[00:00] hope you guys are having a nice let winter morning there [00:01] here's the notes from the team meeting this morning: [00:01] https://wiki.ubuntu.com/DesktopTeam/Meeting/2009-08-25 [00:01] rickspencer3, there is no such thing as WInter in Sydney :) [00:01] hehe [00:01] the days aren't shorter? [00:02] Although there is such ting as a windy winter on the mountains at least. [00:02] Been blowing a gail since yesterday morning. [00:02] hiya TheMuso [00:02] TheMuso, how long does it take to drive to Sydney from where you live? [00:03] rickspencer3: About an hour or so, thats to the inner west of Sydney. [00:04] ah [00:04] so, you guys wanna do a quick Team Meeting, Easter edition? [00:05] Eastern Edition, even? [00:05] Sure. [00:05] the heart of the meeting regard feature freeze [00:05] https://wiki.ubuntu.com/DesktopTeam/Meeting/2009-08-25 [00:05] Only thing I have to add re gnome-speech replacement is that the final MIR for that will be filed today, and once everything is main, I just have to switch a few things around in terms of package dependencies, to get orca using speech-dispatcher by default. [00:05] that's all I had to ask :) [00:06] TheMuso, do you see any risk there? [00:06] rickspencer3: No. [00:06] like, any chance the MIR team won't get to it in time for Feature Freeze? [00:06] sweet [00:06] and afaik MIRs hav ebeen approved post feature freeze in the past, so I assumed that was the same thing here, but thing may have changed and I didn't know about it. [00:07] robert_ancell, you had gdm-config [00:07] gah typing [00:07] TheMuso, no problems [00:07] I'm sure all if find [00:07] rickspencer3, is there some paperwork I need to do for that? [00:07] robert_ancell, nope [00:07] it's all done [00:08] I set the blueprint to "Implemented" for you [00:08] pitti looks across the blueprints, and checks on the status [00:08] cool [00:08] hggdh: pong [00:08] the other agenda item as the burndown chart [00:09] next week I will be working with each team member to savagely postpone work items to get us below the trend line [00:09] I think asac and kenvandine will be mostly impacted by this, though seb128 and robert_ancell as well [00:09] Right, I think I only ahve a couple left, one of which is post feature freeze safe. [00:09] the other will be done by feature freeze, as its to do with MIRs. [00:09] sweet [00:10] any other business? [00:10] nope, just a stack of packages to work through :) [00:10] k ... I'll let you get to it [00:11] thank you gentlemen .. on to Feature Freeze! [00:11] :) [00:11] Indeed. [00:15] jcastro, sorry, already found the Way -- I am setting up a "Triaging Sound Bugs" class, and wanted to know do to get the -classroom for it [00:16] * jcastro nods [00:16] * jcastro cheers! [00:16] jcastro, of course, I did the unthinkable: searched the wiki, and got the answer... thank you, anyway [00:23] robert_ancell, any idea debsign can't find my secret key, consider I *just* created one? [00:23] could it have a pointer to the one that I deleted cached somewhere? [00:24] rickspencer3, does the email on the key match DEBEMAIL? [00:24] it matches MAIL in any case [00:24] rickspencer3, not sure sorry [00:24] this worked before I deleted my old keys and created new ones [00:24] you should set DEBEMAIL and DEBSIGN_KEYID [00:25] cat ~/.devscripts [00:25] if that gives an id then you need to change it [00:25] james_w, I have no .devscripts file [00:25] otherwise see https://wiki.ubuntu.com/DeveloperGuide/Uploading for how to work out how it is searching for the key based on the changelog [00:27] hmm [00:27] seems to be that I added a comment to the new key [00:28] it's always the comments :-) [00:28] ug [00:28] you can either add that comment to your changelog entries [00:28] well .. I want this to work with quickly ... [00:28] or one of the other steps there [00:28] so I'll create a new key without comments [00:28] add a uid to your existing key [00:35] robert_ancell: hey [00:35] huats, hey [00:35] please do the gcalctool update :) [00:35] I haven't found the time to do the 27.90 [00:35] but go ahead for the .91 [00:35] ) [00:35] huats, Should be an easy one - it's just translations this time [00:35] huats, ok, will do ti [00:35] iot [00:35] it [00:35] and I hope you enjoyed your holidays [00:35] i can spell good [00:36] :) [00:36] huats, yeah it was a lot of fun (spent two days in Disneyland being a big kid) [00:36] hello rickspencer3 and james_w btw :) [00:36] hehe looks nice [00:36] salut huats [00:36] comment ça va james_w ? [00:36] :) [00:37] hi hauts [00:37] tres bien merçi, et toi? [00:37] tres bien aussi [00:37] rickspencer3: you should learn to speak french [00:38] soon this channel will be in that language :) [00:38] we need a google translate bot [00:38] :) [00:40] hauts: I start French 101 on September 14th [00:40] I joined alliance France Seattle [00:40] :) [00:41] so soon [00:41] F* it [00:41] I'm just starting over with a new key and no comment [00:41] * rickspencer3 puts pulled hair in garbage can [00:42] great ! [00:42] huats, got my account on ubuntufr, thanks! [00:43] do not hesitate to talk to us (I mean the french collony) [00:43] fta: good... [00:43] ;) [00:43] fta so now you can use it :) [00:43] i did [00:44] great ! [00:44] so thanks to you fta ! [00:53] yeah! [00:53] my first ppa upload ... for the least essential application ever! [00:57] yes! [00:58] quickly works! [01:09] From now on when I hear about quickly I'm going to think "squick" :/ [01:13] rickspencer3: shall we have a call? [01:14] ArneGoetje, hmmm === rickspencer3 is now known as rickspencer3-afk === Cuddles is now known as Richie [06:22] Good morning [06:23] robert_ancell: for bug 412455, you mean that gnome-panel and nautilus already send the d-bus signals? [06:23] Launchpad bug 412455 in xsplash "shut down xsplash when session is ready" [High,Fix released] https://launchpad.net/bugs/412455 [06:23] pitti, yes [06:23] robert_ancell: ah, ok; that was a misunderstanding then [06:23] * pitti updates bug [06:59] OMG gnomescan [06:59] * pitti posts feedback to blueprint whiteboard [07:03] kenvandine: ^ this makes me weep :( how badly do you want this? [08:22] morning o/ [08:25] hey didrocks === vuntz_ is now known as vuntz === tkamppeter_ is now known as tkamppeter [08:57] didrocks: hey! in the rotate (left, bottom) (my favourites), what about decreasing tab opacity when flipping them, so it doesn't end up obscuring the widgets that would be around ? [09:01] fredp: great idea, and easy to do as well, I like it :) I have also to figure out how to bind it more properly and try to link the upper tab to the animation as mpt_ suggested (not so easy without changing notebook gdk behavior) [09:03] anyway they are interesting experiments, I'll continue watching you :) [09:12] fredp: thanks :) === WelshDragon is now known as Richie === mpt__ is now known as mpt [10:01] mvo, I just replied to your questions about the Software Store spec [10:16] hello everyone [10:22] mpt: thanks [10:24] /c/c [10:35] hey glatzor [10:41] hello mpt! [10:53] glatzor, have you been following the SoftwareStore spec at all? [10:55] not closely. [10:55] mpt, I installed it yesterday and had a short look at it [10:56] I was impressed at how smoothly it works already [10:56] but I did notice that the PolicyKit dialog appeared *behind* the Software Store window [10:56] that is most likely a bug of mine [10:56] but I need to double check, it might not pass the xid of the parent [10:56] mvo, ok. :-) I remember we had discussed that issue before [11:14] mpt: I just looked at it and it seems to be a problem with compiz for policykit1 in general. its showing corectly in metacity [11:16] mvo, ok, can you tell whoever is responsible for fixing it that sabdfl will be quite annoyed if it isn't? ;-) [11:23] mpt: either me or robert [11:23] mvo, Ancell? [11:23] yes [11:23] ok [11:26] mvo: how do I work around bug 419152 ? [11:26] Launchpad bug 419152 in software-store "application does not start "No section: 'Desktop Entry'"" [Undecided,New] https://launchpad.net/bugs/419152 [11:29] Riddell: please try installing gnome-menus [11:32] is the system beep controled by gnome? or kernel? [11:32] Riddell: I added the missing dependency now, thanks for the bug [11:33] mvo: something else missing, bug 419154 [11:33] Launchpad bug 419154 in software-store "can not install software, NotAuthorized" [Undecided,New] https://launchpad.net/bugs/419154 [11:33] is that the lack of a polkit1 qt frontend? [11:33] mm, could be [11:34] do you have policykit-1-gnome installed? [11:34] yes [11:34] should work then [11:37] I'm not sure whether that's enough [11:37] pkaction --action-id org.debian.apt.install-packages --verbose [11:37] that might give a clue [11:37] is /etc/xdg/autostart/polkit-gnome-authentication-agent-1.desktop actually run under KDE? [11:38] * pitti doesn't know whether OnlyShowIn=GNOME; is considered for autostart [11:38] check pidof polkit-gnome-authentication-agent-1 [11:38] it is I think [11:38] that would do it [11:38] I don't have that file, is it new? [11:39] james_w: not particularly; it's been shipped in policykit-1-gnome since day one [11:39] i. e. only as new as pk-1 itself (a month or two) [11:40] 0.92-0ubuntu1 and I don't have it [11:40] odd [11:40] james_w: http://paste.ubuntu.com/259771/ [11:40] and update-manager isn't popping up to remind me to update [11:40] /etc/xdg/autostart/polkit-gnome-authentication-agent-1.desktop doesn't exist [11:41] hmm, policykit-1-gnome is out of date [11:41] james_w: hm, maybe it was in /usr/share/autostart/ before or so [11:41] that paste is ok, it should be prompting [11:41] dpkg -L policykit-1-gnome | grep desktop [11:41] so I suspect pitti is correct [11:41] /usr/share/gnome/autostart/polkit-gnome-authentication-agent-1.desktop [11:42] aah [11:42] so it just moved to /etc recently, apparently [11:42] but it's still gnome only [11:42] perhaps we should remove that restriction if there isn't going to be a qt frontend [11:42] huh? http://paste.ubuntu.com/259772/ [11:43] (by release) [11:43] a GNOME dialog is better than no dialog? [11:43] it's from /usr/lib/libpolkit-gobject-1.so [11:43] I would suspect version skew [11:43] we do hope to get polkit 1.0 for qt before release but would be sensible to remove that OnlyShowIn until that happens [11:44] well, without a real Qt implementation we should perhaps rather launch the frontend through kdesudo [11:44] that's what I did with jockey [11:44] I have no /usr/lib/libpolkit-gobject-1.so [11:44] otherwise you'd pull gtk and friends into kubuntun [11:44] Depends: libc6 (>= 2.2), libglib2.0-0 (>= 2.18.0), libgtk2.0-0 (>= 2.17.7), libpolkit-agent-1-0, libpolkit-gobject-1-0, policykit-1 [11:44] weird [11:45] it's shipped by libpolkit-gobject-1-0 [11:45] but it really ought to be a versioned dependency [11:45] bad shlibs for the win [11:47] ok updated libpolkit-agent-1-0 policykit-1 libpolkit-gobject-1-0 and now polkit-gnome-authentication-agent-1 starts [11:48] hmm, no it doesn't [11:49] Riddell: so if you actually prefer using the gnome PK frontend over running the app as root, should I change that in Jockey, too? [11:50] pitti: well jockey has a kde frontend so it's nice to keep it all KDE. software-store doesn't so that's not an issue [11:50] ah, ok [11:51] oh no polkitd has version issues now [11:56] ok I install libpolkit-backend-1-0 and now it all works [11:56] doesn't polkit to binary compatibility and versioned dependencies? should it really be in main? [11:57] mvo: no feedback during the install except that after 30 seconds the install button changed to a "remove" button [11:57] seems there are some depedency bugs; what did you try to install originally? [11:57] pitti: software-store [11:57] Riddell: policykit-1 depends on libpolkit-backend-1-0 [11:58] right, it misses shlibs or symbols files; I'll add some symbols [12:10] Riddell: no "pending " item on the left ? [12:11] mvo: nope [12:12] not that I noticed anyway === MacSlow is now known as MacSlow|lunch [13:11] pitti, i have very little invested in gnomescan... it wasn't really my idea [13:11] i just hate xsane [13:12] pitti, so i am fine keeping what we have :) [13:12] Does anybody know if UNR are doing anything in particular in relation to gdm theming? [13:12] TheMuso, i don't thinks so [13:13] TheMuso, actually pretty darn sure they aren't [13:13] kenvandine: ok [13:13] pitti, want to just kill it? [13:13] kenvandine: Right, so other than getting the default wallpaper that they use for the desktop, nothing else. [13:14] Hrm. We in Studio have the interesting problem where we want the wallpaper for gdm to be different to what is on our desktop. [13:15] * TheMuso will have to work something out... [13:20] TheMuso, yeah, that won't be easy [13:20] you will need to set the default wallpaper for the gdm user in gconf [13:20] i think [13:20] pitti, can you sponsor xsplash 418974 ? [13:21] kenvandine: This is what I have so far. Gdm sets a wallpaper filename/path in gconf on package install for the gdm user. THis filename is then changable via the alternatives system depending on what wallpaper you want. [13:21] However, this doesn't solve the problem of changing the actual window theme. [13:21] not changable, but directable to the file you want. [13:22] Or, we have all theme packages for gdm change the gconf settings... hrm. [13:25] kenvandine: looking [13:26] pitti, thx [13:27] pitti, there is also one more FF related package needing sponsoring [13:27] couchdb-glib 419160 [13:27] and evolution-couchdb will be coming in a bit... then it will be ready to go on the CD [13:28] assuming we get an agreement on erlang [13:28] it's obviously something we have to swallow, but it's really ugly [13:29] yeah [13:32] kenvandine: fixing Vcs-* again [13:34] kenvandine: your postinst needs love [13:34] oh? [13:34] kenvandine: you can't rely on the gdm user being present, since you don't depend on gdm [13:34] what is wrong with Vcs? === MacSlow|lunch is now known as MacSlow [13:34] kenvandine: Vcs-Bzr: needs to be a proper URL (https://code...) [13:34] kenvandine: and there's no Vcs-*-Browser:, just Vcs-Browser: [13:34] sigh... asac insists it shouldn't be [13:35] since for web browsing you don't need to know which particular VCS you use [13:35] true [13:35] kenvandine: right, but then you need to create the gdm system user if it doesn't exist [13:35] i was just following his example [13:35] pitti, ok so for 0.6 i will add to postinst [13:35] to create the user if it doesn't exist [13:35] why does it need to be setuid gdm? [13:35] doesn't gdm start it in the first place? [13:36] -rwxr-xr-x 1 root root 22368 2009-08-12 18:54 /usr/bin/xsplash [13:36] ^ currently [13:39] kenvandine: so with that, only root or gdm can run it, and it suids to gdm [13:39] that looks like a complicated no-op to me? [13:40] pitti, well gdm must start that stuff as root [13:40] or before switching to user gdm [13:40] xsplash? [13:40] yes [13:40] ugh, why does it need root privs? [13:41] kenvandine: so you mean root:gdm 4751 [13:41] it doesn't [13:41] we want it to not run as root [13:41] it is configurable.. originally it ran as root [13:41] but people complained :) [13:42] I thought the greeter would run as gdm [13:42] the greeter probably does [13:42] but i think the mechanism that fires the Init stuff and PreSession stuff is still root [13:42] kenvandine: ok, so you want to drop privileges [13:42] yes [13:42] kenvandine: then suid is not what you want [13:43] the first thing in your program should be [13:43] setuid(uid); [13:43] where uid is from getpwnam("gdm") [13:43] so if a normal user calls it, that will just fail [13:43] and if root calls it, you'll drop all your privileges [13:44] and avoid suid stuff and all the system user handling in postinst [13:44] bratsche, ^^ [13:44] or, rather, "gdm" -> the value from your configure option [13:45] it's a constant in config.h, I presume [13:45] Riddell: added shlibs to pk-1, thanks for spotting [13:45] morning tedg [13:47] kenvandine: summary posted to sponsoring bug [13:47] thx [13:51] kenvandine: couchdb-glib uploaded [13:52] pitti, thx [13:52] * kenvandine hopes evo couch comes soon [13:53] kenvandine: Good morning! [14:07] he, has some decision been made about upnp support in libnice and farsight2? [14:09] pitti, what do you think about that? [14:10] libupnp-igd is needed for that [14:10] yeah... adding more stuff to the CD [14:10] :) [14:10] :p [14:10] which makes pitti's head hurt [14:11] CD is already -3MB or so :P [14:11] Time to drop evolution? [14:11] * Amaranth runs [14:11] and we have more to go on it today :) [14:11] Amaranth, bah humbug :) [14:12] new f-spot will save some space [14:12] new f-spot is already on it [14:12] and we are over [14:12] (apparently) [14:12] never [14:12] newer* [14:12] yeah, f-spot saved 1MB or so [14:13] But the CD was -1MB yesterday when someone wanted to add 3MB to it so... [14:13] me :) [14:13] 13.9 KiB on i386 [14:13] we are +3M now on the amd64 [14:13] and we need to add like 4M [14:13] "# reduce the both the package size and installed size by a few Megs " [14:13] for 0.6.1.0 [14:14] is 0.6.1.0 out? [14:14] yes [14:14] * kenvandine will prepare that :) [14:14] i already started [14:14] cool! [14:14] that will help! [14:14] thanks Laney [14:14] pitti will at least have a smaller heart attack [14:14] well it remains to be seen what the actual benefit is [14:15] probably have it tonight [14:15] kenvandine: what is upnp, anyway? [14:16] media stuff, universal plug-n-play i think [14:16] it's everything [14:16] I'm afraid I don't have an opinion on it, except if you tell me it'll need > 1 MB, then I do have an opinion :-P [14:16] * kenvandine isn't into upnp [14:16] :) [14:16] pitti: Universal Plug'n'Pray :-) [14:16] punching holes in routers is all I know about it [14:16] I know the expansion of the acronym [14:16] pitti: standardized (aka MS) way for mapping ports (nat) on great public routers [14:16] pitti: it's like dynamic port forwarding AFAIK [14:16] UPnP is what a program can use to open a port in your router and what the xbox 360 and ps3 use to talk to PCs [14:16] so it's basically everything [14:16] but it doesn't tell me anything else than "quantum slip stream phase realignment" [14:17] why does 'apt-get purge' remove more packages than 'apt-get remove'? man page states the difference is that 'purge' also removes configuration files [14:17] it does file serving too somehow I think [14:17] pitti: no no. the quantum stuff have to do with beers I'm pretty sure. we're talking about software now :-) [14:17] with the telepathy stack it should permit the user to forget the nat config to make voip calls [14:17] pmatulis: because you "removed" stuff before that left configuration files. [14:17] IPv6!!! (*cough*) [14:17] bigon, that is interesting then [14:17] apt-get remove --purge actually :-) [14:18] pitti: *grin* nice troll! ;-) [14:18] * pitti just wishes people woudl stop inventing the 10.000st workaround of NAT [14:18] pitti: and I want a pony :) [14:18] Nafallo: I'm actually pretty serious [14:18] UPnP is actually the standard way invented so everyone would stop using those other ways [14:18] pitti: yeah. my troll comment was pretty much a troll :-) [14:18] upnp isn't just about nat :P [14:18] kenvandine: so if you think it's sane, and it doesn't need much space, sure [14:18] But it kind of morphed into other things as well so UPnP is a lot of things [14:19] * pitti hugs Nafallo [14:19] it's about controlling lights! [14:19] alex-weej: sorry but that doesn't make sense to me [14:19] Amaranth: I have a pony. at some point I should add IPv6 to it ;-) [14:19] pitti: +1 for ipv6 (btw all the tp stack should be ipv6 aware) [14:19] Amaranth: (pony.magicalforest.net fwiw) [14:20] pmatulis: look in synaptic at the "residual configuration" filter, and you will see a lot of packages that you thought you had removed [14:20] or even better... http://www.magicalforest.se/pony.xhtml [14:20] pitti, ok i will prepare an MIR [14:24] alex-weej: even if that's true (cannot yet find such a filter), why would removing one particular package remove these other packages? the purge should be applied to just that package [14:24] pitti, looks like it will add about 95k [14:25] pmatulis: that's just the logic of apt-get. try another package manager if you dislike that logic [14:25] pmatulis: aptitude may work differently [14:25] kenvandine: *phew* :) [14:25] looks like two MIRs though [14:25] alex-weej: apt-get and aptitude give same results [14:25] alex-weej: i'm just following the man pages [14:26] gssdp and gupnp [14:26] kenvandine: couchdb-glib FTBFSed; meh, I should have test-built it before.. [14:26] alex-weej: and what you're saying contradicts them. thanks anyways [14:26] pitti, what? [14:27] failed to build? [14:27] built for me last night [14:27] well not with his debdiff... [14:28] kenvandine: http://launchpadlibrarian.net/30853330/buildlog_ubuntu-karmic-i386.couchdb-glib_0.4.5-0ubuntu1_FAILEDTOBUILD.txt.gz [14:28] pmatulis: ? [14:28] kenvandine: the debdiff was just debian/changelog; some autotools goo [14:28] yeah [14:28] i know [14:28] it built in my ppa last night [14:29] it shouldn't regenerate the autotools files [14:29] probably fs specific timestamp issue again [14:31] kenvandine: I'll retry the build, let's see [14:32] yeah builds fine locally here [14:32] pitti, i have a bunch of daily builds setup in my daily ppa now [14:32] morning rickspencer3 [14:33] hey rickspencer3 [14:34] hi guys' [14:34] * rickspencer3 wipes sleep from eye [14:34] :) [14:35] kenvandine: nope, failed again [14:35] humm [14:47] pitti: kenvandine: sorry, my computer froze twice this morning [14:47]