [17:05] the "home" stream displays everything together [17:05] seb128, but links, images, videos get displayed in the attachments tab [17:05] and messages only displays status updates [17:06] but you should see everything in the home stream [17:06] seb128: can you binNEW dconf-qt (well -0ubuntu2 has a correct pc file if you prefer to wait) [17:06] didrocks, looking [17:06] seb128, if you aren't seeing everything, it is a bug [17:08] didrocks: word on the street is that those people will receive conflicts when they try to apply that patch since it's already on master [17:08] desrt: yeah, people don't live enough on the edge :-) [17:08] desrt: thanks :-) [17:15] seb128: I will upload the new software-properties dbus stuff now, if you could give it some testing [17:16] seb128: that would be cool [17:16] mvo, sure [17:16] mvo, do you want to test trunk before upload or just when it hits the archive? [17:16] seb128: I uploaded it already, but feel free to bzr-buildpackage trunk [17:16] seb128: so that we can fix any issues before it actually hits the masses [17:17] there is a binary-NEW component (the dbus backend) [17:17] so there will be some delay anyway [17:19] mvo, ok [17:19] seb128: hrm, silly me, need to remove the gksu dependency of course :) [17:19] ;-) [17:19] fixed in trunk [17:20] seb128: ok, please let me know if you find more, r692 should be good otherwise. I have dinner now [17:21] mvo, I will test in a few minutes, but enjoy your dinner and read the irclog when you come back if you want [17:21] (if you plan to come back to read the log) [17:21] thanks seb128 [17:21] yeah, I need to do one more upload [17:22] seb128: the new dconf-qt is available [17:22] didrocks, is there any reason the dev is arch: any? [17:23] didrocks, it seems it's a .pc and symlinks only [17:23] argh, didn't spot that when merging from florian, yeah, there is no .a generated [17:23] seb128: can fix that in next upload, staging the change [17:26] didrocks, newed [17:26] seb128: thanks :) [17:26] I pushed the changed in the vcs, no need for an upload now [17:28] yeah [17:31] ok, time for week-end! [17:31] see you on Monday [17:39] yeah, me too, see you on Monday! [17:40] rodrigo_, didn't you say that you were off monday? [17:42] seb128, oh yes, forgot, so yeah, see you on Tuesday! :D [17:42] rodrigo_, where is the gsd update! ;-) [17:42] rodrigo_, have fun, see you on tuesday, don't worry it can wait a few extra days [17:47] mvo, you had bugs! [17:47] $ cat /usr/share/dbus-1/system-services/com.ubuntu.SoftwareProperties.service [17:47] [D-BUS Service] [17:47] Name=com.ubuntu.SoftwareProperties [17:47] Exec=/usr/share/software-properties/software-properties-dbus [17:47] mvo, software-properties-common: /usr/lib/software-properties/software-properties-dbus [17:47] mvo, lib != share ;-) [17:47] then [17:47] $ sudo /usr/lib/software-properties/software-properties-dbusTraceback (most recent call last): [17:47] File "/usr/lib/software-properties/software-properties-dbus", line 26, in [17:47] from softwareproperties.dbus.SoftwarePropertiesDBus import ( [17:47] ImportError: No module named dbus.SoftwarePropertiesDBus [17:49] mvo, [17:49] $ dh_install --list-missing [17:49] dh_install: usr/lib/python2.6/dist-packages/software_properties-0.60.egg-info exists in debian/tmp but is not installed to anywhere [17:49] dh_install: usr/lib/python2.6/dist-packages/softwareproperties/dbus/SoftwarePropertiesDBus.py exists in debian/tmp but is not installed to anywhere [17:49] dh_install: usr/lib/python2.6/dist-packages/softwareproperties/dbus/__init__.py exists in debian/tmp but is not installed to anywhere [17:49] ... [17:52] mvo, once those installed in my local fixed build [17:52] DialogMirror.py", line 34, in [17:52] from softwareproperties.MirrorTest import MirrorTest [17:52] ImportError: No module named MirrorTest [17:54] mvo, not sure why though [17:54] $ python -c "from softwareproperties.MirrorTest import MirrorTest" [17:54] $ [17:54] ie. that works [17:55] ups [17:55] it worked because I was in the source dir [17:58] mvo, "you had bugs"->"you have bugs" (sorry, I didn't want to mislead you, I didn't fix and upload) [17:58] well I can fix the first 2 [17:58] not sure about the mirrortest one [18:05] mvo, ok, my turn to get dinner but I will read scrollbar later ;-) [18:45] seb128, so will you guys revert to raptor from raptor2 [18:45] or are you going to encapsulate the upstream fix [18:45] ? [19:01] It seems that the last update has switched thunderbird from 5 to 6 and now I cannot see my mail folders any more. [19:02] Is this known? [19:02] tkamppeter, I can still see mine... [19:02] tkamppeter, I know that it has some migration code for your profile. Maybe yours didn't migrate? [19:03] mterry, no migration pop-up appeared. [19:03] mterry, what should I do? [19:04] tkamppeter, no, it's behind the scenes, just moves .thunderbird folders around (I believe). chrisccoulson would know more [19:04] chrisccoulson, have you any idea about what to do? [19:04] hey mterry [19:04] the only migration code is to copy your thunderbird profile to thunderbird-trunk when you run the nightly [19:05] cyphermox, heyo [19:05] seif, not sure yet but I think we will roll back to the old version until the new one is fixed [19:05] tkamppeter, any errors in the error console? [19:05] seb128, ok this makes sense then [19:05] then we will revert the dependency too [19:05] thanks [19:06] chrisccoulson, I have started thunderbird from the command line, there I got a lot of messages. [19:06] I will check with didrocks on monday and ping doko about backporting the fix [19:07] `menu_proxy_module_load': /usr/bin/thunderbird-6.0/thunderbird-bin: undefined symbol: menu_proxy_module_load [19:07] tkamppeter, that message is benign [19:08] (thunderbird-bin:18362): Gtk-WARNING **: Failed to load type module: (null) [19:08] seb128: I ran into an issue updating evolution-exchange yesterday, the package is all ready and done but I couldn't upload it; I pushed it to the team branch though [19:09] cyphermox, it's not in the desktop set? [19:09] nah [19:09] chrisccoulson, these two repeat around 6 times. Can it have to do that I have something like 6 mail accounts configured? [19:09] according to cjwatson it should be possible to add it back to a seed owned by desktop to have it back in the ubuntu-desktop set [19:10] cyphermox, can you email cjwatson about getting it added? I will have a look at sponsoring it but maybe not tonight, I was just passing by after dinner to check if mvo replied to my comments [19:10] ah, I did, hence why I'm asking [19:10] or if someone else wants to just sponsor it [19:10] cyphermox, hum, desktop set need to be seeded? [19:10] tkamppeter, no, those errors are harmless [19:11] tkamppeter, are there any errors in the error console? "Tools -> Error console" [19:11] seb128: just a second [19:11] or you know what, we can just work on this monday; it's no rush [19:12] cyphermox, yeah, let's do this one monday rather ;-) [19:12] ;) [19:12] rather figure out what's wrong with bluez right now [19:12] chrisccoulson: [19:13] tree is undefined [19:13] aFolder is undefined [19:13] tree is undefined [19:14] These three have all some URI starting with chrome:// [19:14] formatURL: Couldn't find value for key: TIME_SESSION_RESTORED [19:15] tkamppeter, can you send me the contents of the startupCache folder in your thunderbird profile? [19:15] followd by a jar:file:///... URI [19:15] (~/.thunderbird/xxxxxxxx.default/startupCache, where "xxxxxxxx" is some random digits) [19:15] chrisccoulson, will do [19:16] then once you've done that, delete the startupCache folder and restart [19:16] i bet it will work [19:16] m_conley ^^ [19:16] that's the second occurrence of that now :( [19:16] i wonder if the cache isn't getting blown away after upgrading [19:17] chrisccoulson: hrm.. [19:18] chrisccoulson, what is your mail address? [19:19] tkamppeter, chris dot coulson AT canonical dot com [19:20] that's it! I can use thunderbird again! [19:20] chrisccoulson, but I will send you the file anyway. [19:20] tkamppeter, thanks [19:21] tkamppeter, i've just thought - the file may get bounced by our e-mail gateway, as it's just a zip file [19:21] not sure if there's anywhere else you can host it [19:21] although you could try sending it first [19:23] chrisccoulson, I have sent it only now. [19:24] tkamppeter, thanks, i've got that now [19:25] chrisccoulson, good luck! Perhaps the migration process has to simply delete the file. [19:25] seb128: hey my friend, thanks for the awsome testing, I will fix the issues now [19:26] mvo, wb ;-) [19:26] seb128: you did not fix anything yet in branch that I can just merge, right ;) ? [19:27] chrisccoulson, mterry, thank you very much for the quick help. [19:27] mvo, no I just fixed locally [19:27] mvo, I can do a merge request if you want but it's a 1 liner for each [19:28] ok, should be fine I will fix it up then :) [19:28] thanks again! [19:28] mvo, around? [19:32] seif: yes [19:33] mvo, do you have any clue about the mirrortest one? if I can workaround it I will test if the software run and works ;-) [19:35] seb128: no, i can not even reproduce it :/ [19:35] seb128: how did you trigger it? [19:35] seb128: I just did a fresh install i na chroot with the other two fixes [19:35] mvo, bzr checkout lp:software-properties [19:35] bzr bd [19:35] rm *kde*deb [19:35] dpkg -i *.deb [19:35] software-properties [19:35] basically [19:36] but I fixed the .install in between and ran the dbus service by hand [19:36] does it make sense? [19:36] hm, that is mostly what I did now, and I don't see the mirror test error, let me try harder [19:37] (cd ~ ; python -c "from softwareproperties.MirrorTest import MirrorTest") [19:37] <- work for me with the new code [19:37] seb128: what does dpkg -S /usr/share/pyshared/softwareproperties/MirrorTest.py [19:37] say? [19:38] python-software-properties: /usr/share/pyshared/softwareproperties/MirrorTest.py [19:38] hm, odd [19:44] mvo, [19:44] t$ strace python -c "from softwareproperties.MirrorTest import MirrorTest" 2>&1 | grep softwareproperties 2>&1 | grep -v ENO [19:44] -t [19:44] open("/usr/lib/python2.7/dist-packages/softwareproperties/__init__.py", O_RDONLY|O_LARGEFILE) = 3 [19:44] open("/usr/lib/python2.7/dist-packages/softwareproperties/__init__.pyc", O_RDONLY|O_LARGEFILE) = 4 [19:44] mvo, it works with python2.6 for some reason [19:45] $ python2.6 -c "from softwareproperties.MirrorTest import MirrorTest" [19:45] $ [19:45] $ python -c "from softwareproperties.MirrorTest import MirrorTest" [19:45] Traceback (most recent call last): [19:45] $ [19:47] seb128: haha, python b0rked again [19:47] seb128: and I thought this would be fixed for good now [19:47] :-( [19:47] dpkg -L python-software-properties |grep Mirror [19:47] ? [19:47] I have [19:47] /usr/lib/python2.7/dist-packages/softwareproperties/MirrorTest.py [19:48] in there [19:48] $ dpkg -L python-software-properties |grep Mirror [19:48] /usr/share/pyshared/softwareproperties/MirrorTest.py [19:48] /usr/lib/python2.6/dist-packages/softwareproperties/MirrorTest.py [19:48] !!! [19:48] DOKO [19:48] yeah, who else... [19:48] mvo, so it would break with python2.6 for you [19:48] "fun" [19:49] I have both, 2.7 and 2.6, but apparently you don't :( [19:49] lets see if the version that I uploaded to the archive builds properly [19:49] *fingerscrossed* [19:50] chrisccoulson: around? [20:11] dobey, yeah, for a few minutes [20:13] chrisccoulson: does https://code.launchpad.net/~chrisccoulson/bindwood/ff6-compat make it work on ff5? if so, why haven't you proposed that for merging?! :) [20:13] dobey, oh, i thought that got merged ages ago :/ [20:13] i'm sure we're using that in oneiric already [20:14] chrisccoulson: oh. did you put it in an SRU for natty too? [20:15] dobey, yeah. bindwood is on my upgrade checklist - https://wiki.ubuntu.com/MozillaTeam/FirefoxUpgradeChecklist/6.0 [20:15] chrisccoulson: and we should get that into bindwood trunk [20:15] dobey - yeah. we should also make sure it's compatible with the current aurora channel (7.0), which is now pretty much API stable [20:16] chrisccoulson: https://bugs.launchpad.net/ubuntu/+source/bindwood/+bug/801050 [20:16] Ubuntu bug 801050 in bindwood "Bindwood does not work with Firefox 5" [Undecided,New] [20:16] wtf, bindwood is still sat in proposed [20:17] that should have been copied across weeks ago [20:17] chrisccoulson: that bug report suggests it doesn't work though? [20:18] dobey, it worked here (in that, i cleared everything in desktopcouch and checked that it recreated my bookmarks when i opened firefox) [20:18] ok [20:21] chrisccoulson: can you take/dup that bug then, and get your branch proposed for merging into bindwood trunk after adding a commit --unchanged --fixes= for whatever the bug # is? i've seen a few, not sure which one your updated packaging used, but seems like a good one for it [20:25] wtf, launchpad won't let me modify some of the bug properties on bug #798484 [20:25] Launchpad bug 798484 in moon "Tracking bug for Firefox 5 transition in Natty" [Medium,Fix committed] https://launchpad.net/bugs/798484 [20:25] dobey, what did you want to modify? [20:26] chrisccoulson: well, i added upstream bindwood to it as affected, and wanted to assign it to you, since you already fixed it. but it won't let me change assignee, only status and importance [20:26] hrmm [20:27] hrmm, weird. maybe having that many affected packages/series/projects causes lp ui to go wonky [20:28] i clicked to change the status, and it made me go to the edit page, where i could assign it [20:37] dobey, I've seen that before too, for bugs with many tasks [20:38] i guess i should try running bindwood on a latest nightly to make sure it still works as well [20:38] although, i couldn't get desktopcouch to work the last time i tried [20:39] dobey, https://code.launchpad.net/~chrisccoulson/bindwood/ff6-compat/+merge/68906 [20:40] what happened to the extras repo that was talked about in orlando last year? do we have packages in that? [20:42] don't know [20:42] we can talk about it in orlando again this year, if you want :) [20:43] chrisccoulson, which one? [20:43] seb128, i think this is what i'm referring to - http://extras.ubuntu.com/ [20:44] oh, not much in there [20:44] heh [20:44] 'news' and 'suspended-sentence' [20:44] whatever those are :) [20:45] i was wondering if bindwood would be better off in there, so we don't end up with it broken for 5 weeks out of every 6 whilst we wait for someone to wave their hand to say "yep, it works", before it's copied from proposed [20:59] chrisccoulson: well, we're doing something similar for u1 stuff anyway :) [21:00] dobey, oh? it might be better to do the same thing with bindwood then :) [21:00] the current process doesn't really seem to be compatible with pushing things out quickly [21:00] bindwood is part of u1, yes :) [21:00] cool :) [21:02] chrisccoulson: if you propose the upstream fix first, and poke me, things can probably go more smoothly for that anyway. [21:04] do you think that cheese could be synced with debian? the only "important" change looks like the apport script === m_conley is now known as m_conley_away [21:23] kenvandine: any rational about the no-gnu-gettext.patch patch in cheese? [21:24] ah that was not you, sorry for bothering you