[00:00] well, guess it's time to meet? [00:00] robert_ancell: TheMuso? [00:00] Hey rickspencer3_. [00:01] https://wiki.ubuntu.com/DesktopTeam/Meeting/2009-07-21 [00:01] so it seemed like it worked well last week, when we only covered highlights? [00:01] yes [00:01] do either of you have agenda items? [00:01] no [00:01] robert_ancell: ? [00:01] no [00:01] alrighty, then [00:02] in terms of action items from last week ... [00:02] ACTION: TheMuso to test accessibility of empathy built with webkit [00:02] * TheMuso tested empathy [00:02] I think I documented it in my activity report. [00:02] * TheMuso checks. [00:02] Yes, I did. [00:03] TheMuso: you mentioned a couple of bugs, yes [00:03] but are you concerned about moving to empathy in Karmic? [00:03] do you feel that the issues can be reasonably addressed, or do you see a significant risk [00:03] TO elaborate, the dialogs and menus are fine, but the contact list is not properly read. The only way to get contact list items read is to open chat windows for each of them, and close the window again. [00:03] Online status combo was not readable properly. [00:03] hmmm === asac__ is now known as asac [00:04] is this a regression for pidgin? [00:04] As far as my knowledge of the architecture goes, I think they can be addressed, but they may end up being bugs in webkit. [00:04] Yes this regresses from the accessible use of pidgin. [00:04] ok [00:04] It also doesn't help that as far as I have checked, there is no expanded orca support for empathy. [00:04] I would like to see release blocking bugs for this [00:04] Ok. [00:04] so that if they are not fixed by, say, beta, we pull empathy and switch back to pidgin [00:05] Ok. [00:05] we should discuss with the rest of the team, but this seems reasonable to me [00:05] thoughts? [00:05] me to. [00:05] robert_ancell: ? [00:05] thoughts [00:06] I agree we should block on these issues [00:06] If we make enough noise people will hopefully step up and fix them [00:06] ACTION: TheMuso to target bugs for empathy accessibility to karmic, and to set to appropriate milestone (beta) [00:06] ok, thanks [00:06] thanks TheMuso [00:06] an announcement ... [00:07] I've arrange for a $50 subsidy us per canonical desktopper so that you can [00:07] expense some bluetooth device ... [00:07] provided that you agree to get it working in Karmic :) [00:07] * TheMuso already has bluetooth hardware that works,. [00:07] this is to encourage wider testing of BT [00:07] TheMuso: great news ... perhaps there is something else that may not work? [00:07] robert_ancell: maybe it's time for you to get a mic - :) [00:08] Well if there is, I can't think of it. I have the two common use cases being a phone and a headset. [00:08] rickspencer3_, I read that before and I will get a mic now :) [00:08] lol [00:08] TheMuso: sounds good [00:08] The only thing that would be useful is DUn/teathering, which is an app issue. [00:09] missing from the wiki is the discussion of making room on the CD [00:09] looks like long term we are going to run out of room [00:09] as we make room for new features from OLS and others, especially [00:09] Yeah. [00:09] the games package was mentioned, as well as Gimp (by me) [00:09] I am proposing to reduce the number of games by default [00:10] robert_ancell: yes, perhaps just the two or three that take up the least room ;) [00:10] There are a lot of packages that ship wav files for sounds. I should look into how they play their sounds, and see if we can't compress them to ogg. [00:10] Not reducing that much!! :) [00:10] well, 1 meg is a significant savings at this point [00:11] lastly, Till has made some progress regarding some printing related issues, also not on the wiki yet [00:11] The games are about 50M decompressed, I'd expect we can halve that [00:12] not sure that impacts either of you [00:12] 50 megs! [00:12] decompressed, I read "compressed" [00:12] 25M compressed (i.e. as .debs) [00:12] * TheMuso hardly prints as it is. [00:12] (TheMuso: dun works in blueman ... though in the end it should work directly in NM/connman yes.) [00:12] in terms of saving space, I'd like to get a little bit systematic about finding targets of opportunity [00:13] sound files is one, as I said. [00:13] I bought a printer the other week and it works great for me [00:13] robert_ancell: you have Till to thank for that [00:13] :) [00:13] tkamppeter, his is impressive indeed [00:13] ok [00:14] that's it [00:14] any other business? [00:14] I'm thinking of attending Linux.conf.au, proposals close this week [00:14] nothing from me. [00:15] robert_ancell: sounds good , but ... [00:15] we general have time off for one conference per year (other than UDS of course ;) ) [00:15] same with supporting travel, etc... [00:15] we should discuss in our next call, though [00:15] sure, will do [00:15] as we can make exceptions for important things [00:16] so that's it for the meeting, you guys can get back to your coffee machines ;) [00:16] there is about 2.6MB of wav files in usr/share/sounds. [00:17] My coffee is a walk of a morning. :) [00:17] TheMuso: you are too healthy [00:17] So if we can get that down somewhat, that would be good. [00:17] my coffee is coffee [00:17] TheMuso: right, plus ogg is open goodness [00:17] Indeed. [00:18] in terms of fitting onto the CD, I'd like to see a bit of a system for figuring it out: [00:18] 1. estimate how much space will we need [00:18] 2. a list of things that can be done to save space [00:18] It all comes down to whether the app needing the sounds can play ogg somehow./ [00:19] Is there a summary of what is on the CD? I was looking for a package list and even finding that is hard [00:19] 3. a decision making process to determine which items from #2 to do [00:19] The seeds are the best place to look./ [00:19] robert_ancell: I think this is hard [00:19] right, you have to look at the seeds [00:19] TheMuso, I've looked at the seeds but I really want the output of the seeds [00:19] I think we can brainstorm the items for #2 [00:19] lp:~ubuntu-core-dev/ubuntu-seeds/platform.karmic and lp:~ubuntu-core-dev/ubuntu-seeds/ubuntu.karmic [00:20] * rickspencer3_ goes back to updating tutorial for quickly [00:20] ug [00:20] TheMuso, yes but how do I convert those into which packages get pulled in? [00:20] robert_ancell: the ubuntu-archive home dir on the people.ubuntu.com server is probably the best place to find such info. [00:21] I think there is a list in there somewhere. pitti or cjwatson would be able to give you more info. [00:22] http://people.canonical.com/~ubuntu-archive/germinate-output/ubuntu.karmic/desktop ? [00:23] robert_ancell: Which printer did you buy? [00:23] HP Deskjed F4280 printer/scanner [00:23] Was about $60 from office works [00:24] robert_ancell: HP does a good part on it, with their HPLIP drivers and also giving me their printers for testing. [00:25] robert_ancell: Currently I work on the new udev-based Plug'n'Print to make it scale, with a laptop and 10 USB printers connected to it. [00:25] tkamppeter, I bought HP because of what I had heard about HPLIP. The scanning works well but xsane is a terrible interface [00:26] tkamppeter, that's a lot of printing!! [00:26] robert_ancell: In Karmic xsane is about to be replaced. [00:26] by something (hopefully) more sane. [00:26] tkamppeter, printing is great in Ubuntu - I had the father test and he said it was easier than in Windows (and I didn't need a support call) [00:26] :) [00:28] Windows is this old-fashioned Plug'n'Play: You put in the CD and play for hours, rebooting, connecting the printer, rebooting again ... And Linux has Plug'n'Print. [00:36] tkamppeter: lol [00:37] rickspencer3_: but its so true. [00:37] TheMuso: I wouldn't know [00:37] I haven't used Windows since I worked at Microsoft [00:38] well, not entirely true, I installed Windows in a vm so I could sniff the USB interacton with this little photo viewer keychain thing I bought at Fry's [00:38] Right. I remember setting printers up for people and having to do a similar procedure to what tkamppeter outline. [00:38] outlined [00:39] I tell this story all the time ... [00:40] about a year ago, I bought the very first eeepc model, like the first week they were out [00:40] got ubuntu running on it ... I plugged it into the usb printer, and went to get a chair so I could sit down and configure the printer ... [00:41] in the moments I was gone, it had configured itself! there was a little pop-up: "your printer is ready" (or similar) [00:44] Ubuntu is at the point where if things work, they work very very well. [00:45] so that doesn't surprise me. [00:51] TheMuso: and if they fail, they fail in spectacular ways? [00:51] ajmitch: Depends on the hardware I think. [00:51] Audio for example, is very much a mixed bag./ === rickspencer3_ is now known as rickspencer3-afk === tkamppeter_ is now known as tkamppeter [06:10] Good morning [07:16] mmm, my panel has now two different volume applets. Is that expected? why are they different? [07:16] * ara is running karmic [07:17] ara: it's a known problem indeed [07:17] but not entirely trivial to solve [07:18] pitti, can you provide some background (or a bug number), please [07:19] ara: you'll still see the classical "applet" with the horizontal slider, from gnome-applets [07:19] but recently, pulseaudio started to have a notification area icon as well [07:19] pitti, I do [07:19] we somehow need to remove the former on upgrades [07:19] disabling the former by default should be easy [07:20] pitti, which one are we going to remove? the gnome one or the pulseaudio one? [07:20] the gnome one, I think [07:20] mmm, the pulseaudio one does not mute on double click, which is a really nice feature === WelshDragon is now known as Fluffles [07:52] Note that the pulseaudio one is provided by gnome-media. [08:10] * robert_ancell is going mad battling glib+dbus+policykit [08:11] hey robert_ancell [08:11] hey pitti [08:11] robert_ancell: what are you working on? gdm config? [08:11] robert_ancell: btw, I was gonna ask you about that [08:12] pitti, yes - I figure we need to have some PolicyKit in gdm to control writing the configuration [08:12] ask away [08:12] robert_ancell: I assume that /etc/gdm/custom.conf is just a backwards-compatibility crutch, and it should actually be configured through gconf? [08:12] robert_ancell: can we use the existing PK magic in gconf for that? [08:12] pitti, No, it appears to be actual configuration though why it hasn't been migrated to gconf I don't know [08:13] otherwise we'd need a PK service to gdm itself to write custom.conf [08:13] hm, that's bad [08:13] there is a dbus service to read and write that config but you need to be root to use it [08:13] oh, there is already? nice [08:13] pitti, I believe they had trouble having a daemon running gconf which is why it ended up like that [08:13] then the PKification shouldn't be too bad [08:14] robert_ancell: ah, I thought they'd use it for the greeter session [08:14] but the greeter might not even run for autologin [08:14] pitti, or so I thought... Are you a PK expert? I'm trying to gather up information to work out how to use it. It seems very intolerant of my errors [08:15] robert_ancell: expert is too much, but I did use it myself in jockey [08:15] but I only ever implemented it myself in Python, not C [08:16] although with the libpolkit-gobject stuff it should actually be much easier in C than with the raw d-bus API in Python [08:17] robert_ancell: so by and large, you define a new privilege in /usr/share/polkit-1/actions/org.gnome.gdm.policy (say "org.gnome.gdm.set_autologin") [08:17] The current problem I have is passing a GError on authentication failure causes the error "arguments to dbus_message_new_error() were incorrect, assertion "_dbus_check_is_valid_error_name (error_name)" failed in file dbus-message.c line 1211". I have no idea why is wrong with my GError [08:17] allow_any =auth_admin should do [08:17] pitti, which has to be in lowercase I found out [08:17] right [08:18] robert_ancell: did you initialize the GError before? [08:18] (or set NULL if it initializes one for itself) [08:18] I usually see this if doing something like [08:18] GError e; [08:18] Yes, it's a standard GError - looking through the dbus code it gets the GError "name" and checks that is "asd.asdasd.asd" or similar [08:18] call_function (..., &e) [08:19] But I can't find out what I'm supposed to do to make a dbus-compatible GError [08:19] hm, that's the part I didn't do in Python then [08:20] might be worth stealing some code from existing PK services, such as ConsoleKit or devkit-disks [08:20] robert_ancell: do you have your current code somewhere? [08:20] Yes, been looking though those [08:20] sure [08:20] * pitti takes a look [08:21] http://paste.ubuntu.com/224142/ [08:22] hm, polkit_authority_check_authorization() doesn't even take a GError? [08:22] So it fails in dbus_g_method_return_error [08:23] uh, I'm afraid I never touched that [08:23] robert_ancell: does it get passed "NULL" or so? [08:23] good morning there [08:23] bonjour seb128 [08:23] hey robert_ancell pitti [08:24] hey seb128 [08:24] robert_ancell, btw I chatted with rick yesterday who said you found polkit complex [08:24] pitti, no, the printf about works [08:24] robert_ancell, did you look at polkit1 or the previous version? [08:25] seb128, I'm just looking at the Ubuntu packages and the latest docs. What is the difference? [08:26] hey pitti, seb128 and robert_ancell [08:26] robert_ancell, I don't use polkit but patches for using the new one are usually dropping some hundred lines of code [08:26] robert_ancell, policykit-1 is the new one [08:27] seb128, what is the Karmic installed version? [08:27] robert_ancell: no idea about the dbus_g_method_return_error(), I'm afraid [08:27] I see libpolkit2 [08:27] pitti, thanks for looking [08:27] oh, indeed, that's policykit 0.9 code [08:27] robert_ancell, we have both because things are not migrated yet to policykit-1 [08:28] robert_ancell: in polkit-1, you don't need to do anything on the client side, and the server-side code is much easier [08:28] http://people.freedesktop.org/~david/polkit-0.92-newdocs/ [08:28] http://cgit.freedesktop.org/PolicyKit/tree/docs/PORTING-GUIDE [08:30] robert_ancell, btw, https://wiki.ubuntu.com/DesktopTeam/Specs/Karmic/LoginExperience too is worth reading if you didn't yet [08:31] polkit-1 looks a lot better. [08:31] robert_ancell, doing the change to add the option menu should be easy enough and would be the place where to have the autologin option or a way to start the gdmsetup dialog [08:31] robert_ancell, right ;-) that's why I mentioned it, I was not sure if you were looking at the new version [08:32] The PolKit web presence could do with a tidy up :) [08:32] robert_ancell: yeah, it got much cleaner in the rewrite [08:33] before that, the client had to do all the juggling to get authentication, etc. [08:34] seb128, do you know if the design team wants to have a password prompt when setting the default user? [08:35] I guess there has to be one [08:35] you mean have to enter your password to change the default autologin user? [08:35] yes [08:35] yes probably or anybody could do it on the login screen [08:36] it needs to ask for an admin user, and her password, yes [08:36] this also means that we need to start polkit-gnome in the gdm session [08:36] yeah [08:39] I'm off for an hour for some errands, bbl [08:39] pitti, see you [08:40] I get a conffile prompt on jaunty->karmic upgrade for gdm/custom.conf - is that a known issue? [08:42] mvo_, no, custom.conf didn't exist in jaunty [08:43] seb128: this is a automatic install, I did not do any customization with it [08:43] seb128: what does normally create it? [08:43] (image was created with ubuntu-vm-builder) [08:44] mvo_, it's shipped with the new gdm in karmic [08:44] mvo_, we used to have a gdm.custom-conf in jaunty [08:44] seb128: its custom.conf, I just checked the log [08:44] mvo_, well it was gdm.conf-custom in jaunty [08:46] seb128: right, it seems like its doing some magic in the preinst that may well confuse the dpkg conffile prompt code [08:46] mvo_, the preinst code just rename the file [08:46] if [ -f /etc/gdm/gdm.conf-custom ]; then [08:46] echo "Renaming /etc/gdm/gdm.conf-custom to /etc/gdm/custom.conf" >&2 [08:46] mv /etc/gdm/gdm.conf-custom /etc/gdm/custom.conf [08:46] fi [08:46] dpkg --compare-versions "$2" le-nl "2.26.1-0ubuntu5"; then [08:47] mvo_, ^ basically [08:47] I see a echo '[daemon]' >> /etc/gdm/custrom.conf here too [08:47] mvo_, that's only if there is an autologin setting to migrate [08:47] and there is a typo in the next line, going to fix this one [08:47] good morning everyone [08:48] hey ch [08:48] hey chrisccoulson [08:48] hey chrisccoulson [08:48] hey seb128, mvo_ [08:48] seb128: I have a look why I get the prompt [08:48] but I promise I did not do customization :) [08:49] mvo_, ok, ignore the typo on the echo $settings line [08:49] mvo_, did you enable autologin in the installer? [08:49] seb128: no [08:49] ok, so the echo lines should not be used [09:00] seb128: is gdm in bzr? [09:00] mvo_, yes, why? [09:01] seb128: to get rid of my conffile prompt :) [09:01] mvo_, apt-get source gdm should tell you ;-) [09:01] mvo_, what do you want to change? [09:01] mvo_, don't worry I'm just curious about what is the bug ;-) [09:03] hey seb128 - i finished the g-s-d update last night, including the fix for the disk space warning that seems to be annoying people [09:03] chrisccoulson, hey good, I was just looking to the sponsoring bug but it didn't change [09:03] seb128: so it seems like gdm checks for gdm.conf-custom and (unconditionally) moves it to the new custom.conf - but it has a different md5sum from the custom.conf that is shipped in new-gdm. because it happens in preinst dpkg thinks the user has modified the file [09:03] i had to make a small change to that though, as i noticed a couple of bugs ;) [09:03] mvo_, oh, how do we fix those sort of issues? [09:03] later all [09:03] seb128: I'm not sure what is ideal, I guess not moving if the old gdm.cutom-conf was not modified [09:04] seb128 - i pushed the changes to my own branch for now. i can push them to ubuntu-desktop unless you want to review the changes first [09:04] seb128: morning [09:04] chrisccoulson, is there many changes? it seemed mostly a .install update a patch from git [09:04] seb128: but that would still leave users with a customized one with a conffile prompt [09:04] chrisccoulson, ie those should be fine for ubuntu-desktop directly [09:04] davmor2, hi [09:05] mvo_, which means the majority of users, almost everybody used gdmsetup one day [09:05] seb128 - yeah, that's basically it. i had to make a small change to the patch from git though, as i noticed some small bugs [09:05] i'll send those upstream today [09:05] seb128: its way to difficult to do this kind of stuff IMO, just look at http://wiki.debian.org/DpkgConffileHandling [09:05] seb128: gnome-codec-install issue was confirmed yesterday by sbeattie and fader_ on #u-testing [09:05] mvo_, I told pitti that we should keep using gdm.conf-custom as naming [09:05] seb128: you think so? I never used it :) but yeah, if its a big portion, then the solution is not good enough [09:05] davmor2, good for them? [09:06] I thought I'd added details for it but that was in a previous attept to write the bug that kept timing out sorry:) [09:06] davmor2, we get some hundred bugs confirmed every day, that's useful work but not worth mentioning on IRC [09:06] davmor2, not even sure what bug you are talking about there [09:06] seb128: but the code in the wiki page does not move between package AFAICS [09:06] mvo_, anybody who enabled timed login, autologin, changed the theme, etc [09:06] * mvo_ nods [09:07] bug 402773 [09:07] Launchpad bug 402773 in rhythmbox "Karmic: gnome-codec-installer is not finding mp3 but rather looking for input-selector" [Low,Incomplete] https://launchpad.net/bugs/402773 [09:07] seb128: yeah, I'm sure that are many people [09:07] davmor2, well the bug is basically useful, it doesn't describe the issue or what to do to trigger it [09:07] I wonder if we could just ship the same file in the new gdm (with all the header text) so that the md5sums of the original match [09:08] davmor2, would be useful if rather than confirming bugs people could describe what they do to trigger the error [09:08] * mvo_ scratches his head [09:08] mvo_, we used to distro patch the gdm.custom-conf to undo upstream change for that purpose before [09:08] mvo_, sometime conffile handling is ridiculously complex [09:09] ++ [09:09] seb128: Yes again sorry for that I got tied and lp kept timing out on the bug I wrote manually so in the end triggered ubuntu-bug in the hope that lp would accept the report [09:09] we could not ship one in the package but create it on the fly (if there is none already) [09:10] davmor2, anyway IRC confirmation are not so useful, could you rather reply to the asked details on the bug? [09:10] davmor2, I don't doubt it's an issue for some user but they don't say what they try to play, how, etc [09:11] seb128: I have. would it be useful to add a screenshot of the end result too? [09:11] mvo_, we don't need to ship one at all, that's only a file for user config [09:11] hmm are there supposed to be scripts installed into /etc/pm/sleep.d? i thought the correct place for all package-managed pm-utils scripts was /usr/lib/pm-utils/sleep.d [09:11] davmor2, you are not new to bug reporting right? anything which make easy to trigger the bug or explain it correctly ... [09:11] davmor2, it's not complicated, a bug just need to be clearly described, ie have easy steps explaining it or a screenshot if you fail to describe those [09:11] ie [09:11] * start rhythmbox [09:11] * add a mp3 [09:12] * double click on it [09:12] or whatever you are doing [09:12] neither the description nor the error are clear in that bug [09:13] "is actually looking for gstreamer input-selector" [09:13] how do you figure that? [09:13] is that an error? something listed in g-c-i [09:13] [09:13] hyperair, not sure you are on the right channel for that question [09:14] seb128: hmm whoops. which channel should i be on then? -devel? [09:14] well you can try [09:14] I mean I've no clue about pm-utils [09:14] and I guess it's the case for most of desktopers [09:14] you can wait for pitti to come back or try another channel [09:17] alright. [09:29] slomo, hi [09:30] seb128: what is the new gdmsetup called - is there one ? [09:30] mvo_, no there is none but robert_ancell is working on writting one [09:30] ok, he will have to make sure it works without a custom.conf :) [09:30] mvo_, ;-) [09:32] seb128: I think we should be fine with http://paste.ubuntu.com/224167 - unless I miss something, but gdm still starts without a custom.conf for me [09:32] seb128: I just tested 2.26 -> 2.27 and no conffile prompt for me anymore (but for a really meaningful test I have to wait until its in the archive and the auto-upgrade tester can make its job) [09:32] mvo_, that's fine with me [09:32] mvo_, do you want to do the upload now? [09:32] seb128: ok, I will commit and we can upload after the freeze [09:32] ok good [09:33] mvo_, https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-karmic-gdmconfig [09:33] hah, got my climbing equipment rented now [09:33] mvo_, if you want to add a note there [09:34] seb128: should I fix the typo (gdm.conf.conf along the way?) [09:34] pitti, you go climbing? sounds dangerous, be careful we need you ;-) [09:34] mvo_, yes please and upload then ;-) [09:34] mvo_: perhaps we should not make custom.conf a conffile in the first place? [09:34] mvo_, you don't want to clean the autologin option at each upgrade [09:34] pitti: yeah, I currently just removed it from being shipped [09:35] seb128: just a fixed rope route, not _that_ dangerous :) [09:35] ups [09:35] pitti, ^ [09:35] pitti, good ;-) [09:35] * seb128 hugs pitti [09:35] seb128: "clean the option"? [09:35] pitti, well if custom.conf is not a conffile it will be overwritten at each upgrade [09:35] how so? [09:35] pitti, better to just not ship a custom.conf at all [09:35] the migration removes the old gdm.conf-custom [09:36] seb128: right, I don't particularly mind [09:36] pitti, if you have a custom.conf in the gdm binary upgrade will install the new version on disk no? [09:36] the postinst coudl create one for having a template for the user to change [09:36] seb128: that would be a conffile [09:36] seb128: hi :) [09:36] mvo_: perhaps we should not make custom.conf a conffile in the first place? [09:36] pitti, I'm not sure to follow the discussion ;-) [09:37] seb128: right; it's currently a conffile, which causes the dpkg prompt [09:37] slomo, hey, is https://bugs.edge.launchpad.net/ubuntu/+source/totem/+bug/402221 a known issue? [09:37] Ubuntu bug 402221 in totem "[Karmic ] GstDecodeBin2: This appears to be a text file" [Undecided,New] [09:37] so if we make it a normal config file, handled by postinst, we avoid the prompt [09:37] pitti, ah right, that would work too [09:37] though I'm in favor of as little postinst work as possible [09:37] seb128: operationally it's pretty much the same, but providing a default is easier for admins to change, IMHO [09:37] right [09:37] ie if having no config in the package and no postinst magic is possible the better [09:37] it's possible [09:38] we could still have an example config in /usr/share/something [09:38] but right, once Robert implements the GUI for configuring gdm, we would have had the same problem again [09:38] seb128: *nod* [09:38] pitti, why? there is no conffile prompt if you don't have a conffile [09:38] is custom.conf would be an user written data [09:38] not something managed by a package [09:39] seb128: right, I meant "with the current version' [09:39] i. e. it doesn't matter how the file is named, we'd have the same problem with a conffile [09:39] -> conffile is bad [09:39] right [09:39] so let's not ship any custom.conf at all [09:39] and let the gdmsetup write one [09:40] mvo_, seb128: so we should fix .install to instlal /etc/gdm/custom.conf into /usr/share/gdm/examples/ ? [09:40] +1 [09:40] pitti: sure, we can do that too [09:40] seb128: one moment [09:40] * mvo_ changes bzr [09:40] mvo_: ah, you already changed it; nevermind [09:41] well just change the rules line to a mv one [09:41] easy enough ;-) [09:41] omph, nasty typo, thanks for fixing [09:41] yeah, just go ahead [09:41] seb128: not known but interesting, thanks :) [09:41] slomo, do you want me to bugzilla it? [09:42] should it still be uploaded even though we are in the freeze? [09:42] mvo_: yes, I think so [09:42] mvo_, ask pitti, I tend to be liberal about small fixes during soft freezes [09:43] target of opportunity, and it's not intrusive [09:43] and fixes upgrades [09:43] but that's only me I don't want to recommend that to everybody ;-) [09:43] seb128: yes, might be a totem bug though but first file it on gst-plugins-base [09:43] slomo, ok, doing that [09:43] ok [09:43] things to avoid now: polkit-1 migration and soname breaks [09:43] things encouraged: fix RC bugs and other bugs which don't change package structure and are tested [09:44] seb128: oh wait [09:44] mvo_: please go ahead; I'm a bit busy with alpha-3 stuff [09:44] pitti: sure, can do [09:44] * pitti hugs mvo_, thanks for the fixes [09:44] seb128: http://bugzilla.gnome.org/show_bug.cgi?id=587704 [09:44] Gnome bug 587704 in gst-plugins-base ""GstDecodeBin2: This appears to be a text file" error when playing files from a samba share" [Normal,Unconfirmed] [09:44] np [09:44] mvo_: oh, and thanks for adopting gdm :-P [09:45] *muuuraaagghhh* [09:45] slomo, doh, just opened a bug so closing it now [09:45] * mvo_ crawls under a rock [09:45] * pitti watches mvo_ run away screaming [09:45] seb128: sorry but at least you have bug closing powers :) [09:46] mvo_: so you did an upgrade test to current karmic? how bad is it? [09:46] slomo, no problem ;-) you don't? [09:46] pitti: suprisingly good actually [09:46] pitti: just the conffile prompt during the upgrade [09:47] slomo, comment added upstream on the other bug and new one closed, thanks! [09:47] pitti: no errors, it still boots afterwards [09:47] seb128: i do but often i have to close bugs where the reporter writes a comment that it can be closed ;) [09:47] mvo_: nice, can you please add that to http://iso.qa.ubuntu.com/qatracker/test/2808 ? [09:48] slomo, btw do you know if anybody fixes libmms issues? got another totem crasher due to it [09:48] pitti: adding it now [09:48] seb128: i don't know, at some point upstream was dead, then someone else took it. no idea if someone is maintaining it currently [09:49] slomo, ok [09:49] mvo_, congrats on take over gdm btw ;-) [09:49] * mvo_ is still hidding under a rock [09:49] oh, we even have system/shutdown back :) ? [09:50] mvo_, you have been living under this rock for how long now? ;-) [09:50] * seb128 hides [09:50] seb128: I tend to shutdown my machine with "sudo halt" :) [09:50] as all real men do? ;-) [09:50] seb128: but thats great to have it back [09:50] well, that certainly works :) [09:50] real men yank out the power chord [09:50] and have journalling [09:50] mvo_, it's temporary until the slack^Wdxteam update fusa [09:51] :/ [09:51] mvo_, otherwise we would have no action to stop the box at all right now [09:51] pitti: yanking out my cord switches me to the battery! =p [09:51] * hyperair uses gnome do to shut down [09:51] not for me, I don't keep the battery in the laptop usually [09:52] hmm saving the battery life? [09:53] yes [09:53] constantly charging/discharging it isn't exactly helpful [09:53] and even if it's fully charged, they still get a constant trickle of charging power [09:54] seb128: could you please have a look at the gdm bzr (just so that I don't upload with embarassing typos/thinkos or anything) [09:54] I wish any notebook vendor would invent a simple switch to physically disconnect the battery without ahving to take it out [09:54] mvo_, ok [09:54] pitti: I guess part of the problem is that this won't stop the heat from the machine to get to the battery [09:54] mvo_: looks good to me [09:55] pitti: thanks! [09:55] * pitti injects "debcommit -r" into mvo's brain [09:55] pitti: new merge proposal (https://code.edge.launchpad.net/~didrocks/python-distutils-extra/update_control_file/+merge/9090). Just to comment on the "erasing the entire debian/control", that's not what my code does. It's only updating the automated generated fields, leaving fields order or added fields pristine. Btw, I added the force-control switch :) [09:55] didrocks: ah, nice [09:56] mvo_, looks good to me too [09:56] hey didrocks [09:57] hey seb128 (seems everyone missed my "good morning" on the chan two hours later ;)) [09:57] (sorry, my "good morning" missed everyone... damn English ;)) [09:57] bonjour didrocks! [09:58] Guten Tag pitti :) [10:01] slomo, any idea about https://bugs.launchpad.net/bugs/402773 too? [10:01] good morning didrocks ;) [10:02] hi chrisccoulson [10:02] hey didrocks! [10:03] Guten Tag mvo_ (it's like a FIFO list with one element at each time ;)) [10:04] seb128: yes, fixed with gs-tplugins-base 0.10.23.3 (input-selector was needed by playbin2 but is in gst-plugins-bad) [10:05] slomo, ok thanks [10:05] Ubuntu bug 402773 in rhythmbox "gnome-codec-installer is not finding mp3 but rather looking for input-selector" [Low,Incomplete] [10:06] slomo, ok thanks [10:07] * didrocks hopes that KMS will soon enable less seb128 flickering :) [10:07] didrocks, lol [10:08] could the ubuntuone-client guys stop turning nautilus is a crash land? [10:09] * seb128 reassign 3 new nautilus crashes to ubuntuone-client [10:09] seb128: all dupes? should we set up a bug patterN? [10:10] pitti, well those are new ones, doesn't help that none of the 15 retracing I looked at this morning all failed to give any debug symbol [10:10] I'm wondering if the retracers are broken [10:11] would be worth checking the log for one of the bugs, perhaps dbgsyms are missing [10:11] pitti, well, that's over 6 different components now [10:11] nautilus, gnome-media,ubuntuone-client,notify-osd, gedit [10:12] * seb128 has a look to the retracer log [10:13] nothing obvious in the log [10:15] * seb128 logs into a amd64 karmic retracer [10:17] what the? [10:17] # ls /usr/bin/gnome-volume-control -l [10:17] -rwxr-xr-x 1 1000 2552 144656 Jul 16 22:40 /usr/bin/gnome-volume-control [10:17] # gdb /usr/bin/gnome-volume-control CoreDump [10:18] ... [10:18] /usr/bin/gnome-volume-control: No such file or directory. [10:18] (no debugging symbols found) [10:18] Core was generated by `gnome-volume-control'. [10:18] Program terminated with signal 11, Segmentation fault. [10:18] #0 0x00000000004100a3 in ?? () [10:18] [10:18] "No such file or directory."? [10:18] seb128: fakechroot bug? try a relative path? [10:18] # /usr/bin/gnome-volume-control [10:18] ** (gnome-volume-control:5799): WARNING **: Cannot open display: [10:18] # gdb gnome-volume-control [10:18] but it might explain broken retraces [10:18] /usr/bin/../lib/debug//usr/bin/gnome-volume-control: No such file or directory. [10:19] gnagnagna [10:19] gdb usr/bin/gnome-volume-control I meant [10:19] /usr/bin/../lib/debug//usr/bin/gnome-volume-control: No such file or directory. [10:19] I tried the apport chroot test suite on current karmic, it worked, but it doesn't cover everything [10:19] I think retracers are officially broken [10:20] should we comment the cron job? [10:20] please just comment the karmic entry in chrootmap [10:20] unless you are sure that juanty/hardy etc. are affected as well [10:21] pitti, I will check that and comment the 9.10 line if jaunty is working [10:22] seb128: thanks muchly [10:22] you're welcome [10:22] brb [10:23] * pitti continues to fight with CD builds [10:46] pitti, the jaunty retrace seem to work so I only commented karmic [10:47] seb128: thanks [10:47] * pitti blames fakechroot [10:47] we will have to retag bugs once that's fixed, we have some hundred of failed retraced this week [10:47] not the perfect time to enable apport by default [11:02] seb128: ok, can you please give me a quick heads-up what you did? I'm done wrestling with CDs for now, and can look into this [11:02] pitti, I just commented the 9.10 line for now [11:02] to get the issue? [11:02] * pitti logs into i386 chroot [11:03] yes [11:03] logged into a karmic amd64 retracer [11:03] apt-get install gnome-media-dbgsym [11:03] got a crashdump from a launchpad bug [11:03] and tried to use gdb on it [11:03] got the file no exist error [11:03] where running the binary by hand works [11:03] ah, you just downloaded the core dump, I guess? [11:03] yes [11:03] which bug#? [11:03] I wanted to try if the coredump was working in gdb [11:04] pitti, bug #402827 [11:04] Bug 402827 on http://launchpad.net/bugs/402827 is private [11:04] seb128: merci [11:04] pitti, bug #402827 [11:04] Launchpad bug 402827 in gnome-media "gnome-volume-control crashed with SIGSEGV in IA__g_closure_invoke()" [Undecided,New] https://launchpad.net/bugs/402827 [11:04] set it public now it's only a mixer crash [11:05] root@ronne:/# /usr/bin/gnome-volume-control [11:05] /usr/bin/gnome-volume-control: /lib/libc.so.6: version `GLIBC_2.9' not found (required by /tmp/tmplPa6MR/usr/lib/libgio-2.0.so.0) [11:06] meh [11:06] was working for me [11:06] or rather I was getting a no display set error [11:07] I'm in the i386 chroot [11:08] anyway, "gdb /usr/bin/gnome-volume-control" reproduces the problem perfectly well [11:08] right [11:08] I'm really clueless about how to debug such issues [11:09] exec-file usr/bin/gnome-volume-control [11:09] that works [11:09] #($*#$ fakechroot *(*#$ [11:09] what is exec-file? [11:09] in gdb [11:10] doing gdb usr/bin/gnome-volume-control works [11:10] same as $1 [11:10] but it doesn't manage to find the debug symbols then [11:10] really? that doesn't even work here [11:10] /usr/bin/../lib/debug//usr/bin/gnome-volume-control: No such file or directory. [11:10] right [11:10] that's what I mean [11:22] 14937 stat64("/home/ubuntu-archive/apport-retracer-i386/chroots/karmic/usr/bin/gnome-mount", {st_mode=S_IFREG|0755, st_size=80860, ...}) = 0 [11:22] 14937 open("/usr/bin/gnome-mount", O_RDONLY) = -1 ENOENT (No such file or directory) [11:22] so, it's not wrapped correctly [11:22] that's all magic to me how that thing is working [11:23] seb128: essentially, it is a preloaded library which wraps all libc calls like open(), stat(), etc. [11:23] and prepends the fake chroot path [11:23] but newer glibcs tend to introduce newer variants of calls, which fakechroot has to catch up with [11:23] pitti, well it didn't change recently in karmic why does it break out of the blue now? [11:23] ah ok [11:23] glibc did change [11:23] so it's doko's fault [11:23] ;-) [11:24] oh, nice [11:24] EGLIBC BROKE OUR DEBUG PROCESS! [11:24] it does work in a karmic dchroot [11:25] Keybuk, right ;-) [11:25] pitti, can we retrace jaunty bugs also there? [11:25] seb128: that's what I'm going to check right now [11:25] I wouldn't like yet another duplication of retracer setups [11:26] I'm testing jaunty dchroot with intrepid/jaunty/karmic fakechroots now [11:27] /CoreDump" is not a core dump: File format not recognized [11:27] huh? [11:27] oh, it's an amd64 core === yofel_ is now known as yofel [11:28] pitti, want a i386 one? [11:28] looking [11:28] meh, suddenly all bugs I'm looking at are amd64 [11:29] anyway, I'm fairly sure it works in the jaunty dchroot [11:29] pitti, bug #402940 [11:29] Bug 402940 on http://launchpad.net/bugs/402940 is private [11:30] pitti, bug #402202 [11:30] Launchpad bug 402202 in evince "evince crashed with SIGSEGV in cmsDoTransform()" [Medium,Incomplete] https://launchpad.net/bugs/402202 [11:33] meh, no luck in karmic dchroot [11:33] I would argue to have karmic working and break jaunty if we can have only one right now [11:34] ie quick way would be to switch to karmic and comment jaunty [11:34] right, if only karmic would work in the karmic dchroot [11:34] but apparently it doesn't [11:34] oh, nice [11:34] it does work in a karmic dchroot [11:35] ? [11:35] I tested that with /usr/bin/gnome-mount [11:35] ah [11:36] hrmpf [12:03] seb128: is there a bug for the two mixer applets already? [12:04] 2 mixer applets? or 1 mixer applet and 1 g-v-c-applet? [12:04] i'm sure i've seen a bug for that [12:05] pitti - bug 401294 [12:05] well, applets, notifications, etc. [12:05] pitti, you seem to be strongly annoyed by that one ;-) [12:05] bug #401294 [12:05] hmmmm [12:05] seb128: I need it for the release notes [12:06] chrisccoulson, bot is being slow recently [12:06] seb128: it seems fixing that for new installs/live images is trivial, it just needs a solution for upgrades? [12:06] * hyperair tossed the panel applet away [12:06] thanks [12:06] seb128 - i was just about to ask if the bot was sleeping ;) [12:06] the other one went into the notification area didnt it? [12:06] pitti, right, I'm still not clear what we should be doing [12:07] should we not just disable the mixer applet in gnome-applets? [12:07] well, if the pulseaudio notification icon is the new thing, then we should use it [12:07] pitti, the switch to the new capplet in karmic renewed the oem complains about lack of feature in this version [12:07] and disable the old applet by default [12:07] pitti, so I'm trying to not run into things which will make a potential rollback harder [12:07] I see [12:08] (that's a separate issue, though, isn't it?) [12:08] I'm still not sure how much oem need things which are not working [12:08] should we not try to get these missing features in to the new applet? [12:08] and what we can do for karmic [12:08] rather than shipping 2 [12:08] chrisccoulson, nobody said we should ship 2, I'm just trying to not wipe the applet from configs if we are going to roll back [12:09] but both applets bring back the very same dialog? [12:09] yes, the capplet is orthogonal [12:10] it's just than we can roll back to the previous capplet quite easily [12:10] but once the applet has been cleaned from the user config it will not be added again automatically [12:10] why is it a notification icon now? [12:10] isn't an applet better? [12:10] james_w, I expect because the redhat guys don't have applet in their gnome-shell world [12:10] s/better/preferable/ [12:11] hmm [12:12] seb128: right, but the upgrade issue is a bit less distracting than new installs/live systems [12:12] anyway, I'll release-note it [12:12] pitti, I think you are making a bigger deal of that that it is, ie it's mainly visual glitch [12:12] which is expected from an alpha version anyway [12:12] we had the issue for several jaunty alphas too [12:12] right, but I expect duplicates to be filed [12:12] so I'd like to point it out as a known issue [12:13] right, makes sense [12:13] given that they launch the same dialog, why not just use the applet anyway? [12:13] I guess it's unsupported now? [12:16] james_w, well the applet doesn't control pulseaudio directly I think [12:16] james_w, my understanding is that alsa mixer should be set to max and pulseaudio used for control [12:16] ok [12:16] having a mixer changing alsa volume break that model [12:23] ok, I applied an apport workaround to the i386 karmic fakechroot, testing [12:28] pitti, you are a superstar ;-) [12:28] meh [12:28] seems I'm not :/ [12:28] bug 402972 [12:28] no bot [12:29] crappy retrace again [12:29] oh, wait [12:30] no nautilus dbgsym for 2.27.4 [12:30] we lose [12:30] ? [12:30] how come [12:31] * pitti wants soyuz ddeb support [12:32] pitti, they are on http://ddebs.ubuntu.com/pool/main/n/nautilus/ weird [12:33] so index generation failed; I'll trigger a rebuild [12:33] oh, not the ubuntu2 version [12:33] but I'll apply/test the workaround for amd64 first [12:34] yeah, -ubuntu2 dbgsym failed for i386, lpia and powerpc [12:34] (armel too) [12:35] meh, not my day today [12:35] didrocks, how "failed"? [12:35] failing desktop CD builds, failing ddebs, failing fakechroot [12:35] time for holidays [12:35] I just looked at ddebs. I'm heading to LP [12:35] * seb128 hugs didrocks [12:35] didrocks, right, I figured so far that they were not there ;-) [12:36] dpkg-deb: building package `nautilus-dbgsym' in `../nautilus-dbgsym_2.27.4-0ubuntu2_i386.ddeb'. [12:36] seb128: oh sorry, read too fast and missed your " oh, not the ubuntu2 version" [12:36] pitti, can you check if those have been stored where they should but just not moved? [12:37] pitti, where I those stored and where is the code running again? [12:37] I know I had a look during some distro sprint but didn't take notes [12:38] seb128: macaroni.ubuntu.com, ubuntu-archive, see crontab -l [12:38] * pitti runs for d in 20090716 20090717 20090718 20090719 20090720 20090721 20090722; do ~/ddeb-retriever/ddeb-retriever $d; done [12:38] pitti, thanks, looking [12:38] ah that's it [12:38] this should re-fetch all the ddebs from the last week [12:38] previous time lp pages changed and parsing broke [12:39] I remember now [12:39] seb128: https://bugs.edge.launchpad.net/ubuntu/+source/gnome-media/+bug/402827 [12:39] Launchpad bug 402827 in gnome-media "gnome-volume-control crashed with SIGSEGV in IA__g_closure_invoke()" [Medium,New] [12:39] seb128: it's not really nice, but at least it somewhat works again [12:39] pitti, see you are a rockstart I knew it ;-) [12:40] all the gstreamer packages are outdated, but that shouldn't really make a difference? [12:40] seb128: well I think I leave it running for a while again [12:40] seb128: please stop them again if they still wreak havoc [12:40] pitti, no it shouldn't [12:40] pitti, ok [12:41] pitti, do you have magic to retrace everything which failed this week? [12:41] no, that requires log parsing, collect the bug numbers, and using ./mass-tag.py [12:41] pitti, will do that [12:42] * pitti hugs seb128 [12:42] * seb128 hugs pitti back [12:42] thanks; I need to get back to unf*** CD builds, I'm afraid [12:42] Keybuk, hey [12:42] Keybuk, could you do a bootchart with /usr/lib/nautilus moved away to compare with it being there? [12:43] Keybuk, just to give us a clue about how much time in nautilus startup is being spent there [12:43] by local testing it seems to take 3 seconds on a 8 seconds startup with empty cache [12:46] http://ddebs.ubuntu.com/pool/main/n/nautilus/ has the current ddebs now [12:47] regenerating package indexes now [12:48] pitti, \o/ [12:49] another mysterious failure there but if that works now... [12:49] well, let's see for how long [12:49] I wish that apt-ftparchive wouldn't need a million years [12:49] then I could run it more often and avoid dropping packages === MacSlow is now known as MacSlow|lunch [13:19] seb128: can I do it once I have the SSD-based Mini? that's the "official" benchmark [13:19] Keybuk, yes sure, thanks [13:20] I've just started looking at what is taking time nautilus start by using federico's printf and strace way [13:31] pitti: it takes a million years even with the db cache? [13:33] yes === MacSlow|lunch is now known as MacSlow [14:11] pitti, hum, we install contact-lookup-applet by default, I think we can stop doing that === rickspencer3-afk is now known as rickspencer3 [14:25] hey rickspencer3 [14:27] hi seb128 [14:38] mvo_, did you start working on porting synaptic to gtkbuilder? [14:38] hey rickspencer3 [14:38] seb128: yes, but just briefly, the glade code there is pretty old, I got a lot of error from glade-3 [14:38] urg [14:38] seb128: then I got a bit anoyed and did something else :) [14:38] mvo_, how likely is it for karmic? [14:39] mvo_, I'm just updating the GNOME3 spec [14:39] seb128: 50/50 - I would like to do it, but my workload is relatively big currently [14:39] mvo_, ideally we want everything using gtkbuilder before ll [14:39] seb128: ll ? [14:39] lazy lizard? [14:39] karmic+1 [14:39] yeah ;-) [14:40] ie not having to use libglade in the default lts install would be cool [14:40] seb128: oh? it's not part of gnome any more? [14:41] pitti, gtkbuilder is what should be used now [14:41] seb128: I meant contact-lookup-applet [14:41] pitti, oh, it has never been [14:41] seb128: ok, I will bump priority, is there a list of stuff that needs to be ported? [14:42] mvo_, don't, we have a good 8 months for it [14:42] pitti, c-l-a didn't change since gutsy [14:43] pitti, it gives an indication of how active upstream is ;-) [14:43] and I don't know of anybody using it, the bug list also doesn't suggest it's a lot [14:45] ok [14:45] seb128: so, it saves us 29 kB :) [14:45] lol [14:46] well we can let if you want [14:46] it's just on the "still use libglade" list [14:46] and I was surprised to see we have it on the CD [14:46] seb128: oh, let's kick it for uncruftification's sake [14:46] * pitti unseeds [14:47] seb128: done [14:47] pitti, danke [14:47] will be gone after alpha-3 [14:48] seb128: out of interest, how many libglade and libgnome rdeps do we still have on the default install now? [14:48] (roughly) [14:49] pitti, mvo: https://blueprints.edge.launchpad.net/ubuntu/+spec/desktop-karmic-gnome-3 [14:49] bottom of the whiteboard has the alpha3 current list [14:49] "libglade default installation:" [14:50] ah, thanks [14:51] pitti, so after next GNOME updates round (ie next week) we should be around 7-8 of those to tackle === andreasn_ is now known as andreasn [15:00] pitti, so when not building the applet you get a message telling you that the mixer is not available in new GNOME at next login [15:00] do we want that? [15:00] for upgrades? [15:00] yes [15:01] it's a wart, but we can't leave it there forever [15:01] seb128: but why wouldn't we build the applet? [15:01] No, we don't want that. [15:01] I thought we just want to disable it by default? [15:01] The warning is not okay. [15:01] pitti, how do you do that? [15:01] tedg, you would just silently clean the config? [15:02] seb128: oh, I thought it'd be a gconf schema change or so [15:02] seb128: (at least for new installs) [15:02] seb128: Either that or create a "NULL" mixer applet. [15:02] tedg, well that's what the null applet do basically [15:02] the null applet is what asks you if you want to remove it from your config i think [15:02] not sure why they decided to display a message now [15:02] seb128: Not cause an error on login ;) [15:03] it would be easy to patch the null applet to remove from the config without asking [15:03] it used to just clean the config I think [15:03] tedg, it's not an error it's an information dialog telling you what is happening to your applet [15:03] chrisccoulson, right, which is basically what I was asking, if we want to do that [15:03] The problem is people who have home directories that are on different releases of Ubuntu. [15:03] seb128: It reads like an error. I wouldn't say it's informational at all. [15:04] tedg, have you seen the dialog? [15:04] well, it's confusing, since from the perspective of the user you still have it [15:04] Yes, remember, I write all these little applets. My panel config looks like crap. I get all kinds of fun errors :) [15:04] (yes, it's the pulse notification thing, but looks all the same) [15:05] pitti, right, what I though [15:12] what pulls in gnome-media on xubuntu? [15:12] (sorry, i can't check right now at work) [15:13] xubuntu-desktop -> xfce4-xfapplet-plugin -> gnome-applets -> gnome-media [15:13] perhaps? [15:13] oops, no, that's just a suggests [15:14] it seems xubuntu users are getting gnome-media, which has gnome-volume-control-applet, which goes crazy when pulaeaudio is not available [15:14] we should probably patch the desktop file to only start in gnome [15:16] chrisccoulson, or add a recommends on pulseaudio [15:16] since the mixer applet and dialog depends on it now [15:17] seb128 - we could do i suppose. but the dependency on libpulse should be enough shouldn't it? or do we add a PA dependency for every application that uses it? [15:17] or should libpulse depend on PA? [15:18] chrisccoulson, the fact that you use libpulse doesn't mean you depends on pulse [15:19] ie the old capplet was able to use gstreamer or alsa [15:19] yeah, thats true actually [15:19] but the new codebase apparently does require pulseaudio to be running to be useful [15:20] well, it requires it to be installed. it doesn't need to be running, as it will automatically start it if it isn't [15:20] but that's the issue with the high CPU usage [15:20] right [15:23] i did suggest a fix to upstream to lower the CPU usage, but i don't know if they'll consider it [15:28] charlie-tca - how does gnome-media get installed on your xubuntu machine (what depends on it)? [15:28] I have no idea. Does alsa install it? [15:29] i doubt it [15:30] aptitude why gnome-media [15:31] metacity, gnome-session [15:31] gdm [15:31] Laney - thanks - i didn't know that [15:31] tis useful sometimes [15:31] there's also whynot [15:31] charlie-tca - thanks. that makes sense [15:32] You are welcome [15:32] Thanks, Laney [15:32] it seems xubuntu currently uses half of gnome just for the login screen now [15:32] Yeah [15:32] I know... === SiDi_ is now known as SiDi [15:39] pitti, bug #401294 fix uploaded for new installs [15:39] Launchpad bug 401294 in gnome-applets "sound applet and notification icon create duplication" [High,Confirmed] https://launchpad.net/bugs/401294 [15:39] (ie removed the mixer applet from the default profile) [15:39] seb128: yay, thanks [15:40] seb128: should we keep the bug open for upgrades, though? [15:40] pitti, the gnome-applets task is still open for that [15:40] ah, I see [15:40] pitti, I've the change to no build the applet I just need to silent the null applet cleaning the config on upgrade [15:43] seb128: I need to rebuild the ubuntu desktops anyway (ecryptfs), so that might just make it [15:47] pitti, I think retracers are still broken, should I stop those? [15:48] seb128: *sigh* please [15:48] pitti, they find the binaries now but not the debug symbol I think [15:48] pitti, ie http://launchpadlibrarian.net/29390644/Stacktrace.txt [15:48] for bug #402673 [15:48] Launchpad bug 402673 in notify-osd "notify-osd crashed with SIGSEGV in tile_destroy()" [Medium,New] https://launchpad.net/bugs/402673 [15:49] seb128: so it seems it finds the executable file again, but not the debug symbols somehow [15:52] pitti, right, similar to this morning when doing "exec-file usr/bin/binary" [15:53] pedro_, fyi retracers are broken so don't touch crash bugs for now, ie stop cleaning crashdump we might be able to retrace later [15:54] seb128, got it, may you please ping me when they start working again? [15:54] pedro_, will do! [15:54] seb128, thanks pal ;-) [15:54] you're welcome ;-) [16:17] pitti, we should install gnome-disk-utility by default btw [16:17] pitti, it has been accepted as official GNOME 2.28 component [16:18] seb128: ack [16:19] asac, gnome-bluetooth and webkit have been accepted for 2.28 [16:19] seb128: please seed it (I'm @ phone) [16:20] pitti, if you can do it quickly after your phone call please do, I will have a look otherwise but I didn't touch any seed for ages so I need to read documentation on where to get those again, etc [16:20] seb128: ok, will do [16:20] danke [16:34] pitti, gnome-applets updated too to not build the mixer applet and clean config on upgrade [16:35] seb128: wow, you rock! [16:35] ;-) [16:39] seb128: nice, the patch was just commenting out code in null-applet? interesting [16:39] pitti, well the null applet is made to clean applets not built on upgrade, GNOME is just displaying annoying dialog at next login which I commented [16:40] going to open a bug upstream for discussion now [16:40] ah [16:40] seb128: so that would affect other removed applets as well then? [16:41] pitti, right, other applets in the list being inbox and cd player [16:41] dunno what inbox is and cd player has no been built for years [16:41] ah, it has an explicit list [16:41] that sounds fine [16:41] yes, it's in the null-applet.c [16:44] i'm wondering if some of the other less useful applets should be removed? [16:44] ie, battstatus, which uses a deprecated /proc/acpi interface, or HAL [16:46] and modemlights seems to require network-admin, which is not installed by default anyway === rickspencer3_ is now known as rickspencer3 [17:28] seb128: ok. where is the announce? [17:28] * pitti purges devicekit [17:28] Keybuk: \o/ [17:29] asac, http://mail.gnome.org/archives/devel-announce-list/2009-July/msg00005.html [17:29] asac, http://mail.gnome.org/archives/devel-announce-list/2009-July/msg00005.html [17:29] ups [17:30] pitti: you're updating DK-power before α3? [17:30] Keybuk: no, but locally [17:30] while I wait for infinity unbreaking live CD builds [17:30] ah [17:31] Keybuk: I added a Conflicts/Replaces devicekit to dk-p for cleanup love [17:31] is anyone familiar with gstreamer here? [17:32] it seems that the default installation of ubuntu has the alsa plugin used instead of the pulseaudio plugin. [17:32] this was a non-issue before, but now that we've got the volume control applet which integrates into pulseaudio, you get a lot of annoying output [17:32] in the case of banshee, it lists "mono" as the application, rather than showing a nice Banshee icon [17:33] hyperair, slomo [17:34] ooh he's around today! [17:34] at least, his irc client is. =\ === ember_ is now known as ember [17:52] bryce: ping [18:02] seb128: I just discussed some gnome-power-manager patches with hughsie and got a +1 for committing them; do you think you can be my commit bot for them? [18:03] seb128: he said "just go ahead and commit trivial fixes, just consult me for UI/policy changes" [18:03] pitti, seems a good opportunity for you to ask for commit rights to GNOME? ;-) [18:03] heh, seems I should [18:03] seb128: I'm currently doing a g-p-m patch review (while waiting for CD builds) [18:03] you can probably have davidz vouching for you too [18:04] seb128: I'll look at the procedure next week then [18:05] pitti, http://live.gnome.org/NewAccounts [18:05] seb128: can I throw three format-patches at you, and you commit them? [18:05] pitti, yes [18:05] tab opened, thanks [18:05] doing them then [18:06] pitti, I will go for dinner now but will commit that after dinner [18:26] seb128: http://people.canonical.com/~pitti//tmp/gpm/ [18:39] tedg: is anything actually using gnome-power-manager-inhibit ? [18:40] and what is it for, exactly? [18:40] pitti: I do :) [18:41] pitti: http://gould.cx/ted/blog/A_small_utility_you_probably_don_t_know [18:41] tedg: ah, so it's not something other distro parts call? [18:41] tedg: ok, thanks [18:41] It came from a UDS session where people said that we couldn't do sleep by default because of command line apps. [18:42] ah, thanks [18:49] seb128 - i've pushed the g-s-d changes now [18:53] anyone know where i can find Jens Granseuer and what his IRC handle is? i can't see him on any channels i'm on, unless he has a really obscure IRC name [19:01] Heya do you remember the discussion about pulseaudio+skype? [19:01] there's a simple workaround which may be perhaps powered-up [19:02] using "pavucontrol" one can go to the configuration tab, turn off the input and output of pulse, make a skype call, turn on input and output again and enjoy banshee again. [19:02] the same for jack and thus ardour [19:03] perhaps a simple control for this could be put in the default desktop, dunno, I tell you, you do whatever you please :) [19:06] interesting eh. [19:06] well skype + pulse worked nicely for me when i was in x86 [19:06] but i'm getting a hell lot of underruns in x64. [19:07] i'm not sure whether it's something that changed in karmic or whether it's to do with x64. [19:07] the last time i used x86 was in intrepid. [19:09] I constantly get underruns with pulseaudio+skype, but using alsa it works. As it's a buggy closed application, having it working is enough. [19:09] But the problem is there also for jack apps so it's worth thinking about it [19:09] @_@ [19:09] stupid skype. [19:10] if only ekiga worked properly through my uni's strange firewall [19:10] then i'd install it at home and use ekiga instead. [19:11] no wait, ekiga has underruns too. [19:12] i'm beginning to think it's the pulseaudio alsa plugin [19:12] rather than all the other apps [19:12] hell, even sdl can't access pulseaudio thorugh alsa [19:13] ekiga works for me :) but I can't find a sip provider having a flat offer for the whole europe - but we are getting a bit offtopic [19:13] and plus, ekiga is going to be replaced by empathy I suppose. And empathy-sip does not work at all for me. [19:14] hyperair you know the drill: report the bugs! :) [19:17] hmm i haven't tried empathy sip [19:25] bye for now [19:27] man, my day was going so smooth until about 1.5 hours ago, suddenly I was getting pinged from all directions! [19:27] [19:27] * pitti stomps on rickspencer3's DSL cable [19:30] * bryce pings rickspencer3 [19:30] (kidding) [19:30] heh [19:30] * rickspencer3 pongs in random directions [19:31] I've been pointing people at http://err.no/personal/blog/tech/2006-10-10-12-05_contentless_pings.html - I don't mind getting pinged a lot, it's just maddening when they don't say what they want :-) [19:32] ping bryce [19:32] * chrisccoulson runs [19:33] bryce: in terms of "contentless pings" it seems cultural [19:33] some channels seem to prefer it, and some channels it pisses people off [19:34] hey rickspencer3, ping? ;-) [19:34] chrisccoulson, he's on #control-center on irc.gnome.org usually but seems busy recently [19:35] chrisccoulson, ie not often on irc [19:35] pitti, ok thanks [19:35] thanks seb128. i have another patch for him [19:37] « stop ping, make hugs » [19:37] * seb128 hugs didrocks [19:37] * didrocks hugs seb128 back [19:37] * chrisccoulson hugs didrocks [19:38] * didrocks hugs chrisccoulson too ;) [19:59] * rickspencer3 gets set to install alpha 3 on my netbook - iso testing [20:06] chrisccoulson, re, cf cleaning other old applets I think it's a good idea indeed [20:06] chrisccoulson, could you email ubuntu-desktop@lists.ubuntu.com with a suggestion? [20:07] seb128 - yeah, i can do. i think it's only the 2 applets i've already mentioned so far [20:09] anyone getting segfaults with gksudo? [20:11] james_w, no [20:14] just me then :-) [20:15] I'll try and debug later [20:16] james_w: again? :-) [20:16] one broken sudo per day is clearly enough [20:16] heh [20:22] seb128: the current desktop CD is a work of perfection [20:22] what should I complain about now? [20:22] ;-) [20:22] gvfs crashes when connection to a ssh location there [20:22] do you get that issue too? [20:23] Mount(0): SFTP auf piware.de -> sftp://piware.de/ [20:23] Type: GDaemonMount [20:23] hum, k [20:23] I opened my server bookmark and have nautilus [20:23] how to reproduce now? [20:23] well it works on my install but crash on current iso in kvm [20:23] oh, trying [20:24] just go on an empty workspace, ctrl-L, ssh://someip [20:26] indeed, I get that too [20:35] pitti: ping [20:35] j/k [20:36] seriously, looks like the iso tracker is getting ticked off [20:36] * pitti sends some flowers to rickspencer3 [20:36] but need some testing for amd64 and server? [20:38] I pinged the server guys (see #u-devel), they are on it [20:38] kewl [20:39] unr seems to be working fine, and someone else tested it while I was installing :) [20:41] nice [20:42] kubuntu netbook seems to be in some trouble, but by and large I hope that the images work now [20:43] ok [20:43] I'll try when I get back ... off to a lesson [20:44] * rickspencer3 downloads kubuntu netbook iso [20:46] * didrocks is going to get some rest now. Have a good evening/day everyone. [22:05] good night everyone [22:07] pitti, bug #403223 is the gvfs issue [22:07] Launchpad bug 403223 in gvfs "gvfsd-sftp crashed with SIGSEGV in strlen()" [High,Triaged] https://launchpad.net/bugs/403223 [22:07] pitti, it crashes when the host is not known and it should display a dialog about it [22:07] pitti, 'night [22:13] good night pitti [22:13] (few minutes too late ;)) [22:26] * chrisccoulson is not looking forward to another early start tomorrow [22:27] chrisccoulson, do you get a free day in exchange at least? [22:29] seb128 - hopefully i'll finish before lunch tomorrow, if everything goes well [22:30] but yeah, i will probably get some time off somewhere [23:02] bleah, launchpad's out of service [23:15] well not out of service just read only.. [23:15] for an hour. [23:17] read only, except that you can't view bugs which is slightly annoying [23:18] yes I hear you, but upgrade is good. [23:20] upgrades are fine, I just wish they didn't _always_ have to happen in the middle of my workday :-/ Esp. on the eve of an alpha release. [23:20] : -/ best done at night early morn.. [23:20] ok I'm done venting :-) [23:21] bryce it is good to vent sometimes. [23:22] I am glad they have it available as readonly now. But launchpadlib scripts still seem to get borked, unfortunately [23:22] : HTTP Error 503: Service Unavailable [23:25] bryce: hi [23:26] bryce: how possible do you think is it to use hw accelleration in the new gdm? [23:26] oh man [23:27] good morning robert_ancell [23:27] rickspencer3, hi rick [23:27] hi robert_ancell, rickspencer3 [23:28] mat_t, depends [23:28] * bryce waves to robert_ancell [23:28] rickspencer3: did "oh man" mean "it's way too late to think about it now dude"? ;) [23:28] hey bryce [23:28] bryce: go on [23:29] mat_t: it just makes me worry about what you have in mind [23:29] :) [23:29] :D [23:29] * rickspencer3 is curious to hear [23:29] basically I'm wondering if we could use compiz as w-m [23:30] with all compiz effects normally available [23:30] mat_t, in fact I recall some discussions early on about this [23:30] eh ... we can't reliably run compiz on the desktop after log in [23:31] mat_t, but keep in mind compiz doesn't work for everyone [23:31] * rickspencer3 thinks there are still many blacklisted cards [23:31] right, sure [23:31] no, not many blacklisted cards anymore, just bugs here and there [23:31] sweet [23:31] robert_ancell: can probably speak to that ;) [23:31] rickspencer3, I don't think there are any blacklisted cards in Karmic as I understand how the blacklisting works [23:32] that's great! [23:32] in fact I think with either the current or next compiz update we'll be dropping the last two card blacklists [23:32] or that's the start of many bugs :) [23:32] bryce, that's already been released to Karmic [23:32] it was just a couple 8xx cards [23:33] If you're running the latest Karmic then look at /usr/bin/compiz-manager, BLACKLIST_PCIIDS is empty and WHITELIST contains all the drivers [23:33] anyway, sorry to derail your question mat_t [23:33] bryce: so we could have a compiz-run option for most users and a non-composited fallback for example. [23:33] I'm not really trying to avoid it [23:34] rickspencer3: not to worry, that's why I'm here - to discuss it :) [23:34] mat_t, well all of these are really !xorg questions to be honest [23:34] right [23:34] yes, xorg can do it, the question is more of do you really want it ;-) [23:35] does it not take ages to load compiz? [23:35] and then it's just a matter of wrestling gdm and compiz into compliance. Nothing would need altered from xorg's perspective [23:35] rickspencer3, yep [23:35] but we do load it at some point anyway [23:35] mat_t, the other angle that I've seen discussed is using the compositing functionality in metacity [23:35] right [23:36] which gives you less bling but (in theory) better performance and stability [23:36] basically, I'd like at least for the gdm window to fade in and out [23:36] mat_t: does changing the oppacity on a gtk window approach the effect you would like? [23:37] rickspencer3: that could do the job [23:38] robert_ancell: would that have a hope of working? [23:38] the opacity? [23:38] yes [23:38] right, like put a timer on that slowly changes the opacity [23:39] (I have no idea if this works in gtk, but does in dhtml ;) ) [23:39] :) [23:39] I have never used it but I'd expect so. Unless you are doing any complex effects standard GTK+ and metacity should be able to do a simple fade [23:40] that would be cool [23:40] * rickspencer3 perhaps I could use quickly to quickly test that out? [23:40] robert_ancell: btw, did you have a chance to look at the comments I left under the gdm-config spec? [23:41] mat_t, I didn't see them, will have a look this morning when LP is back up [23:43] robert_ancell: so the plan is to access the config directly from gdm (not sure if that's a new decision though) [23:44] I think that's the best way of doing it - the gdm daemon already has a dbus interface to the config [23:45] robert_ancell: also, the recommendation was not to have the bg editable by the user, as it could cause "sudo-user conflict", basically the login screen should remain a "public space" [23:47] mat_t, not sure what you mean about sudo-user conflict? [23:48] robert_ancell: so it's easy to imagine a situation where two or more users share a machine, and one puts something in the login screen that others don't like :) [23:49] mat_t, sure - then how do we configure the background? It could have different PK privileges [23:49] robert_ancell: I'd stick with one, neutral background [23:50] mat_t, I can just see the endless bug reports now... "Unable to configure login background" :) [23:50] robert_ancell: why? [23:51] mat_t, just because it can't be done! [23:51] oh I see what you mean [23:51] :) [23:52] was the bg easily configurable in the old gdm? [23:52] Well, in saying that we just need to point them to a forum post saying how to edit the config file [23:52] right [23:53] I'd support not having it easily user configurable - there's no sense in changing the background if you can't change the rest of the theme [23:53] exactly [23:54] I thought gdm got its background from the GNOME wallpaper setting, which would be for the gdm user [23:54] so that should be changable [23:54] robert_ancell: the time difference is crazy between us, it's 12am here :) I've got to head to bed... [23:55] mat_t, it sure is! Thanks for the feedback, see you later [23:56] TheMuso, can you log in as the gdm user? [23:56] np, take care - drop me an email if you need anything! [23:59] 'night everyone