[00:00] <Bojhan> mpt: hey
[00:00] <Bojhan> mpt: pm?
[01:20] <rickspencer3> can someone /msg me so that I can test something in the indicator-applet?
[01:22] <Bojhan> rickspencer3: is there a bot I can leave msg's for someone in this channel?
[01:23] <rickspencer3> Bojhan: thanks!
[01:24] <Bojhan> rickspencer3: hehe, it was a question as well tohugh
[01:24] <rickspencer3> Bojhan: sorry, I didn't actually read it
[01:24] <rickspencer3> I don't know
[01:35] <bratsche> Laney: I posted a new patch, but I don't have hardware to test this with.  If you happen to have some bluetooth stuff and can test sharing files, that would be awesome.
[02:07] <LaserJock> tedg: around?
[02:07] <tedg> LaserJock: Yes.
[02:07] <LaserJock> tedg: do you have any bugs about the indicator applet not showing up with Pidgin?
[02:07] <tedg> Though, I've had a glass of wine ;)  Expect more sarcasm than usual.
[02:07] <LaserJock> lol
[02:08] <tedg> LaserJock: Yeah, there seems to be one where it disappears.  I haven't been able to track it down though.
[02:08] <LaserJock> it disappears or never appears?
[02:08] <tedg> Everyone seems to be doing what they're told, they just disagree on it.
[02:08] <tedg> Disappears.
[02:09] <LaserJock> hmm
[02:09] <tedg> If you want to see if Pidgin is actually indicating, you can run /usr/lib/indicator-applet/listen-and-print and that will print everything outputing an indicator.
[02:09] <LaserJock> for me it's like it doesn't see the indicator at all
[02:09] <tedg> Are you getting notifications?  You might have the plugin disabled?
[02:10] <LaserJock> hmm, didn't check to see if it was disabled
[02:11] <LaserJock> which plugin is it?
[02:11] <LaserJock> I've got libnotify popups and message notification
[02:12] <LaserJock> ah, "libnotify popups" did the trick
[02:16] <tedg> LaserJock: Heh, now you can test the notification system :)
[02:18] <LaserJock> tedg: yes, it was kinda lonely since I don't use evo
[02:18] <LaserJock> I  still wish I could get bzr-notify to be useful though :(
[02:18] <LaserJock> but as I understand it, it's being worked on some
[02:20] <Bojhan> Do you guys know, I want to leave a message for mpt but I can't seem to reach him trough irc, is there another way I could try contacting or leaving a message, or should I just come back later?
[02:20] <LaserJock> email?
[02:20] <Bojhan> LaserJock: Not on his website
[02:21] <LaserJock> Bojhan: it's on Launchpad
[02:25] <tedg> LaserJock: I do have my own bzr notification plugin.  I haven't done much with it, but I could push it somewhere if you're interested.  It only shows notifications if the operation takes longer than 30 seconds.
[02:26] <LaserJock> my problem is that I can't keep the notifications around long enough
[02:26] <LaserJock> is that something that's easy to tweak?
[02:26] <LaserJock> I keep putting my cursor on them and *poof* :-)
[02:27] <tedg> LaserJock: Not really :(  One of the things that we're discussing for the Karmic version is the idea of a hotkey to keep the around longer.  A "hold" key per se.
[02:27] <LaserJock> I talked to Mark about possibly putting bzr notifications in the indicator applet but he said that was not the right way to do it
[02:28] <tedg> I'm not sure how good of an idea it is though.  It seems nice, but it's entirely indiscoverable.
[02:28] <LaserJock> it's just frustrating when I know my computer is trying to tell me something but I keep missing it
[02:28] <LaserJock> there's no way to get a log, for instance, that I've seen
[02:28] <tedg> LaserJock: ~/.cache/notify-osd.log
[02:29] <LaserJock> ohhhh
[02:29] <LaserJock> that's fantastic
[02:29] <LaserJock> I just need a GUI for it
[02:29] <tedg> Heh, I only have so many answers :)
[02:29] <LaserJock> like "Notification Log" when right clicking the indicator applet
[02:30] <LaserJock> hmm, this is really weird
[02:30] <LaserJock> but I almost wish I had a notifier of the notification :-)
[02:30] <tedg> It would be interesting if Bazaar could make the terminal it's running in call for attention... seems like it'd be possible.
[02:31] <tedg> Probably could exploit the fact that X has almost no security :)
[02:31] <LaserJock> I go away from laptop for 5 min, come back and have a "you have 5 notifications waiting" thing
[02:31] <LaserJock> the thing is queuing notifications
[02:32] <LaserJock> it's *good* information, I hate to be losing it all the time
[02:32] <tedg> The problem there is probably most of them you don't care about... login/logout or IMs, the display in Pidgin is better than anything we could do.
[02:32] <LaserJock> hmm
[02:32] <LaserJock> I can't recall getting a notification I didn't care about
[02:32] <LaserJock> at least at some level
[02:32] <LaserJock> if I don't care about them I turn them off :-)
[02:33] <tedg> See, and now I've helped you understand by adding a flood of notifications from your IM buddies ;)
[02:33] <tedg> Things like "7 new mail messages" from Evolution is useful when it's popping up.  But in 10 minutes is just useless.
[02:34] <LaserJock> well, looking at the log it seems more useful than I thought
[02:34] <LaserJock> perhaps a simple pygtk "aggregator" would do the trick
[02:35] <LaserJock> it's more of what I was thinking the indicator applet was going to be
[02:35] <tedg> Cool, it'd be interesting to see what you come up with.  We couldn't come up with something that we thought would work, but that doesn't mean there isn't an answer :)
[02:36] <LaserJock> yeah, I'm sure plenty of people would find it annoying
[02:36] <LaserJock> but right now the notifications are pretty much all stuff I'd like to be able to queue
[02:36] <LaserJock> something like gwibber for notification is what I'm thinking
[02:37] <LaserJock> perhaps tying into the FUSA "global status" thing would cut down on "noise"
[02:38] <tedg> Yeah, that would be interesting.  The upstream GNOME guys are putting status into gnome session for that purpose.  I don't believe that's the right place to put it, but if it's there, we should use it.
[02:38] <LaserJock> only queue from non-messaging type apps (pidgin, gwibber, etc.) while Available, but present everything that happened while Away
[02:39] <LaserJock> anyway, that's just a braindump after seeing that log
[02:39] <LaserJock> thanks for pointing that out, very useful
[02:41] <LaserJock> heh, I wonder if you could even present the notifications graphically as a timeline
[02:44] <tedg> That would be interesting.  Have you seen the movies of the Wizbit timeline?  It's pretty interesting how it compresses data as you zoom in and out.
[02:47] <LaserJock> it would be rather cool to see
[02:47] <LaserJock> but methinks my programming skills are not up to that task ;-)
[07:29] <pitti> Good morning
[08:55] <seb128> hello there
[08:55] <seb128> mvo: there?
[08:55] <Tm_T> where?
[08:55] <robert_ancell> morning all
[08:56]  * pitti hugs seb128, bonjour
[08:56] <seb128> mvo: robert_ancell is getting the "do you want to migrate the logout button to a fusa applet" after every login
[08:56]  * pitti hugs robert_ancell as well, good morning
[08:56]  * seb128 hugs pitti, guten tag!
[08:56]  * robert_ancell tries to remember his german lessons...
[08:56] <Tm_T> guten morgen
[09:00]  * didrocks hugs seb128, pitti, robert_ancell & Tm_T :)
[09:01] <seb128> lut crevette didrocks
[09:01] <crevette> salut les jeunes :)
[09:01] <didrocks> plop le vieux :p
[09:02] <crevette> good morning to the people who don't speak french
[09:02]  * seb128 got lot of retracer emails and hug pitti again for fixing those
[09:03] <pitti> \o/
[09:03] <pitti> I checked it this morning, all quiet
[09:03] <crevette> hey pitti, I wanted to know how should I proceed to have dgb package for bluetooth
[09:03] <crevette> dbg
[09:03] <pitti> crevette: what's wrong with the -dbgsyms?
[09:05] <crevette> how could I get the dbgsysms ?
[09:05] <pitti> crevette: https://wiki.ubuntu.com/DebuggingProgramCrash
[09:06] <crevette> ah okay, so I need to ask people to add that to their sources list
[09:06] <seb128> pitti: what was your wiki page about email filtering again?
[09:06] <pitti> seb128: https://wiki.ubuntu.com/Bugs/HowToFilter
[09:06] <seb128> thanks
[09:06] <seb128> robert_ancell: ^
[09:07] <pitti> seb128, robert_ancell: http://people.ubuntu.com/~pitti/talks/lp-bug-mail-filtering/lp-bug-mail-filtering.pdf as well
[09:07] <pitti> well, I guess it's not much useful without hearing the talk, though
[09:08] <robert_ancell> thanks
[09:12] <crevette> pitti, another question :/, (I don't have access to a Linux smachine at work), is it possible to tell apport to don't check if crashing application comes from an official package (for https://bugs.edge.launchpad.net/ubuntu/+source/bluez/+bug/343859)
[09:13] <mvo> seb128: oh?
[09:13] <mvo> robert_ancell: what kind of system is it? a hardy->intrepid->jaunty upgraded one?
[09:14] <robert_ancell> mvo: jaunty upgraded
[09:14] <mvo> robert_ancell: could you please kill the running update-notifier and then run "update-notifier --debug-hooks"  and give me the output in a pastebin (it will probably be long)
[09:14] <mvo> ?
[09:15] <robert_ancell> mvo: http://paste.ubuntu.com/138093/
[09:16] <seb128> crevette: it does that by default
[09:16] <robert_ancell> mvo: This caused the notification to pop up again
[09:17] <crevette> seb128, are you sure, I'm used to see when one of my ppa is crashing, that apport can't report the bug because the package is not an official ubuntu package
[09:17] <seb128> crevette: right, which is what you were asking to get no?
[09:17] <seb128> ah no
[09:18] <seb128> crevette: well you have the .crash on your disk, sending it to launchpad would be wrong since that's an ubuntu version
[09:18] <seb128> crevette: you can still apport-retrace your local version
[09:18] <crevette> yeah :/, but I would be interested by the crasher backtrace
[09:20] <crevette> anyway, I need to focus on my day job
[09:20] <seb128> crevette: use apport-retrace locally?
[09:28] <mvo> robert_ancell: thanks, looking
[09:31] <robert_ancell> mvo: I noticed when I closed the notification it crashed, see bug 348883
[09:33] <crevette> seb128, does apport generate debug information even if the package is not official? so the user can upload them and I can retrace on machine.
[09:36]  * mpt replies to Bojhan
[09:37] <seb128> crevette: yes, the .crash is in /var/crash too
[09:37] <mvo> robert_ancell: thanks, thanks that trace too, it looks lik eits crashing in the debug code :P
[09:38] <robert_ancell> mvo: haha
[09:47] <seb128> robert_ancell: https://wiki.ubuntu.com/UbuntuDevelopment/PatchTaggingGuidelines
[09:54] <hyperair> sigh .it's all i can do to not reply very sarcastically to the person who just posted in ubunt-devel regarding the whole update-notifier change.
[10:00] <mvo> robert_ancell: I'm preparing a fix for the log problem and will upload to my ppa, then the log should be more useful, currently its broken :/
[10:00] <robert_ancell> mvo: ok, will test when ready
[10:02] <maxb> hyperair: why sarcastically? I thought it was a reasonably summation of the situation
[10:03] <hyperair> maxb: sarcastically not  to him, but regarding the whole decision of the desktop team in general
[10:03] <hyperair> i'm still feeling rather sore about it
[10:03] <maxb> aha
[10:03] <maxb> Well, +1 on that :-)
[10:03] <hyperair> yeah
[10:04] <maxb> The whole notifications/indicators thing has been hideously mishandled IMO
[10:04] <hyperair> to put it rather bluntly, because a few idiots can't tell where to click,  the rest of us have to endure the stupid window popping up, and delayed notifications of updates, and this new behaviour CANNOT BE TURNED OFF
[10:04] <hyperair> #$%^&*(
[10:05] <seb128> troll troll troll
[10:05] <seb128> not really the channel for that guys
[10:05] <hyperair> right.
[10:05] <mvo> it can be turned off via gconf
[10:05] <seb128> maxb: what has been mishandled?
[10:05] <hyperair> mvo: oh it can? well my icon went missing the moment i turned it off
[10:05] <hyperair> >=(
[10:06] <mvo> hyperair: the key in gconf is "auto_launch"
[10:06] <hyperair> mvo: that's the key i messed with
[10:06] <hyperair> mvo: and the icon effectively disappeared for me
[10:06] <mvo> hyperair: and it does not work? have you restarted update-notifier since?
[10:07] <hyperair> regarding that...
[10:07] <hyperair> perhaps not
[10:07] <maxb> seb128: Well maybe there's undercurrents that I'm not aware of, but it comes across to me as an ordinary user, as a small team at Canonical shoving an unwanted change down everyone's throats... and after FeatureFreeze no less
[10:07] <mvo> please try it, maybe that should be clearer in the gconf description
[10:07] <mvo> ^--- hyperair
[10:07] <mvo> hyperair: if that does not help, please report it as a bug
[10:07] <hyperair> mvo: is there a gconf schema about this?
[10:07] <hyperair> mvo: alright
[10:07] <mvo> hyperair: yes
[10:08] <hyperair> maxb: i thought that was done before FF
[10:08] <seb128> maxb: no change landed after featurefreeze no, and it's not unwanted you always have vocal minorities on lists but that's not always representative of the userbase
[10:08] <mvo> hyperair: we are in string freeze, so I'm a bit hesitant to change the gconf schema doc, but maybe we can create a wiki page for it or something
[10:08] <hyperair> mvo: alright
[10:09] <maxb> I wonder if we don't have a vocal *majority* in this case
[10:10] <seb128> maxb: how do you know?
[10:11] <seb128> maxb: your view is biaised on geeks users I gues
[10:11] <seb128> maxb: we will not know the normal non technical users reaction before having pushed that in a stable version to those users
[10:12] <seb128> maxb: following you way of thinking we would never do any change
[10:12] <seb128> maxb: it's good to try new things, look how it goes and adjust, if that's a failure at least that will have been experimented we can roll back next cycle and we will have learnt something
[10:16] <mvo> robert_ancell: its uploaded to my ppa now, in the meantime culd you please show me ~/.update-notifier/hooks-seen please?
[10:17] <robert_ancell> mvo: doesn't exist
[10:18] <robert_ancell> mvo: sorry, exists as hooks_seen but is empty
[10:19] <Laney> bratsche: No, sorry. Perhaps you can ask in #ubuntu+1?
[10:21] <mvo> robert_ancell: ok, thanks
[10:21] <mvo> robert_ancell: that should not be empty :)
[10:21] <mvo> robert_ancell: do you run i386 or amd64 ?
[10:21] <robert_ancell> mvo: i386
[10:22] <mvo> robert_ancell: ok, I build you a deb, the ppa probably takes some more time until its finished
[10:40] <mvo> robert_ancell: I uploaded a fix for the logging to http://people.ubuntu.com/~mvo/update-notifier/ - could you please install update-notifier and upate-notifier-common from there and run the --debug-hooks thing again?
[10:44] <robert_ancell> mvo: http://paste.ubuntu.com/138142/
[10:47] <mvo> robert_ancell: but the file is still empty I assume?
[10:47] <robert_ancell> mvo: Now contains:
[10:47] <robert_ancell> fusa-applet.note 1223780646 0
[10:47] <robert_ancell> firefox-3.0-restart-required 1237808887 0
[10:48] <robert_ancell> mvo: A bunch of notifications occurred about previous crashes etc
[10:48] <mvo> robert_ancell: ok, please check if you still see the dialog on the next start of u-n / next login
[10:48] <mvo> robert_ancell: right, that is apport
[10:48] <mvo> robert_ancell: maybe/hopefully the log fix fixed the file writing as well :)
[10:49] <robert_ancell> mvo: brb
[10:50] <robert_ancell> mvo: just restarted session, everything seems good now!
[10:52] <seb128> pitti: what is the criterious for failed retracing again?
[10:53] <seb128> pitti: ie shouldn't bug #346102 have been closed?
[10:53] <mvo> robert_ancell: excellent, thanks a lot
[10:53] <robert_ancell> mvo: no prob.  thanks for the fix
[10:53] <seb128> mvo rocks at bug fixing as you can see ;-)
[10:53]  * mvo is also pretty good at creating them 
[10:54] <seb128> hehe
[10:54] <seb128> mvo: just curious but that was a corner case or something? weird that nobody ran into that bug before
[10:54] <pitti> seb128: failed-retrace -> "mostly ??" (do you want the precise algorithm?)
[10:55] <pitti> seb128: we don't auto-close bugs if there weren't obsolete packages
[10:55] <seb128> pitti: ok, I was not sure about that one, thanks
[10:55] <seb128> I was not sure if we decided that "??" for the first n functions was enough to autoclose or not
[10:56] <pitti> we didn't decide that yet
[10:56] <pitti> I'm not saying I'd reject the idea, of course
[11:09] <seb128> mvo: is https://bugs.edge.launchpad.net/ubuntu/+source/update-notifier/+bug/348110 the issue you just fixed too?
[11:09] <seb128> crash in hook_file_mark_as_seen()
[11:21] <crevette> sorry to ask dumb question, but what should I look at in a stacktrace. I should examine each functions to see if parameters are sane for instance?
[11:21] <crevette> http://launchpadlibrarian.net/18611864/Stacktrace.txt is a good candidate to show me
[11:21] <crevette> :)
[11:24] <seb128> crevette: start looking to the crash line
[11:24] <seb128> crevette: is this case hs = 0x0
[11:24] <seb128> crevette: you might have some code which doesn't handle the case where it's not set or something
[11:26] <crevette> at #2 the src seems weird src = {b = "¥\210Ó`\020"}
[11:32] <pitti> mvo: hm, I just saw james_w's question about auto-running Ubuntu CDs in nautilus and update-manager
[11:32] <pitti> mvo: where is the logic that brings up the "ubuntu cd inserted" dialog?
[11:35] <seb128> pitti: update-notifier I guess and nautilus does handle autoruns
[11:35] <seb128> pitti: I think we should just disable the nautilus autorun code, I'm not sure there is any linux CD around using that
[11:35] <seb128> 95% of the cases you will get a microsoft autorun and the dialog will not work
[11:35] <pitti> seb128: I think it's mainly meant for windows CDs and using wine
[11:35] <pitti> but that doesn't work anyway
[11:36] <pitti> seb128: I agree; at least for jaunty, the integration just isn't there yet
[11:36] <pitti> seb128: I just noticed the request for packging gnome-pilot 2.0.17
[11:36] <seb128> another component unmaintained in ubuntu
[11:36] <pitti> is anyone interested in gnome-pilot?
[11:36] <seb128> I've no device to use that and never used it
[11:36] <pitti> neither have I
[11:37] <pitti> seb128: I'd do the update now, mainly since our current version is so old, but I can't test it really either
[11:37] <dobey> heh
[11:37] <seb128> do the update and rely on users to do the testing ;-)
[11:37] <seb128> I'm not sure many user are using that anyway
[11:37] <pitti> crevette, ember, didrocks: does any of you happen to have a device which works with gnome-pilot and is interested in updating?
[11:37] <pitti> seb128: okay
[11:38] <didrocks> pitti: I have any device yet, that uses gnome-pilot. So, it will be difficult for testing :)
[11:39] <pitti> it should have some maintainer in Ubuntu, though
[11:39] <seb128> everything should
[11:39] <seb128> in practice that's not that easy ...
[11:42] <crevette> pitti, gnome-pilot is for palm devices, right ?
[11:42] <pitti> or we should remove it from the default install
[11:42] <pitti> crevette: for PalmOS, yes
[11:43] <crevette> my father gave me one few years ago, but I never used it, I'm not really interested in those devices...
[11:43] <crevette> pitti, I can send it to you if you want
[11:43] <crevette> :)
[11:43] <pitti> crevette: that's USB?
[11:43] <crevette> Paml is so 90's for me :)
[11:43] <pitti> well, I don't exactly have spare cycles to maintain this either, I'm afraid
[11:43] <pitti> but if nobody wants to, we should remove it from default install
[11:44] <seb128> pitti: there was a discussion on the list about dropping it from the default install recently
[11:44] <crevette> pitti, oh no, it's the one with the craddle on serial I guess
[11:44] <crevette> I would vote for dorpping it from default install
[11:45] <crevette> pitti, this is this one http://mobileministrymagazine.com/uploaded_images/PalmVx-789461.jpg
[11:45] <pitti> it's not even in the menu
[11:46] <pitti> ah
[11:46] <seb128> pitti: it is in system preferences
[11:46] <pitti> crevette: a friend of mine has an old Palm 3x with a serial cable; but heck, I don't even have a serial port anywhere
[11:47] <crevette> pitti, me neither
[11:47] <crevette> I just have a laptop with usb nowaday
[11:47] <didrocks> seb128: with "Bienvenu" and not "Bienvenue" in French :D
[11:49] <mvo> pitti: the logic is inside update-notifier, it hooks into libhal
[11:50] <mvo> seb128: I unified the debug log stuff some days/weeks ago and I guess noone ran into a notification since
[11:50] <mvo> seb128: that unification had a bug
[11:50] <seb128> ok
[11:50] <mvo> seb128: the bug is the same
[11:51] <seb128> mvo: see the other bug number I posted there, you might want to close that as well
[11:51] <seb128> ok good
[11:51] <mvo> thanks seb128
[11:51]  * seb128 hugs mvo, good work!
[11:51] <seb128> makes me wonder about the beta testing though
[11:51] <seb128> we are in a beta testing week and nobody noticed the issue ...
[11:52] <seb128> admittedly we do rather new installs testing than upgrades but still
[11:52] <pitti> seb128: bug 348940
[11:52] <seb128> pitti: thanks
[11:52] <crevette> dobey, around?
[11:53] <crevette> mvo, seb128, what are you talking about, missing notification icon when there is updates ?
[11:53] <dobey> crevette: hi
[11:53] <pitti> seb128: I sub'ed you, since AFAIK you ack FFEs for GNOME?
[11:54] <crevette> dobey, hello, I seen there was an update of the icon-naming-spec recently, but no changelog, do you know what was changed, and if it is possible to have the changelog?
[11:54] <seb128> pitti: I'm not sure about main, I've been delegated by MOTU for universe
[11:55] <pitti> ah, I see
[11:55] <seb128> pitti: I think for main that's still u-r deciding, ie slangasek
[11:55] <pitti> I'll put it into the desktop PPA for now then
[11:55] <seb128> pitti: I gave my +1 on the bug for the record ;-)
[11:55] <crevette> does version 0.8.90  mean the version is a like a beta for 0.9 ?
[11:56] <crevette> ah damn it I forgot to open a FFe for nemiver ...
[11:56] <dobey> crevette: http://webcvs.freedesktop.org/icon-theme/icon-naming-utils/ChangeLog?view=log
[11:57] <dobey> crevette: eh, icon-naming-utils will be going away soon anyway.
[11:57] <crevette> dobey, ah I'm stupid, I should have look here.
[11:57] <crevette> dobey, why ?
[11:57] <dobey> crevette: http://launchpad.net/icontool
[11:58] <crevette> dobey, but the spec will continue ?
[11:58] <dobey> crevette: will be merging stuff into there, and it will do more stuff than just the symlink compat crap
[11:58] <seb128> pitti: btw the discussion I mentionned was "Strawman: remove vendor-specific configuration tools from default install" on -discuss
[11:58] <pitti> seb128: yes, I noticed that, but I didn't reply back then
[11:59] <seb128> pitti: https://bugs.launchpad.net/bugs/334446
[11:59] <crevette> icon-namings-utils permits  to maintain compatibility for gnome and KDE icons, right ?
[11:59] <seb128> pitti: you have some palm users subscribed to this bug, you can probably ask for testing feedback there ;-)
[11:59] <dobey> crevette: yes, it just creates symlinks for old icon names, so apps will still get the new icons even though they haven't all migrated
[11:59] <pitti> seb128: nice, thanks
[12:03] <crevette> dobey, do you have an idea for https://bugs.edge.launchpad.net/bugs/348683, si we use application-x-executable ?
[12:04] <crevette> sorry to bother you, don't hesitate to kick me
[12:08] <dobey> there is my opinion/idea :)
[12:11] <crevette> dobey, actually the patch comes from ubuntu
[12:11] <dobey> well the bug is in launchpad, so sure
[12:12] <crevette> the patch is http://patches.ubuntu.com/g/gnome-panel/extracted/09_default_icons.patch
[12:12] <crevette> and we also provided the icon
[12:12] <dobey> i don't see what that has to do with my suggestion?
[12:12] <dobey> which is to say "don't do that"
[12:12] <dobey> :)
[12:13] <crevette> dobey, okay so for you we should drop the patch
[12:13] <dobey> yes, that is what i suggested
[12:13] <crevette> I should test without the patch to see what that look like
[12:14] <dobey> either don't put an icon there, put an empty icon, or fix the code so that all menu items have the same height
[12:14] <crevette> dobey, ah you think the icon is here because it make sures the height is the same on all menu entries?
[12:14] <dobey> i think that's the only valid reason to stick such an icon there
[12:14] <crevette> I don't know the rationale of the patch ...
[12:15] <dobey> (although, using an icon to do that is the wrong fix)
[12:15] <crevette> seb128,, do you know what is the rationale behind http://patches.ubuntu.com/g/gnome-panel/extracted/09_default_icons.patch
[12:15] <seb128> crevette: we got a menu saying that having an icon would be better than the empty upstream way
[12:15] <seb128> crevette: and that icon was suggested
[12:16] <seb128> that was years ago
[12:16] <crevette> s/we got a menu/we got a bug/ I guess
[12:17] <crevette> okay
[12:25] <dobey> bbiab
[13:27] <rickspencer3> I did a clean install of Jaunty Beta last night, it's working quite
[13:27] <rickspencer3> well
[13:27] <rickspencer3> it's the first time I installed a 64 bit build as well
[13:28] <rickspencer3> The only issue so far is that I am experiencing the "Intel Slowness"
[13:28] <rickspencer3> in terms of X
[13:36] <kenvandine_wk> rickspencer3: what video card?
[13:36] <kenvandine_wk> 965 by chance?
[13:36] <rickspencer3> seb128: I see that CD Creator is in System Tools
[13:36] <kenvandine_wk> rickspencer3: try changing to uxa and reporting you results on this wiki page
[13:36] <kenvandine_wk> https://wiki.ubuntu.com/X/UxaTesting
[13:36] <rickspencer3> kenvandine_wk: yah, thanks
[13:37] <kenvandine_wk> uxa  made a huge difference for me... went from unusable to damn fast
[13:37] <rickspencer3> kenvandine_wk: ack
[13:37] <rickspencer3> bryce has been working on this problem for months
[13:37] <pitti> hey rickspencer3
[13:37] <kenvandine_wk> yeah, hard problem
[13:37] <rickspencer3> hey pitti
[13:38] <rickspencer3> I think bryce is pretty close to having it cracked, so I'll wait for him to get online in case there's some extra testing or something I can do to help him out
[13:38] <pitti> UXA doesn't change much FWIW; just causes some rendering artifarcts
[13:38] <rickspencer3> pitti: It looks like "System Tools" menu item crept into the beta
[13:39] <pitti> rickspencer3: yeah, unfortunately :( we quickly discussed it some days ago, but I don't remember the conclusion; seb128?
[13:39] <rickspencer3> well, we shall remove it soon :)
[13:39] <crevette> kenvandine_wk, in config without worg.conf I can just put this part of config and it'll work?
[13:39] <kenvandine_wk> yes
[13:39] <crevette> neat
[13:39] <seb128> rickspencer3: what do you suggest?
[13:39] <seb128> the rational is that this one opens the burn: location
[13:40] <kenvandine_wk> crevette: works great here... my box is completely unusable without it :/
[13:40]  * pitti wants the freeze to end and to get his 30 fix committed bugs closed
[13:40] <seb128> that was in the places menu before but that's not really a place either
[13:40] <crevette> kenvandine_wk, my display is so slow, I need to try that
[13:40] <rickspencer3> Accessories I suppose is the only other logical one
[13:40] <kenvandine_wk> crevette: which video card?
[13:40]  * seb128 is having a meeting with davidbarth and looking through notify-osd bugs
[13:40] <pitti> seb128: speaking of which, the a11y menu also only has one entry (orca)
[13:40]  * kenvandine_wk would rather see a button in nautilus
[13:40] <seb128> rickspencer3: would be confusing to have one for the GUI and one for the nautilus location though
[13:40] <crevette> intel x3100, but I don't remember the display chipset
[13:41] <seb128> pitti: yeah, I told TheMuso and he said he would fix after beta
[13:41] <kenvandine_wk> crevette:   lspci -nn | grep VGA
[13:41] <pitti> ah, nice
[13:41] <crevette> kenvandine_wk, I don't have my laptop here
[13:41] <kenvandine_wk> ok
[13:42] <rickspencer3> seb128: I'm not sure what you mean
[13:42] <rickspencer3> I think it's okay to have more than one way to get to the fuctionality
[13:42] <seb128> rickspencer3: there is one item in sound&video already which opens the GUI
[13:43] <rickspencer3> right, brasereo
[13:43] <seb128> I'm not sure if having 2 similars icons for different things would be confusing
[13:43] <kenvandine_wk> i would rather see a button in nautilus that takes you to burn:///
[13:43] <rickspencer3> I think brasereo is very tuned to creating audio cds (though I know it does more)
[13:44] <rickspencer3> and CD/DVD creator is more of an archiving tool
[13:44] <rickspencer3> I think it would work fine in accessories
[13:44] <seb128> ok good
[13:44] <rickspencer3> and also you can get to it by right clicking on things, and saying "burn"
[13:44] <rickspencer3> what do you think?
[13:44] <seb128> I was considering putting it back in the places menu
[13:44] <seb128> but accessories works too
[13:44] <rickspencer3> hmmm
[13:45] <seb128> I would like to get mpt opinion on that one too
[13:45] <mvo> asac: could you please pastebin me the lshal output of your system with the 3g card?
[13:45] <rickspencer3> I was going to suggest ivanka, but either way
[13:45] <seb128> right
[13:45] <rickspencer3> oops, brb
[13:45] <mpt> seb128, I think Accessories makes a bit more sense than Places
[13:46] <ivanka> did someone call?
[13:46] <mpt> A CD/DVD that you haven't inserted yet isn't really a Place in the sense that all the other items in the menu are
[13:47] <ivanka> mpt: is this what we were talking about the other day?
[13:47] <mpt> ivanka, yes
[13:47] <ivanka> then I agree with you
[13:47] <mpt> the "CD/DVD Creator" item in the Places menu
[13:48] <mpt> seb128, on a slightly related topic, is Terminal in "Accessories" upstream, or is it in "System Tools"?
[13:48] <kenvandine_wk> mpt: that is upstream
[13:48] <mpt> kenvandine_wk, which is?
[13:48] <kenvandine_wk> terminal in acces
[13:48] <mpt> hm
[13:49] <kenvandine_wk> has been like that for quite a while in upstream gnome
[13:49] <kenvandine_wk> maybe since 2.16 or 2.18
[13:51] <ivanka> for a non-technical user it might seem a little incongruous to have it in Accessories
[13:51] <ivanka> Terminal, that is
[13:51] <kenvandine_wk> some distros have patched it to put it in system tools
[13:51] <kenvandine_wk> but personally i don't even like having system tools
[13:51] <kenvandine_wk> clutter...
[13:52] <ivanka> yes
[13:52] <ivanka> clutter doesn't help anyone
[13:52] <ivanka> tech or not
[13:52] <kenvandine_wk> not sure where else to put it
[13:52] <ivanka> what other options are there?
[13:52] <ivanka> at least System Tools implies some sort of direct conversation with the machine
[13:53] <ivanka> Like the mechanics garage vs just topping up the screen wash in the car
[13:53] <kenvandine_wk> i think that stuff fits better under administration
[13:54] <ivanka> What else is in System Tools? (just looking btw)
[13:55] <kenvandine_wk> nothing else
[13:55] <kenvandine_wk> just cd/dvd creator
[13:55] <ivanka> What is the difference between System > Administration and System Tools?
[13:55] <kenvandine_wk> historically it has been a dumping ground for power user stuff
[13:55] <ivanka> k
[13:56] <rickspencer3> hehe
[13:56] <kenvandine_wk> i think things that tend to go there should really go in administration
[13:56] <rickspencer3> about $.50
[13:56] <kenvandine_wk> like disk utilities, etc
[13:56] <kenvandine_wk> gparted puts itself in system tools i think... at least upstream does
[13:56] <rickspencer3> ivanka: we'd like help looking at all the menus, bur for karmic
[13:56] <kenvandine_wk> that clearly belongs in administration
[13:57] <rickspencer3> for now, I would like us to ensure that there is no "system tools" menu under applications
[13:57] <ivanka> mpt and I have already looked at some usability testing we can do for the menus
[13:58] <rickspencer3> ivanka: great
[13:58] <rickspencer3> I would love to see a UDS session around cleaning up the menus
[13:58] <ivanka> We'll add it to the list :)
[13:59]  * kenvandine_wk wonders what upstream gnome hackers will be at UDS... i would really like to see that stuff cleaned up directly upstream for 2.28
[13:59] <dobey> ugh. sometimes when i resume my laptop in jaunty, it seems that the framebuffer is all messed up, and X crashes :(
[13:59] <ivanka> sorry guys, I have to go to a meeting
[14:00] <ivanka> for now I agree with MPT on the subject of CD/DVD creators
[14:00] <dobey> kenvandine_wk: i'll be there :)
[14:00] <ivanka> and agree that Terminal could do with a new home
[14:00] <kenvandine_wk> dobey: i was thinking non-canonical folks :)
[14:00] <ivanka> Not all that actionable
[14:01] <rickspencer3> but to be clear, for Jaunty, the only menu related task is to move CD/DVD Creator
[14:01] <kenvandine_wk> rickspencer3: agreed
[14:01] <ivanka> rickspencer3: yes
[14:01] <rickspencer3> and we're moving it Accessories, it sounds like we are all agreed
[14:01] <rickspencer3> seb128: do you need a bug on that?
[14:02] <ivanka> rickspencer3: yes and bye for now
[14:02] <seb128> rickspencer3: if there is no bug yet please do open one
[14:02] <ember> there's one similar bug #344431
[14:03] <rickspencer3> ember: thanks kindly, I'll use that
[14:03] <kenvandine_wk> seb128: i can take that
[14:04] <kenvandine_wk> i get great pleasure from reducing the number of categories in the menu... :)
[14:04] <rickspencer3> ok, assigned to ken
[14:06] <rickspencer3> does anyone else think it's interesting that I installed the beta last night, and my burning question was about pushing menus around?
[14:07] <rickspencer3> it seems to be a very solid release for a beta
[14:07] <rickspencer3> and my hardware is newish as well
[14:09] <rickspencer3> robert_ancell: how is it going?
[14:09] <robert_ancell> rickspencer3: good.  just squashed bug 345645
[14:09] <rickspencer3> sweet!
[14:10] <robert_ancell> and I managed to close my UK bank account at lunch so they can stop snail mail spamming me in australia :)
[14:11] <rickspencer3> hehe
[14:17] <pitti> rickspencer3: yeah, I have a very good feeling about it as well :)
[14:17] <pitti> and it's the first time I could use the beta freeze to work on my bugs instead of fighting fires
[14:36] <pitti2> pitti2: ping
[14:37] <seb128> hey pitti2
[14:37] <mvo> mvo: ping
[14:37] <pitti> wanted to try whether it crashes for me as well
[14:37] <pitti> pitti2 is pidgin
[14:39] <pedro_> mvo: hello, may you comment on bug 323154 later?
[14:39] <mvo> pedro_: sure, that feature request is *tricky* :)
[14:41] <pedro_> mvo: Thanks!. yeah seems to be one of those :-P
[14:42] <kenvandine_wk> pitti: should i subscribe ubuntu-release as well as ubuntu-main-sponsors for things that need sponsoring now?
[14:42] <kenvandine_wk> or actually just ubuntu-release?
[14:42] <pitti> kenvandine_wk: just u-m-s, unless there is a FF involved
[14:42] <kenvandine_wk> ok
[14:43] <pitti> if there is a FFE, both
[14:43] <kenvandine_wk> so moving the cd/dvd creator menu item?
[14:44] <seb128> kenvandine_wk: doit
[14:44] <kenvandine_wk> seb128: i did it :)
[14:44] <seb128> ok good
[14:44] <kenvandine_wk> just making sure i subscribe the right people :)
[14:44] <seb128> just subscribe the sponsor team to the bug
[14:44] <kenvandine_wk> ok
[14:44] <kenvandine_wk> done
[14:44] <seb128> that's not a feature but a bug fix
[14:52] <pitti> asac: investigating the firefox-3.1 retracer hang now
[14:53] <pitti> asac: indeed it seems that calling debian/rules setup calls run-mozilla.sh with xulrunner-bin--gre-version
[15:13] <seb128> ivanka, mpt: any opinion on what we should do for menu entries which have no icon?
[15:13] <mpt> seb128, can you give an example?
[15:14] <seb128> use the menu editor and add a menu item? ;-) we don't have one in the default installation
[15:14] <seb128> upstream just displays no icon
[15:14] <crevette>  https://bugs.edge.launchpad.net/bugs/348683 ?
[15:14] <seb128> we added one ages ago which leaded to this bug
[15:14] <seb128> crevette: thanks I can manage that discussion I think ;-)
[15:14] <crevette> sorry :/
[15:15] <mpt> ah right
[15:15] <mpt> seb128, I thought there was a generic application icon? iirc it was a purple diamond
[15:15] <seb128> hum, the one nautilus displays when it doesn't know the mimetype or for binaries?
[15:16] <mpt> yes
[15:16] <seb128> do you think we should use that?
[15:16] <seb128> or the no-icon upstream way?
[15:16] <mpt> seb128, I suggest using that same generic-app icon
[15:17] <seb128> mpt: ok thanks
[15:17] <dobey> i really think no-icon is better
[15:17] <seb128> why?
[15:17] <dobey> but eh
[15:18] <dobey> it looks cleaner
[15:18] <mpt> Cleaner like the smile of someone with a missing tooth
[15:19] <seb128> lol
[15:19] <dobey> having a bunch of the same icon in the menus isn't really helpful
[15:19] <mpt> true, but usually this will happen only once or twice in any given menu
[15:19] <dobey> mpt: right, because the analogy is we want all the icons to look the same? :)
[15:23] <pitti> asac: hm, debian/rules patch does that as well; I can certainly dpkg-divert away /usr/lib/xulrunner-1.9.1b3/xulrunner-bin, but that will break once the next xulrunner version will land
[15:31] <pitti> asac: ^ done now
[16:05] <asac> pitti: yes. we will remove the xulrunner binary invocation in postinst
[16:06] <asac> for now the divert should be good enough
[16:06] <pitti> asac: it's not the postinst, it's debian/rules
[16:07] <asac> pitti: oh
[16:07] <asac> thats really unfortunate
[16:07] <asac> you need that for unpack?
[16:07] <asac> and patch
[16:07] <asac> hmm
[16:07] <asac> so meaning i should really try to fix jemalloc
[16:08] <asac> bummer
[16:08] <asac> we use the MALLOC_OPTIONS=O option as you can see in rules
[16:08] <asac> but that only works around against general chroot bustage
[16:08] <asac> seems we need more in fakechroot
[16:11] <davmor2> pitti: bug 338340 can be marked as fix released can't it?
[16:16] <asac> mvo: yes?
[16:28] <davmor2> pitti: also bug 193731 is now covered by the Broadcom-STA in jockey so can this be fix released too?
[16:32] <mvo> asac: I was looking at the NM integration stuff again for update-notifier, you send me some stuff a while ago on irc, could you paste that again to me please?
[16:32] <asac> mvo: what kind of stuff was that? code?
[16:33] <mvo> asac: lshal/dbus examples iirc
[16:33] <mvo> asac: essentially I want to figure out if the default gw is a 3g device, lshal should make it easy find find the device for the interface name
[16:34] <mvo> asac: and I wonder if I should querry NM for the default route or just ask the routing tabel
[16:34] <pitti> asac: it's calling debian/rules patch, and that calls xulrunner
[16:34] <pitti> asac: I'm not sure why applying patches needs that
[16:35] <asac> pitti: was cdbs fixed?
[16:35] <pitti> asac: fixed how?
[16:35] <asac> last i know patch rule alone didnt unpack embedded tarballs
[16:35] <mvo> asac: I'm off for dinner now for ~45min, if you still have the example you gave me that would be reat, otherwise I try to dig up the stuff myself
[16:35] <asac> so its not suitable to get "patched" sources
[16:35] <pitti> asac: well, I just apt-get source'd, debian/rules patch, and that called it
[16:36] <pitti> davmor2: 338340> if it's not happening for you any more, feel free to close it
[16:36] <asac> pitti: yes. right. i just think thats not enough to get properly unpacked sources
[16:36] <asac> for embedded tarballs (unless cdbs was fixed)
[16:36] <pitti> asac: oh, I see
[16:36] <asac> pitti: but lets talk ablout that later
[16:36] <pitti> asac: right; not urgent
[16:36] <asac> pitti: the --gre-version in rules i will look into
[16:36] <pitti> asac: enjoy dinner
[16:37] <pitti> davmor2: closed the other one, tahnks!
[16:39] <davmor2> pitti: I've fix release 338340 for now
[16:40] <pitti> good night everyone!
[16:42] <asac> pitti: cu tomorrow
[16:43] <asac> mvo: you can use nm-tool
[16:43] <seb128> 'night pitti
[16:43] <huats> hello  everyone !
[16:43] <asac> if you dont want to listen for dbus changes
[16:43] <asac> mvo: that would be polling approach though
[16:44] <bratsche> Who did the new update manager stuff?  This is really hot!
[16:44] <asac> mvo: otherwise you can start with nm-tool code
[16:44] <asac> mvo: thats C though
[16:44] <asac> mvo: will introspection files help you?
[16:46] <mvo> asac: so type will tell me "3g" ?
[16:46] <mvo> asac: well, there are two problems a) use the net when it becomes available b) do not use 3g if the user does not want to
[16:46] <mvo> asac: I was wondering how to detect 3g reliable for b)
[16:47] <mvo> I have some idea about a) (I hope)
[16:47]  * mvo is really away for dinner now
[16:51] <seb128> mvo: 5pm is rather tea time ;-)
[16:51] <seb128> oh right, european time is +1 :-)
[17:52] <bryce> rickspencer3: yeah uxa vs. exa for release is worth some discussion
[17:57] <kenvandine_wk> bryce: how are you feeling about uxa?
[17:57] <bryce> kenvandine_wk: I would use the words "open minded"
[17:57] <kenvandine_wk> hehe
[17:58] <bryce> kenvandine_wk: I have not had much trouble with it, but then I've not had the perf issues others reported either
[17:58] <kenvandine_wk> i am scared either way...
[17:58] <bryce> so I'm judging based on quantity of bug reports others have sent in
[17:58] <kenvandine_wk> i know if i installed jaunty on my main desktop and didn't know to go edit xorg.conf to enable uxa
[17:58] <kenvandine_wk> i would be rather upset
[17:58] <bryce> so far, I'm uncomfortable with the number of uxa bugs still open
[17:59] <kenvandine_wk> cause we enable compiz by default... and it is aweful on that hardware
[17:59] <kenvandine_wk> i wish we could be more selective on enabling compiz
[17:59] <bryce> I'm also fearful that many of the bugs we're seeing are going to end up requiring kernel code changes
[17:59] <bryce> so post-release fixup of such issues seems like it'd be tough
[18:00] <bryce> OTOH, it's entirely unlikely we'll get fixes of any form for the EXA problems that we're looking at now
[18:01] <bryce> I know that for people for whom uxa works, it works great and they're very enthusiastic about putting it in by default
[18:03] <bryce> but as it currently stands we already have soooo many bug reports on -intel, it's tough to justify moving to yet another incompletely tested new tech... esp. when we know it's bringing new bugs...
[18:03] <kenvandine_wk> agreed
[18:05] <kenvandine_wk> bryce: i see this as a rock and a hard place kind of situation :/
[18:08] <seb128> I think it's way too late in the cycle to switch the xorg acceleration method on intel
[18:08] <seb128> what is wrong with current driver and exa?
[18:09] <kenvandine_wk> seb128: for some intel hardware... the experience is aweful
[18:10] <seb128> was it better previous cycles?
[18:10] <seb128> and if so what did change?
[18:10] <kenvandine_wk> well, for me it wouldn't be so bad if compiz wasn't enabled by default
[18:10] <kenvandine_wk> my desktop with a 965 card is unusable with a default install
[18:10] <kenvandine_wk> it is fine with metacity
[18:11] <kenvandine_wk> and blazing fast with compiz with uxa enabled
[18:11] <bryce> seb128: have you re-tested since we moved to 2.6.3?
[18:11] <seb128> bryce: I'm running current jaunty on my laptop, didn't upgrade for an hour though
[18:11] <kenvandine_wk> 30+s to render the alt-tab switcher
[18:11] <seb128> weird
[18:12] <seb128> my laptop is intel 965
[18:12] <kenvandine_wk> although, i haven't tested exa on this box in over a week
[18:12] <seb128> and it's just fast using the default config
[18:12] <kenvandine_wk> your's is the GM965 mine is G965
[18:12] <bryce> seb128: 2.6.3 was updated over a week ago so you have it; have you re-tested uxa recently?
[18:12] <kenvandine_wk> mine is a dell desktop with integrated intel video
[18:12] <seb128> bryce: no I didn't exa is fast here and I've no real issue out of this crash in libdrm on user switching
[18:12] <kenvandine_wk> i have that crash too :/
[18:12] <seb128> kenvandine_wk: was it working correctly in intrepid?
[18:13] <kenvandine_wk> i have never used intrepid :)
[18:13] <seb128> kenvandine_wk: https://bugs.edge.launchpad.net/ubuntu/+source/libdrm/+bug/348428 is the crash on user switch for the record
[18:13] <kenvandine_wk> recently in a VM :)
[18:13] <kenvandine_wk> yeah, seen it
[18:13] <seb128> kenvandine_wk: well on whatever driver was available before the one we use now
[18:13] <kenvandine_wk> well, yes
[18:14] <kenvandine_wk> compiz is fine on it with xorg 1.4
[18:14] <seb128> can't we figure what intel did in their driver to trigger that issue?
[18:14] <kenvandine_wk> so probably fine in intrepid
[18:14] <seb128> so that's an xorg issue and not an intel driver one?
[18:14] <seb128> ie current jaunty drive on old server works fine?
[18:15] <kenvandine_wk> the user switching bug is a little different for me... it doesn't hang, Xorg just crashes and i get a new login window
[18:15] <kenvandine_wk> i doubt it would
[18:15] <kenvandine_wk> driver version is quite different
[18:15] <kenvandine_wk> and i think libdrm was 2.4.x
[18:15] <kenvandine_wk> actually, this box came preinstalled with hardy and it worked fine
[18:16] <kenvandine_wk> which means some of those folks that bought the dells pre-installed that want to upgrade to jaunty might feel some pain :/
[18:16] <seb128> I had real hang issues earlier in the cycle due to vsync
[18:17] <kenvandine_wk> bryce: any way to switch between uxa and exa based on hardware ids?
[18:17] <seb128> ups
[18:17] <seb128> vblank I mean
[18:19] <seb128> https://bugs.edge.launchpad.net/xserver-xorg-video-intel/+bug/320813 was the one I was having
[18:20] <seb128> but that has been fixed
[18:21] <seb128> kenvandine_wk: do you have any vblank mention in /var/log/*?
[18:22]  * kenvandine_wk looks
[18:23] <kenvandine_wk>  /var/log/kern.log:Mar 23 16:49:26 trabajo kernel: [264493.894828] [drm:i915_get_vblank_counter] *ERROR* trying to get vblank count for disabled pipe 1
[18:24] <seb128> the linux log on this bug has
[18:24] <seb128> "  * drm/i915: capture last_vblank count at IRQ uninstall time too
[18:24] <seb128>     - LP: #320813
[18:24] <seb128>   * drm/i915: add get_vblank_counter function for GM45
[18:24] <seb128>     - LP: #320813"
[18:24] <seb128> so maybe similar changes are required in other code
[18:24] <seb128> ie G45
[18:25] <seb128> though vblank should be disabled in mesa for intel right now
[18:25] <seb128> do you enable it in your xorg.conf or something?
[18:27] <kenvandine_wk> i don't
[18:27] <kenvandine_wk> i have one line added to my xorg.conf, enabling uxa
[18:27] <seb128> ok, so I don't know
[18:28] <seb128> maybe tjaalton_ has an idea about that ;-)
[18:28] <kenvandine_wk> weird i am getting those errors now and not actually seeing any problems
[18:28] <seb128> but I think your issue is a specific chipset one
[18:28] <seb128> and the solution is to get that fixed and not to switch everybody to a new technology between beta and stable
[18:29] <kenvandine_wk> i would be afraid to switch to uxa now
[18:30] <seb128> we can turn off compiz on G965 if required
[18:30] <kenvandine_wk> i would be happy with that... lowest risk imho
[18:30] <kenvandine_wk> not sure what other hardware might have similar issues
[18:31] <seb128> watching the intel and compiz bugs should give some clues about that
[18:31] <kenvandine_wk> i should also try out exa again... i don't remember when i did that last
[18:31] <seb128> is there lot of users having similar issues?
[18:31] <kenvandine_wk> not sure, bryce?
[18:31] <kenvandine_wk> he said if is afraid of the amount of bugs we have now
[18:33] <seb128> urg, xserver-xorg-video-intel has lot of open bugs indeed
[18:33] <bryce> hmm
[18:34] <seb128> not so many bugs with compiz mentionned or slow in the title though
[18:34] <bryce> seb128: my gut suggests the user switching thing is more likely mesa than libdrm, but there's not proof so far
[18:35] <seb128> bryce: I will get a gdb log when I'm back home
[18:35] <seb128> I don't have a second box to ssh my laptop there
[18:35] <seb128> and valgrind on the xserver just failed yesterday
[18:35] <seb128> valgrind complains about setuid binaries
[18:35] <mvo> seb128: tea time!
[18:36] <mvo> seb128: tea time is good time
[18:36] <seb128> and when I unset the setuid it didn't manage to start xorg
[18:36] <seb128> mvo: ;-)
[18:37] <bryce> kenvandine_wk: it is definitely possible to quirk uxa/exa via hardware ids, however from the evidence I've heard so far there isn't a strong enough correlation between problems and hardware chip types.  For one person with problems on 965, there's another for whom it works great.
[18:37] <kenvandine_wk> bryce: but is it all the same 965?  GM965 vs. G965?
[18:38] <kenvandine_wk> i will test exa again in a minute
[18:38] <kenvandine_wk> my xorg.conf file hasn't been modified in longer than i thought
[18:38] <bryce> seb128: there's one or two performance bugs in -intel that have turned into me-too storms
[18:40] <bryce> kenvandine_wk: well, "965" is covering a huge variety of slightly different chips
[18:40] <bryce> even with GM965 there's a lot of variety
[18:40]  * kenvandine_wk hates hardware
[18:41] <seb128> kenvandine_wk: when did you change your xorg config?
[18:41] <bryce> https://wiki.ubuntu.com/X/UxaTesting - if you look at GM965 8086:2a02, the most common 965, you see uxa works fine for almost everyone, but two people have seen lockups
[18:42] <bryce> er, one person sees lockups, one crashes
[18:42] <kenvandine_wk> 03/05
[18:43] <seb128> that linux was closed after that
[18:43] <seb128> so maybe it's fixed for you ;-)
[18:43] <kenvandine_wk> will test in a few
[18:44] <bryce> https://wiki.ubuntu.com/X/UxaTesting?action=diff&rev2=93&rev1=84
[18:44] <bryce> hmm, those are the changes since the date I upgraded us to 2.6.3
[18:45] <bryce> not a lot of data, but 3 say better, and 3 say mixed
[18:46] <kenvandine_wk> i really thought i had tried right before alpha-6
[18:46] <bryce> I was hoping to see it more overwhelmingly say better
[18:46] <kenvandine_wk> guess not :)
[18:46] <kenvandine_wk> brb
[18:53] <kenvandine_wk> ok, good news
[18:54] <kenvandine_wk> exa seems fine now :)
[18:56] <kenvandine_wk> not quite as smooth as uxa
[18:56] <kenvandine_wk> but usuable
[18:56]  * kenvandine_wk tries user switching :)
[18:56] <kenvandine_wk> woot
[18:56] <kenvandine_wk> that actually worked
[18:56]  * kenvandine_wk tries that again with uxa
[19:06] <kenvandine_wk> ok, user switching doesn't work for me with uxa
[19:06] <kenvandine_wk> but is fine with exa
[19:06]  * kenvandine_wk makes a note on the wiki
[19:07] <seb128> kenvandine_wk: so exa wins again
[19:11] <hyperair> kenvandine_wk: what did you try with?
[19:18] <kenvandine_wk> seb128: yeah... uxa performs a bit better, but exa is usable... but no user switching with uxa
[19:18] <kenvandine_wk> hyperair: ?
[19:19] <hyperair> kenvandine_wk: sorry, i didn't scroll up earlier
[19:19] <hyperair> performance with compiz was good for me with UXA and EXA
[19:19] <hyperair> around the same
[19:19] <hyperair> but with 3d apps, UXA was marginally better, and both completely sucked
[19:21] <hyperair> and mine's the 8086:2a02 one
[19:24] <kenvandine_wk> 8086:29a2
[19:24] <kenvandine_wk> is mine
[19:35] <pedro_> bryce: have you seen something similar to bug 349113?
[19:35] <bryce> dancing...
[19:36] <bryce> pedro_: hmm, the description of that bug is too short
[19:36] <bryce> pedro_: hard to say without more info but maybe it's just the dpi change we put in for jaunty?
[19:37] <bryce> pedro_: ask the reporter if he can git-bisect to see which change specifically solved his issue
[19:37] <bryce> there's not much time left for jaunty to debug cosmetic issues, but if he can identify the specific patch we need, I can take a look at including it
[19:37] <pedro_> bryce: I asked to the reporter trough IRC the same(about dpi change),  but he stated that the git version fixes the issue for him so not sure, I'd prefer to ask ;-)
[19:37] <pedro_> bryce: alright, i'll ask him again, thanks!
[20:35] <Bojhan> mpt: ping me if or when you are here
[20:59] <tjaalton_> kenvandine_wk: the vblank warning is harmless
[21:00] <kenvandine_wk> yeah, seems so
[21:00] <kenvandine_wk> doesn't cause a problem for me
[21:00] <tjaalton> kenvandine_wk: and you can enable vblank in the compiz configuration, so if you had it there -> problems
[21:01] <tjaalton> compiz actually defaulted using it for a while
[21:02] <hyperair> i've always disabled it
[21:03] <hyperair> on intel it's fine, but on nvidia, some windows start quivering