SIGs/Desktop/Meeting-20071017

Desktop SIG Meeting of 2007-10-17

Present

• MatthiasClasen (mclasen)
• DavidZeuthen (davidz)
• RayStrode (halfline)
• Will Woods (wwoods)
• Colin Walters (walters_)
• Lennart Poettering (mezcalero)
• Adam Jackson (ajax)
• JesseKeating (f13)

Notes

• Ray wants to land new gdm in rawhide soon; needs newer dbus-glib; still pretty raw currently
• Revamped graphical boot is on the wishlist for F9 again, still depending on kernel modesetting
• David wants to eliminate pam_console
• Lennart will work on improved volume control, based on Pulse'Audio

Log

Oct 17 14:00:43 <mclasen>	ok, is it time for a desktop sig meeting ?
Oct 17 14:02:10 <mclasen>	sorry that I didn't have time to send out an agenda
Oct 17 14:04:06 <mclasen>	do we have things to discuss today ?
Oct 17 14:05:26 <davidz>	I think f8 is frozen already right?
Oct 17 14:05:34 <mclasen>	pretty much
Oct 17 14:06:34 <mclasen>	I didn't see an announcement yet, but the 16th was supposed to be the freeze date
Oct 17 14:06:52 <davidz>	so perhaps next meeting we should start talking about f9 features?
Oct 17 14:07:00 <davidz>	I mean, it's a bit premature perhaps to do it now
Oct 17 14:07:09 *	davidz , for one, would like to prepare and think about it
Oct 17 14:07:30 <mclasen>	no, I think it is fine
Oct 17 14:07:38 <mclasen>	I had that on the agenda for last week, actually
Oct 17 14:07:44 <mclasen>	but we didn't get around to it
Oct 17 14:08:37 <davidz>	so.. next week or this week?
Oct 17 14:08:43 *	davidz a bit confused, sorry
Oct 17 14:08:51 <mclasen>	I'm fine starting such a discussion today
Oct 17 14:09:01 <mclasen>	given that there is no other topic right now...
Oct 17 14:09:11 <davidz>	so
Oct 17 14:09:17 <davidz>	maybe IRC isn't the best forum
Oct 17 14:09:30 <davidz>	maybe asking contributors on the list to send proposals to the list... would be better?
Oct 17 14:09:50 <halfline>	well, do we have any F8 blockers?
Oct 17 14:10:07 <halfline>	if so, those should still get fixed, or demoted even after today, yea?
Oct 17 14:10:15 <mclasen>	sure
Oct 17 14:10:17 *	davidz all done with f8 blockers
Oct 17 14:10:40 <mclasen>	the freeze means that from now on, you are not supposed to build stuff unless it fixes a blocker or serious issue
Oct 17 14:10:56 <mclasen>	and I believe rel-eng will start selectively tagging things
Oct 17 14:11:02 <mclasen>	ie you need to ask for it to be moved
Oct 17 14:11:10 <mclasen>	f13: is that right ?
Oct 17 14:13:15 <mclasen>	halfline: do you want to discuss any f8 blocker bugs here, or should we start to fantasize about f9 ?
Oct 17 14:15:37 <mclasen>	davidz: I'd hope that we have a feature proposal/brainstorming phase for f9 on the lists as soon as f8 goes out
Oct 17 14:15:55 <mclasen>	but of course, we all have our own agendas already...
Oct 17 14:17:09 <halfline>	let's fantasize
Oct 17 14:17:32 *	walters_ feels like he walked in on something
Oct 17 14:17:35 <halfline>	so one thing i did was get gdm branched early for F9
Oct 17 14:17:48 <halfline>	i'd like to build development snapshots of gdm
Oct 17 14:18:01 <halfline>	we will need fork dbus-glib though and update it as well
Oct 17 14:18:11 <halfline>	since the new gdm requires a newer version of dbus-glib than we ship
Oct 17 14:18:23 <halfline>	davidz: is there any reason we didn't update dbus-glib in the F8 cycle?
Oct 17 14:18:33 <mclasen>	just sheer amount of rebuilds
Oct 17 14:18:34 <halfline>	does it change so names?
Oct 17 14:18:39 <mclasen>	thats what kept me from doing it after test2
Oct 17 14:19:06 <davidz>	halfline: feel free to update it
Oct 17 14:19:13 <davidz>	halfline: unless there's .so name breakage
Oct 17 14:19:18 <davidz>	the we have to coordinate the transition
Oct 17 14:19:27 <halfline>	right
Oct 17 14:19:28 <davidz>	s/the/then
Oct 17 14:19:35 <halfline>	i wouldn't mind forking that package early
Oct 17 14:19:42 <halfline>	but if there are 50 packages that depend on it
Oct 17 14:19:46 <halfline>	i don't really want to fork all of them
Oct 17 14:19:50 <mclasen>	when I was looking after test2, I found ~50 packages that depend on it
Oct 17 14:19:57 <halfline>	okay
Oct 17 14:20:02 <halfline>	so probably not a good idea then
Oct 17 14:20:03 <halfline>	well
Oct 17 14:20:09 <halfline>	actually maybe it doesn't matter
Oct 17 14:20:16 <mclasen>	of course, most of them probably don't really depend on it
Oct 17 14:20:29 <halfline>	we won't be getting f9 composes for some time
Oct 17 14:20:52 <halfline>	so until we do start getting them, it doesn't really matter if the so name breaks, yea?
Oct 17 14:21:23 <halfline>	it won't break anyone's yum update i mean.  it may break people's machines that are testing the new gdm
Oct 17 14:21:34 <mclasen>	no, it won't break peoples machines
Oct 17 14:21:45 <mclasen>	unless they are testing the new gdm
Oct 17 14:21:52 <halfline>	right
Oct 17 14:22:03 <mclasen>	I think doing the fork is probably fine
Oct 17 14:22:04 <halfline>	so it might be worth doing anyway
Oct 17 14:22:27 <halfline>	okay i'll look into it
Oct 17 14:22:41 <mclasen>	you can send out a heads-up to -devel list
Oct 17 14:22:50 <halfline>	yea sounds like a good idea
Oct 17 14:23:09 <mclasen>	then people can fork their packages at their own speed
Oct 17 14:23:14 <halfline>	alternatively, i could see why the new gdm needs the new dbus-glib and patch it so that it doesn't need it
Oct 17 14:23:26 <mclasen>	that would be an option, too
Oct 17 14:23:34 <mclasen>	worked for packagekit
Oct 17 14:23:49 <davidz>	so new dbus-glib changes soname?
Oct 17 14:23:52 <halfline>	ah good, that might be the nicer option
Oct 17 14:23:53 <davidz>	is that a fact?
Oct 17 14:23:58 <halfline>	i don't think we know
Oct 17 14:24:02 <halfline>	know one's checked.
Oct 17 14:24:04 <halfline>	it's your package :-)
Oct 17 14:24:13 <halfline>	s/know/no/
Oct 17 14:24:15 <davidz>	well
Oct 17 14:24:19 <davidz>	if there's no soname changes
Oct 17 14:24:22 <davidz>	then ...
Oct 17 14:24:25 <davidz>	we just update it
Oct 17 14:24:38 <halfline>	you mean for f8?
Oct 17 14:24:41 <davidz>	no, for f9
Oct 17 14:24:42 <wwoods>	most of the desktop-oriented stuff on the blocker list relates to keyring handling and NM
Oct 17 14:24:52 <davidz>	bit late in the game to update for f8 I think
Oct 17 14:25:00 <halfline>	yea
Oct 17 14:25:22 *	davidz curses fd.o gitweb
Oct 17 14:25:25 <davidz>	it's so bloody slow
Oct 17 14:25:50 *	davidz checks out the source instead
Oct 17 14:25:57 <halfline>	i think we're good
Oct 17 14:26:16 <wwoods>	oh, and media keys (e.g. thinkpad_acpi stuff, bug 321421 - but that's a complicated one)
Oct 17 14:26:25 <halfline>	LT_* are the same in the configure
Oct 17 14:26:25 <halfline>	for both
Oct 17 14:26:29 <davidz>	halfline, mclasen: there's an soname bump
Oct 17 14:26:40 <walters_>	hm, why?
Oct 17 14:26:50 <mclasen>	just because...
Oct 17 14:26:50 <halfline>	err
Oct 17 14:27:05 <halfline>	davidz: there is?
Oct 17 14:27:16 <halfline>	are we looking at the same source trees?
Oct 17 14:27:35 <davidz>	hmmmm
Oct 17 14:27:40 <mclasen>	anyway, lets not get bogged down in that soname issue
Oct 17 14:27:41 <davidz>	changed on 2007-02-13
Oct 17 14:27:50 <mclasen>	what are the improvements that we expect out of the new gdm ?
Oct 17 14:28:05 <mclasen>	actually, we should probably start to fill a wiki page with that information...
Oct 17 14:28:10 <mclasen>	to make poelstra happy
Oct 17 14:28:44 <halfline>	well, it's completely different
Oct 17 14:28:48 <halfline>	right now it's very raw
Oct 17 14:29:01 <halfline>	needs a lot of work just to get some parity with what were' currently shipping
Oct 17 14:29:07 <davidz>	halfline: no, I'm confused - will report back on the soname thing on #fedora-desktop later
Oct 17 14:29:14 <halfline>	davidz: ok
Oct 17 14:29:29 <halfline>	one thing it will offer is more of a session before you login
Oct 17 14:29:43 <halfline>	things like gnome-power-manager will be available at the login screen
Oct 17 14:29:52 <halfline>	so you can see your battery usage and suspend
Oct 17 14:30:09 <halfline>	the greeter is completely differnt
Oct 17 14:30:14 <halfline>	so existing themes probably won't work
Oct 17 14:30:26 <halfline>	unless we take some steps to make them work
Oct 17 14:30:53 <halfline>	there is also this idea of a factory greeter, which needs some investigation
Oct 17 14:30:58 <halfline>	the way that it works is, it's always present
Oct 17 14:31:07 <halfline>	and then the theory goes, anytime you need to lock the screen
Oct 17 14:31:11 <halfline>	we jump back to it
Oct 17 14:31:14 <halfline>	or fast user switch
Oct 17 14:31:54 <davidz>	well
Oct 17 14:31:57 <davidz>	we can't let gdm stay
Oct 17 14:32:01 <davidz>	until X is fixed
Oct 17 14:32:09 <davidz>	remember only one server can do DRI etc.
Oct 17 14:32:18 <davidz>	but I think we want g-p-m and n-m to run in gdm
Oct 17 14:32:20 <davidz>	as features
Oct 17 14:32:36 <davidz>	along with another feature in both g-p-m and n-m being a "Make these settings default" button
Oct 17 14:32:40 <davidz>	in their preferences dialog
Oct 17 14:32:43 <davidz>	I can look at the latter
Oct 17 14:32:54 <halfline>	right the factory is just an idea at this point
Oct 17 14:32:58 <halfline>	it's not even on by default right now
Oct 17 14:33:04 <davidz>	it's just PolicyKit and writing a small "copy gconf keys from user directory to system-wide location" thing
Oct 17 14:33:08 <davidz>	another thing
Oct 17 14:33:14 <davidz>	is that we want to introduce another layer in gconf
Oct 17 14:33:26 <davidz>	to do the "copy gconf keys from user directory to system-wide location" feature
Oct 17 14:33:29 <davidz>	anyho
Oct 17 14:33:49 <davidz>	so "introduce another layer into gconf" should be abother feature
Oct 17 14:33:57 <davidz>	now
Oct 17 14:34:04 <davidz>	that's at least three features that relates to gdm
Oct 17 14:34:10 <mclasen>	thats a mini-feature, if anything
Oct 17 14:34:10 <davidz>	(which is fine; I'm just counting)
Oct 17 14:34:29 <mezcalero>	it would be cool if someone would go through all laptop drivers and add the required dmi modalias lines to them
Oct 17 14:34:31 <davidz>	mclasen: maybe! but it may require spec file changes across the entire distro
Oct 17 14:34:42 <mclasen>	davidz: I think we can do without that
Oct 17 14:34:53 <mclasen>	unless you want to be all fhs-correct and move the defaults to /usr/share
Oct 17 14:35:11 <davidz>	mclasen: that would be nice... it's still a pretty important thing to get right so should be tracked as a feature such as to not fall through the cracks (like it kinda did for f8)
Oct 17 14:35:34 <halfline>	another thing is graphical boot
Oct 17 14:35:51 <halfline>	we need to kick rhgb in the shins until it falls on its face
Oct 17 14:36:07 <halfline>	i mean, we need to retire rhgb
Oct 17 14:36:22 <davidz>	mezcalero: definitely worth adding as a feature too
Oct 17 14:36:32 <walters_>	(totally my random opinion, but i don't feel rhgb has a lot of value until we also fix the ~10 seconds of kernel developers saying "Hi Mom!" too)
Oct 17 14:36:36 <halfline>	i don't know where we stand wrt to drm mode setting though?
Oct 17 14:36:52 <davidz>	we need the X hackers to comment on that
Oct 17 14:36:58 <davidz>	ajax, krh, arlied and so on
Oct 17 14:37:09 <halfline>	given ajax is the only one in the room
Oct 17 14:37:14 <halfline>	ajax: speak up!
Oct 17 14:37:17 <mezcalero>	and could we please get rid of sendmail?
Oct 17 14:37:19 <ajax>	do what now?
Oct 17 14:37:30 <davidz>	ajax: you said you weren't going to speak for a month!
Oct 17 14:37:30 <halfline>	tell us where we are on drm mode setting
Oct 17 14:37:47 <ajax>	davidz: no, i said i need no one to speak to _me_ for a month.  subtle difference!
Oct 17 14:37:54 <davidz>	ajax: aha!
Oct 17 14:38:02 <ajax>	halfline: uh.  probably working for intel and radeon for F9?
Oct 17 14:38:10 <halfline>	cool, good enough
Oct 17 14:38:16 <halfline>	more than i was expecting in fact
Oct 17 14:38:21 <davidz>	so do we say "only graphical boot for intel, radeon"? Fine with me fwiw
Oct 17 14:38:28 <halfline>	yup
Oct 17 14:38:41 <halfline>	and everyone else we can say "8 second speed up during boot"
Oct 17 14:38:53 <halfline>	(and we'll quietly whisper the reason is because we dropped rhgb)
Oct 17 14:39:39 <ajax>	not having graphical boot on nv might be unpopular but whatever.
Oct 17 14:40:00 <halfline>	walters_: so the rhgb-replacement I started on actually gets a graphical boot up way earlier than rhgb.  it starts even before the initial ram disk runs through
Oct 17 14:40:10 <walters_>	halfline: nice
Oct 17 14:40:15 <davidz>	other things; 1. I want to replace pam_console entirely - most of the work been done for f8 and there's consensus for what to do for f9; and 2. I want to PolicyKit-ify a few more apps.. maybe even replace userhelper with a small helper that uses PolicyKit to determine whether it's fine to launch a given (X11) app as root - such as to centralize how privileges are managed . Am also talking to other things (s-c-printer for one) in the
Oct 17 14:40:15 <davidz>	 distro using PolicyKit
Oct 17 14:40:16 <ajax>	also, when i say radeon, i mean <=r400, which is not exactly the top-end kit anymore
Oct 17 14:40:41 <davidz>	s/using/about using/
Oct 17 14:40:42 <ajax>	getting r500 and r600 modesetting kernel-side is... possibly a much bigger challenge
Oct 17 14:41:00 <spot>	ajax: even with amd's newfound docs?
Oct 17 14:41:14 <ajax>	yes.
Oct 17 14:41:14 <halfline>	davidz: how do we handle the dbus/consolekit chicken and egg problem?
Oct 17 14:41:20 *	mezcalero wants fsck from nautilus
Oct 17 14:41:25 <walters_>	halfline: the dbus came first.
Oct 17 14:41:27 <ajax>	spot: i really don't want to talk about how much of a disaster that is right now.
Oct 17 14:41:35 *	mezcalero and automatic fsck on dirty unmount for usb sticks
Oct 17 14:41:36 <ajax>	i have two drivers and neither one works
Oct 17 14:41:36 *	spot backs away slowly
Oct 17 14:41:41 <halfline>	meaning dbus needs to know how has console, and consolekit uses dbus
Oct 17 14:41:55 <walters_>	halfline: really though, i think dbus should stop being in the business of fine-grained message control
Oct 17 14:41:55 <ajax>	and i'm supposed to wedge that into the kernel too!  SIGN ME UP
Oct 17 14:42:00 <davidz>	halfline: shouldn't be a problem; let's discuss offline
Oct 17 14:42:01 <ajax>	anyway
Oct 17 14:42:04 *	ajax retreats
Oct 17 14:42:17 <walters_>	halfline: any apps which have at_console=true in the policy config should instead take all messages and talk to policy kit on their own for permissions probably
Oct 17 14:42:27 <spot>	ajax: look out, you're about to fall into that box of DRUGS!
Oct 17 14:42:50 <halfline>	okay, walters_ you want to join in on the conversation davidz and I are going to have offline?
Oct 17 14:43:00 <walters_>	halfline: sure
Oct 17 14:43:12 <davidz>	halfline: one thing I'm looking at is replacing "at_console" in dbus with "requires_polkit_privilege" - e.g. also teching the bus daemon about PolicyKit
Oct 17 14:43:42 <halfline>	davidz: okay, that's a little nicer, more fine grained policy control
Oct 17 14:43:47 <halfline>	still policykit uses dbus too :-)
Oct 17 14:43:58 <davidz>	no, that's in a convenience library
Oct 17 14:44:07 <davidz>	libpolkit does not use D-Bus. At all
Oct 17 14:44:07 <halfline>	ah okay
Oct 17 14:44:13 <halfline>	perfect
Oct 17 14:44:36 <halfline>	we may still need to get ConsoleKit to write out /var/run/console/* for backward compat
Oct 17 14:44:37 <davidz>	also.. I plan to make ConsoleKit maintain /var/run/console too
Oct 17 14:44:40 <davidz>	for backwards compat
Oct 17 14:44:42 <halfline>	jinks
Oct 17 14:44:44 <davidz>	wow
Oct 17 14:44:46 <davidz>	scary
Oct 17 14:44:52 <f13>	Sorry, Ive been in another meething.
Oct 17 14:45:23 <davidz>	halfline: I did half of it http://gitweb.freedesktop.org/?p=ConsoleKit.git;a=commit;h=b2be103bd606291319dc312f07d1f3fcbfdf634c
Oct 17 14:45:37 <f13>	mclasen: to answer your question, yes, as of tonight the freeze is on and builds for fedora 8 release will have to go through releng to get into the final release.
Oct 17 14:46:34 <mclasen>	f13: cool
Oct 17 14:46:53 <f13>	was there any releng type questions?
Oct 17 14:48:02 <mclasen>	f13: no, you pretty much answered them
Oct 17 14:48:22 <mclasen>	f13: earlier on we had a fesco type question about feature proposals/brainstorming for f9
Oct 17 14:50:03 <mclasen>	ok, more f9 plans: mezcalero is going to continue his work on making PA kick ass in F9
Oct 17 14:50:21 <mclasen>	for F9 the topic will be better desktop integration
Oct 17 14:50:29 <halfline>	awesome
Oct 17 14:50:30 <mclasen>	ie make volume control not suck
Oct 17 14:50:55 <mclasen>	clean up the sound capplet, maybe
Oct 17 14:50:58 <halfline>	i know clarkbw was doing some mock ups on ways the volume control could be improved
Oct 17 14:51:07 <mezcalero>	btw, someone sent me a screenshot of the vista volume control
Oct 17 14:51:11 <mclasen>	that is good; we need those mockups
Oct 17 14:51:14 <halfline>	mezcalero: clarkbw ?
Oct 17 14:51:15 <davidz>	mezcalero: yeah, was just about to mention that
Oct 17 14:51:21 <davidz>	mezcalero: it was on p.g.o IIRC
Oct 17 14:51:28 <mezcalero>	i am not sure who did
Oct 17 14:51:30 <mezcalero>	looked good
Oct 17 14:51:36 <mezcalero>	i guess i could check me inbox
Oct 17 14:51:39 <mclasen>	it was jdahlin talking about vista volume control
Oct 17 14:51:45 <mclasen>	on pgo
Oct 17 14:51:46 <davidz>	yeah
Oct 17 14:52:04 <mezcalero>	it's basically a pavucontrol with a sane ui
Oct 17 14:52:07 <mclasen>	and clarkbw was showing vista screenshots here in the grotto yesterday
Oct 17 14:52:22 <davidz>	he linked this http://msdn2.microsoft.com/en-us/library/Aa511278.Sound08(en-us,MSDN.10).png
Oct 17 14:52:30 <mezcalero>	and they are doing vertical volume controls and just show the text on top
Oct 17 14:52:40 <mezcalero>	which is probably fine as long as you only want to show the appname
Oct 17 14:52:42 <mclasen>	I'm not so sure about that
Oct 17 14:52:57 <mclasen>	I guess we'll have to try if vertical works with text
Oct 17 14:53:23 <mezcalero>	yay, vertical text
Oct 17 14:53:31 <mclasen>	nah
Oct 17 14:53:37 <davidz>	vertical text - that's so KDE
Oct 17 14:53:43 <mclasen>	I meant if vertical sliders work with horizontal text
Oct 17 14:54:01 <mezcalero>	and apparently vista includes a peak meter for each stream
Oct 17 14:54:06 <mezcalero>	or at least for each device
Oct 17 14:57:34 <mclasen>	mezcalero: we should probably start to write down the volume control plans on the wiki, too
Oct 17 14:57:43 <mclasen>	once clarkbw hands over those mockups...
Oct 17 14:58:12 <mezcalero>	some novel people wanted to do som UI work too
Oct 17 14:59:36 <mezcalero>	s/novel/nivell
Oct 17 14:59:39 <mezcalero>	s/novel/novell
Oct 17 14:59:40 <mezcalero>	bah
Oct 17 14:59:55 <mclasen>	my own personal f9 feature is the intlclock stuff that has already landed in the panel
Oct 17 15:00:04 <f13>	mclasen: oddly enough we were just talking about that in another meeting.
Oct 17 15:00:31 <mclasen>	I apologize for the fact that the intlclock in the current f9 package is pretty crashy
Oct 17 15:00:35 <f13>	mclasen: I think a relevent comment otmake is that Features can be proposed at any time for F9, FESCo most likely won't start review/acceptance until after Nov 8th.
Oct 17 15:00:44 <mclasen>	need to spend some time debugging it next weekend
Oct 17 15:01:15 <mclasen>	f13: I was mostly wondering if fesco will send out a plea for f9 ideas/features/proposals
Oct 17 15:01:49 <mclasen>	or if that will be self-organizing
Oct 17 15:02:15 <mclasen>	also, we should probably send out pointers to the things we have already started to spec out on the wiki at some point
Oct 17 15:02:28 <mclasen>	to invite comments from the broader public
Oct 17 15:02:31 <f13>	mclasen: FESCo and/or poelcat will be sending out the request for features stuff.
Oct 17 15:02:40 <mclasen>	ok, cool then
Oct 17 15:02:55 <f13>	mclasen: we're really looking to poelcat to drive it, making use of resources as he deems necessary
Oct 17 15:03:25 <mclasen>	ok
Oct 17 15:03:27 <mezcalero>	i'd personally buy the person who implements a sane fsck dbus service for usage in nautilus with support for the "dirty" shutdown flag of fat an ice cream
Oct 17 15:03:45 <mezcalero>	i.e. have an usb stick automatically fscked
Oct 17 15:03:52 <mezcalero>	with a nice gui progress bar
Oct 17 15:03:55 <mezcalero>	if it is plugged in
Oct 17 15:04:07 <mezcalero>	but wasn't unmounted properly before
Oct 17 15:04:11 <mezcalero>	http://www.geocities.com/thestarman3/DOS/DirtyShutdownFlag.html
Oct 17 15:04:33 <f13>	yeah, just mounting read-only is rather rude.
Oct 17 15:04:47 <f13>	'course it doesn't help with fsck.hfsplus segfaults on x86_64 :/
Oct 17 15:07:26 <mclasen>	do we have other F9 features in the pipeline that are worth talking about today ?
Oct 17 15:07:38 <mclasen>	I know that hadess is looking to work on thinkfinger integration
Oct 17 15:08:09 <halfline>	well we've already run over an hour now
Oct 17 15:08:13 <davidz>	mclasen: hmm.. how does hadess work on that fit with Ben Konrath's work?
Oct 17 15:08:42 <halfline>	i think it weould be cool if we made a fake fingerprint on screen based on the fingerprint data
Oct 17 15:09:06 <halfline>	even if it's not the real fingerprint, but a generated one that's unique for the given data
Oct 17 15:09:11 <davidz>	halfline: fingerprint data is probably not available from libthinkfinger
Oct 17 15:09:14 <mclasen>	davidz: I'd hope that it ties nicely together
Oct 17 15:09:33 <mclasen>	I've already mentioned fingerprint enrollment as a thing to add to the dialog to ben
Oct 17 15:09:34 <halfline>	davidz: i don't know anything about the implementation really, but i just think it would be neat
Oct 17 15:09:36 <davidz>	mclasen: I think it's probably the same feature...
Oct 17 15:09:44 <mclasen>	and I've asked clarkbw for some design input on that
Oct 17 15:09:47 <mclasen>	but he probably forgot
Oct 17 15:10:19 <davidz>	halfline: just for fun! http://blog.fubar.dk/?p=35
Oct 17 15:10:24 <davidz>	halfline: (not thinkfinger)
Oct 17 15:10:42 <mclasen>	halfline: you are right, we should close for today
Oct 17 15:11:15 <halfline>	davidz: cool
Oct 17 16:04:59 *	clarkbw realizes he missed the whole meeting...
Oct 17 16:08:25 <clarkbw>	mclasen: http://www.gnome.org/~clarkbw/designs/volume%20control/vista/
Oct 17 16:08:38 <clarkbw>	for the screenshots of vista we looked at yesterday