[00:01] <rickspencer3> TheMuso, robert_ancell - good morning
[00:01] <robert_ancell> rickspencer3, hey
[00:02] <TheMuso> Hey guys.
[00:02] <TheMuso> robert_ancell: Hope you haven't updated yet today.
[00:02] <robert_ancell> TheMuso, not yet, problem?
[00:02] <rickspencer3> you can tell we haven't because we are online
[00:03] <robert_ancell> :/
[00:03] <TheMuso> robert_ancell: see /topic in #ubuntu-devel.
[00:04] <TheMuso> rickspencer3: I saw pitti ask you to ask me about gnome-speech replacement. Whats the deal there?
[00:05] <rickspencer3> right
[00:05] <rickspencer3> shall we go over the team meeting, Eastern edition?
[00:05] <rickspencer3> https://wiki.ubuntu.com/DesktopTeam/Meeting/2009-09-15
[00:06] <rickspencer3> TheMuso, robert_ancell , ready?
[00:07] <robert_ancell> yes
[00:07] <rickspencer3> ok
[00:07] <TheMuso> sure
[00:07] <rickspencer3> so first, reiterate the deciscion regarding Empathy vs. Pidgin
[00:07] <rickspencer3> note that I consider this discussion closed
[00:07] <rickspencer3> RESULT: we are sticking with the POR:
[00:07] <rickspencer3>     * Empathy will be installed in new installs, pidgin won't be
[00:07] <rickspencer3>     * Empathy will be installed on upgrade installs, pidgin won't be removed
[00:07] <rickspencer3>     * Empathy will offer to import pidgin settings on first run
[00:08] <rickspencer3> I am just stating this here for the record (again)
[00:08] <rickspencer3> questions are fine, but we are committed to this course
[00:08] <rickspencer3> questions?
[00:08] <robert_ancell> how many important bugs remain before Karmic?
[00:08] <rickspencer3> robert_ancell, for Empathy and Pidgin?
[00:08] <robert_ancell> Empathy
[00:08] <rickspencer3> I can't answer that
[00:08] <TheMuso> none for me, I just have to prepare myself for the backlash from blind users that think they have to have everything the way they want them or else... :)
[00:09] <rickspencer3> there are none that are release target atm
[00:09] <robert_ancell> ok
[00:09] <rickspencer3> TheMuso, can they still use pidgin?
[00:09] <rickspencer3> note my second point
[00:09] <TheMuso> rickspencer3: Oh sure, but its not on the CDs by default, which is what they will get annoyed about/
[00:09] <rickspencer3> also, we should capture their concerns as legitimate bugs and fix them as soon as is feasible
[00:09] <rickspencer3> TheMuso, right
[00:10] <TheMuso> Thats if they file them. The blind community are hopeless at bug filing.
[00:10] <TheMuso> overall
[00:10] <rickspencer3> I suppose this represents a significant regression, and they are probably entitled to be annoyed
[00:10] <TheMuso> A few who do help out in that regard.
[00:10] <rickspencer3> ok
[00:10] <rickspencer3> I take these concerns very seriously ...
[00:10] <TheMuso> Ok I appreciate that.
[00:10] <rickspencer3> but I understand that Empathy is usable by blind users
[00:10] <rickspencer3> just not convenient
[00:10] <TheMuso> Right.
[00:11] <TheMuso> Its usable, but not the point that pidgin is.
[00:11] <TheMuso> to the point
[00:11] <rickspencer3> in general, I think that blind users have the right to expect to use the same tools as other users
[00:11] <TheMuso> As for gnome-speech replacement, I see the two work items I need to address, one which I can, and one which is ongoing.
[00:11] <rickspencer3> so we should fix these issues asap
[00:11] <TheMuso> Right.
[00:11] <rickspencer3> ok, that's the next section
[00:11] <rickspencer3> first, kenvandine gave the partner update
[00:12] <rickspencer3> a big bug fix for couchdb came this week
[00:12] <rickspencer3> and there are some complaints about the usability of the u1 client preference dialog
[00:12] <rickspencer3> those will need a ui freeze exception to get fixed
[00:13] <rickspencer3> not too much to report on Kubuntu front
[00:13] <TheMuso> yeah read that
[00:13] <rickspencer3> they seem pleased by their progress with the netbook
[00:13] <rickspencer3> we talked a lot about bugs
[00:13] <rickspencer3> neither of you have targeted bugs atm, so you are off the hook
[00:14] <rickspencer3> then we talked about work items
[00:14] <rickspencer3> TheMuso, this is where pitti mentioned you
[00:14] <rickspencer3> TheMuso, you have two open work items regarding gnome-speech-replacement
[00:14] <TheMuso> rickspencer3: Right.
[00:15] <TheMuso> Make that one.
[00:15] <rickspencer3> seems one is just writing the test plan and release notes
[00:15] <rickspencer3> oops, wrong project, sorry
[00:15] <TheMuso> np
[00:15] <rickspencer3> go ahead, TheMuso , what's the status of these work items?
[00:15] <TheMuso> in any case, the other is ongoing, and is not that easy. :)
[00:15] <TheMuso> rickspencer3: Ok debugging pulseaudio output is still ongoing, and the promotion of packages to main is done.
[00:16] <TheMuso> The whiteboard now reflects that.
[00:16] <rickspencer3> ah
[00:16] <rickspencer3> k
[00:16] <rickspencer3> we were looking at these:
[00:16] <rickspencer3>     * Promote speech-dispatcher and libdotconf to main: TODO
[00:16] <rickspencer3>     * Debug speech-dispatcher's pulseaudio output support: TODO
[00:16] <rickspencer3> so, the first is done
[00:16] <TheMuso> yeah
[00:16] <TheMuso> yes
[00:16] <TheMuso> I just marked it as such.
[00:16] <rickspencer3> and the second you just work on up to the bitter end?
[00:17] <TheMuso> Indeed.
[00:17] <rickspencer3> get as much done as you can?
[00:17] <rickspencer3> okay
[00:17] <rickspencer3> thanks
[00:17] <TheMuso> np
[00:17] <rickspencer3> TheMuso, any red flags? concerns? regarding pulsesaudio output support?
[00:18] <TheMuso> Only that the upstrea author of the code has not been very helpful.
[00:18]  * rickspencer3 nods
[00:18] <rickspencer3> ok ..
[00:18] <rickspencer3> robert_ancell,
[00:18] <TheMuso> As he simply adapted, somewhat badly, the code from another project for speech-dispatcher.
[00:18] <robert_ancell> pulse has been playing better for me the last few days
[00:19] <TheMuso> robert_ancell: Great.
[00:19] <rickspencer3> Update Compiz triaging instructions https://wiki.ubuntu.com/DebuggingCompiz:
[00:19] <rickspencer3> robert_ancell, are you still going to do that for Karmic, or shall we postpone?
[00:19] <robert_ancell> I will do that today
[00:19] <rickspencer3> robert_ancell, I didn't mean that to crack the whip
[00:20] <rickspencer3> if it's not important, or can be done after the release
[00:20] <rickspencer3> feel free to set it to postpone
[00:20] <robert_ancell> :) No, it should be done and it wont take too long
[00:20] <rickspencer3> ok
[00:20] <rickspencer3> the great bulk of remaining items are related to firefox, and asac feels he has those under control
[00:21] <rickspencer3> last topic was regarding this page:
[00:21] <rickspencer3> https://wiki.ubuntu.com/Bugs/FindRightPackage
[00:21] <rickspencer3> the qa team is working to change the mix of bugs so that most of them come from ubuntu-bugs rather than the web interface
[00:21] <rickspencer3> so we should have good instructions on this page to help with that goal
[00:22] <rickspencer3> (which will benefit us dramatically)
[00:22] <robert_ancell> cool
[00:22] <rickspencer3> could both please look at that page and make sure:
[00:22] <rickspencer3> 1. all the areas that have packages are covered
[00:22] <TheMuso> right
[00:22] <rickspencer3> 2. that coverage is accurate
[00:23] <rickspencer3> in other words, make the instructions correct and complete
[00:23] <TheMuso> right
[00:23] <rickspencer3> TheMuso, I can't image this:
[00:23] <rickspencer3> ubuntu-bug alsa-base
[00:23] <rickspencer3> is terribly good advice these days ;)
[00:24] <rickspencer3> though I could be wrong
[00:24] <TheMuso> Audio is a difficult one, as there are many bits that are part of the stack. I need to think that over and see if I can come up with something better.
[00:25] <rickspencer3> TheMuso, symptom based reporting script?
[00:25] <rickspencer3> TheMuso, please discuss this possibility with pitti
[00:25] <TheMuso> rickspencer3: I believe pitti and seb128 added support for audio, but I don't know how to call it.
[00:26] <TheMuso> for symptom based reporting.
[00:26] <rickspencer3> ACTION: TheMuso to check on how to call the audio symptom script and update the wiki accordingly
[00:26] <rickspencer3> TheMuso, ok? ^
[00:26] <TheMuso> fine by me
[00:27] <rickspencer3> robert_ancell, I wonder if "graphic environment" might be too broad
[00:27] <rickspencer3> I think there is a symptom script for this ... but also, I wonder if we should try to tease out compiz issues?
[00:27] <rickspencer3> ACTION: robert_ancell to check on Graphic Environment section
[00:28] <rickspencer3> robert_ancell, have time for this? ^
[00:28] <robert_ancell> I'll have to ask Bryce but running everything through one point may actually be a good idea
[00:28] <robert_ancell> yes
[00:28] <rickspencer3> robert_ancell, bryce is on leave ... but you can ask tselliot
[00:29] <rickspencer3> his real name is Alberto, and he is virtual bryce for September
[00:29] <robert_ancell> :)
[00:29] <rickspencer3> TheMuso, robert_ancell any other business?
[00:30] <TheMuso> No.
[00:30] <robert_ancell> no
[00:30] <rickspencer3> ok then
[00:30] <rickspencer3> thanks for your time, as usual :)
[00:31] <TheMuso> np
[00:31]  * rickspencer3 taps gavel
[00:50] <TheMuso> c/
[03:28] <Keybuk> hmm
[03:28] <Keybuk> xsplash is hanging around like an ex on facebook
[04:37]  * ccheney notices thinkpad x200 are pretty cheap currently on the lenovo cpp site
[04:37]  * ccheney thinks the new x210 will be coming out soon
[06:16] <mac_v> Amaranth: hei , you up ? is Bug #391479 due to compiz ?
[06:18] <Amaranth> mac_v: I suppose it probably is since one person says it doesn't happen with metacity
[06:18]  * Amaranth stabs gnome-power-manager
[06:18] <mac_v> hm... testing it out
[06:18] <mac_v> what's up with gpm?
[06:19] <hyperair> many issues
[06:19] <Amaranth> trying to see if jaunty's compiz has issues with suspend to see if a bug is in compiz or the intel driver but gpm won't recognize I unplugged
[06:19] <hyperair> heh
[06:19] <Amaranth> and it just took 15 seconds before it started doing a suspend manually
[06:19] <mac_v> heh , gpm started giving my screen burn-ins!
[06:19] <hyperair> burn ins?
[06:19] <Amaranth> wtf
[06:20] <Amaranth> it just woke up then suspended again :/
[06:20] <hyperair> when i press my suspend key, my screen turns off instead of suspending
[06:20] <mac_v> yeah! the screensaver didnt activate , so it was all nite with the same view!
[06:20] <Amaranth> mac_v: you have a CRT?
[06:20] <mac_v> LCD
[06:20] <mac_v> hehe , when i suspend and resume everything freezes ;)
[06:20] <Amaranth> "Your system encountered a serious kernel problem."
[06:22] <Amaranth> oh, it's just a warning
[06:22] <Amaranth> a little dramatic there
[06:22] <mac_v> i like ubunut+1's new topic :)
[06:22] <mac_v> " Topic for #ubuntu+1 is: Neither karmic nor the buildds are in a working state right now. "
[06:23] <Amaranth> yep, buildds got hosed so packages were published that weren't actually finished building or something
[06:23] <Amaranth> I'm afraid to reboot :P
[06:23] <Amaranth> wait, I just did reboot...
[06:23] <mac_v> \o/ me too
[06:24] <Amaranth> I think that problem may already be fixed. Someone on the planet posted instructions for chroot'ing in and upgrading packages to get their system working again
[06:24] <mac_v> i had to fight to get it to reboot , anyways i rarely shutdown ;p
[06:24] <Amaranth> 3 suspends so far and no sign of this bug, this is bad news for me
[06:25] <Amaranth> hehe, another "Your system encountered a serious kernel problem."
[06:26] <mac_v> hm.. the fan warns me the BIOS is not recognized o.0 ! , is there a list or something for which is recognized?
[06:27] <Amaranth> your fan talks to you?
[06:27] <mac_v> ;) during boot
[06:27] <Amaranth> I'm kind of excited, I finally get to use git bisect
[06:28] <mac_v> acerhdf: unknown (unsupported) BIOS version Acer, inc./Aspire 5670     /v1.3219, please report, aborting!
[06:28] <Amaranth> oh, yeah, that'll be a list of some kind
[06:28] <superm1> Amaranth, the excitement wears off quickly after you realize how many times you have to compile to find the right commit :)
[06:28] <mac_v> it sounds like "Danger Will Robinson"
[06:28] <mac_v> ;p
[06:29] <Amaranth> superm1: Yeah, I'm thinking of just rolling back suspicious commits manually
[06:29] <Amaranth> it gets worse: the problem may be in a compiz-fusion plugin so I'll have to git bisect at least 3 git branches
[06:30] <Amaranth> although I guess I can rule that out easily enough by disabling those plugins and trying to reproduce
[06:30] <Amaranth> gpm is not seeing plug/unplug events at all
[06:31] <mac_v> Amaranth: i think the synaptic/update manager bug i showed earlier is infact due to compiz , i switched the window manager to metacity and not probs 0.0
[06:33] <Amaranth> why do we always have a "everything died, you can't boot and we can't build fixed packages" event near the end of development cycles?
[06:34] <Amaranth> last one was libc6 finally getting updated and discovering the updated build flags we were using made it die
[06:34] <Amaranth> that one nicely trashed the buildds too
[06:35] <Amaranth> ah, this one was just an upstart problem
[06:41] <mac_v> argh ! xsplash overrides the !"quiet" kernel option :(
[06:48] <mac_v> hyperair: do you know the bug# for the Alt+f2 problem?
[06:49] <mac_v> has it started working for you?
[06:52] <Amaranth> 6 suspends with jaunty compiz without problems, 6 suspends with karmic compiz without problems
[06:52] <Amaranth> This bug is either already fixed or just got a lot more annoying
[06:53] <Amaranth> I so hope this bug is already fixed because I _really_ do not want to do more than 6 suspends on each bisect run :/
[07:03] <Amaranth> ok, it's an intel bug
[07:03] <Amaranth> kwin users have the same problem
[07:03] <Amaranth> phew
[07:04] <mac_v> hehe ;)
[07:05] <Amaranth> mac_v: can you assign me to that update-manager bug instead of mvo?
[07:05] <mac_v> sure
[07:05] <Amaranth> I'm too frustrated with launchpad slowness to deal with it right now
[07:05] <mac_v> Amaranth: also are you still having problems with the logout plugin?
[07:05] <Amaranth> haven't gotten a chance to mess with it
[07:06] <Amaranth> it's on the bottom of my compiz todo list tbh since it's just a little shine instead of a bug
[07:06] <mac_v> yeah
[07:06] <Amaranth> I'll try to take a look tomorrow though since a little more shine might help people ignore the 500 or so bugs open in compiz right now :P
[07:07] <mac_v> hehe ;p .. also an FYI , i updated the latest karmic compiz , and no problems for me
[07:07] <Amaranth> mac_v: the panel bug seems to only be easy to trigger on the guest session
[07:08] <Amaranth> otherwise it is very intermittent
[07:08] <Amaranth> unless you just meant fine in general :)
[07:08] <mac_v> hrm... didnt try with guest session , let me try
[07:12] <hyperair> mac_v: what alt+f2 problem?
[07:13] <hyperair> mac_v: also, doesn't xsplash only start when X starts?
[07:13] <hyperair> mac_v: my xsplash doesn't start until after usplash shows me all my messages
[07:17] <mac_v> hyperair: the alt-f2 not working when panel has transparency
[07:18] <hyperair> mac_v: sorry i didn't look into that any more, since i removed my panel's transparency (it's pseudo, not composited)
[07:18] <mac_v> hehe ;)
[07:18] <hyperair> :)
[07:18] <hyperair> i've instead used compiz to match gnome-panel windows for transparency
[07:19] <mac_v> hmm , my xsplash after the latest updates doesnt show the usplash messages , or maybe its due to the boot ppa !
[07:19]  * mac_v should stop testing every ppa at the same time! :(
[07:20] <hyperair> hahah
[07:20] <hyperair> i remember sarvatt wrote a script for cleanly removing a PPA
[07:20] <hyperair> you know, reset every package to the non-PPA package and so on
[07:20] <mac_v> yeah thats for the xorg edgers , should look into that one :)
[07:21] <mac_v> for now , i'v just pinned my boot ppa to a lower priority , so i wont get prompted for an update unless i choose \o/
[07:22] <Amaranth> mac_v: xsplash isn't supposed to show those messages
[07:22] <mac_v> :(
[07:22]  * mac_v wants messages ... grr...
[07:22] <Amaranth> iirc the idea is to start X as soon as possible and show a progress bar but if you need a password or something for encryption it flips back to usplash to ask then flips back to xsplash
[07:22] <Amaranth> but xsplash doesn't show any messages
[07:22] <mac_v> yeah
[07:23] <Amaranth> since I don't have KMS that flip back and forth bit is rather annoying
[07:23] <mac_v> ATI?
[07:23] <Amaranth> intel but with KMS on I have no brightness control
[07:23] <mac_v> crap the screen just blanke on me!
[07:23] <mac_v> blanked*
[07:23] <Amaranth> they need to port the various methods of changing the screen brightness from the xorg driver to the kernel
[07:24] <Amaranth> right now you only get brightness control if your panel supports some ACPI standard or something
[07:25] <mac_v> grr... more gpm testing!
[07:35] <pitti> Good morning
[07:36] <pitti> TheMuso: there's no symptom script for audio right now; would be great to get one, of course
[07:37] <mac_v> pitti: morning  ... any idea where i could check for a recognized BIOS version? i get this error in my kernel log > acerhdf: unknown (unsupported) BIOS version Acer, inc./Aspire 5670     /v1.3219, please report, aborting!
[07:37] <pitti> TheMuso: there's the beginnings of it in the totem package hook; I guess some of it can become part of the sound hook; even though it's just very coarse, it's already better than nothing
[07:37] <mac_v> maybe it could solve the gspca module problem
[07:38] <pitti> mac_v: hm, shouldn't that be displayed on boot, or in the bios setup?
[07:38] <mac_v> no problem is displayed in the bios setup
[07:44] <TheMuso> pitti: hrm ok.
[07:44] <mac_v> but the fan does work without any problems
[07:49] <robert_ancell> pitti, morning!  btw I made a BZR branch for gnumeric when I was working on that bug to cherrypick.  The gnumeric packing style is _very_ old
[07:49] <pitti> robert_ancell: oh, sorry; I just did it yesterday
[07:50] <pitti> robert_ancell: yeah, it's ancient
[07:50] <robert_ancell> pitti, ok, I will merge it into the branch
[07:56] <robert_ancell> pitti, oh, is there a reason why we don't update gnumeric to the latest version?
[07:57] <Amaranth> people still use gnumeric?
[07:58] <pitti> robert_ancell: it's not part of gnome, thus was just bound by the debian import freeze
[07:58] <pitti> robert_ancell: (i. e. nobody cared to update it until FF)
[07:58] <pitti> Amaranth: "still"? gnumeric is awesome
[07:58] <pitti> I like it so much better than OO calc
[07:58] <pitti> anyway, need to reboot, brb
[07:59] <robert_ancell> Amaranth, hey
[07:59] <Amaranth> robert_ancell: howdy
[08:26] <pitti> ArneGoetje: hm, do you know why language-support-writing-fi stopped depending on openoffice.org-voikko and mozvoikko?
[08:33] <ArneGoetje> pitti: yep
[08:34] <pitti> ArneGoetje: I cleaned up http://people.canonical.com/~ubuntu-archive/component-mismatches.txt now by explicitly seeding the ones which were previously pulled in through l-s-translations-*
[08:34] <ArneGoetje> pitti: that dependency is now handled by language-selector and only if mozilla products or oo.o is installed for Finnish users
[08:34] <pitti> and I'll remove the obosolete -zh sources
[08:34] <pitti> ArneGoetje: ah, good to know; then I'll seed them
[08:35] <seb128> good morning there
[08:35] <ArneGoetje> pitti: but not the zh transitional package, right?
[08:35] <seb128> Keybuk, thanks for fixing this upstart dbus restart issue ;-)
[08:35] <pitti> ArneGoetje: I seeded the transitional binaries (we need them until after karmic+1)
[08:35] <pitti> hey seb128
[08:36] <ArneGoetje> pitti: yep
[08:36] <pitti> ArneGoetje: but the old source packages for language-pack-gnome-zh etc. can go
[08:36] <ArneGoetje> pitti: right
[08:36] <Amaranth> seb128: you mean you like having your desktop working?
[08:37] <seb128> Amaranth, my desktop was never broken, upgrades just started to close my sessions which is not the best hing
[08:37] <seb128> thing
[08:37] <Amaranth> ah
[08:37] <Amaranth> only problem with upgrades I had was network manager dying and not coming back but it seems to start itself again now
[08:37] <seb128> Keybuk, btw do you want to get bugs about your changes assigned to you?
[08:38] <seb128> Keybuk, ie bug #430494
[08:40] <ArneGoetje> pitti: is aplha6 through now?
[08:40] <pitti> ArneGoetje: no, tomorrow
[08:40] <Amaranth> pitti: would this be a bad time to upload something then? :)
[08:40] <pitti> Amaranth: major bug fixes are okay
[08:41] <pitti> but alpha-6 CDs are currently being built
[08:41] <pitti> Amaranth: n-m> had that, too, seems fixed now
[08:41] <Amaranth> I dunno how major it is but the fix for compiz making your panels invisible got reverted in the time between me telling mvo about it and him pulling a new git snapshot
[08:41] <pitti> Amaranth: except after suspend, then I have to tick the "enable network" box manually
[08:41] <ArneGoetje> pitti: could you do a code review for language-selector for me and push it into alpha6? Currently there are nasty bugs which can cause crashes and make l-s unusable. I have fixed them in my branch, but mvo didn't have time to review them yet.
[08:41] <Amaranth> pitti: other way around for me, I tried to disable networking while doing suspend testing and it kept turning back on after resume
[08:42] <pitti> ArneGoetje: no, I can't push it into alpha-6; CDs are being built now
[08:42] <ArneGoetje> pitti: :(
[08:42] <pitti> ArneGoetje: can you please mail me about the code review?
[08:42] <ArneGoetje> pitti: sure
[08:42] <pitti> we can upload it, but it's not guaranteed to go in
[08:42] <Amaranth> I guess the compiz fix doesn't matter so much then if it can't make alpha 6 anyway
[08:42] <seb128> Amaranth, lot of people don't use CD images but daily karmic updates
[08:42] <ArneGoetje> pitti: I guess it would be enough if it's in the archive available as an update
[08:43] <seb128> or rather most users install updates and will not be limited to CD testing
[08:43] <pitti> ArneGoetje: *nod*
[08:43] <Amaranth> I guess it's a good thing my patch for panel shadows is there since you can at least see the shadow of the panel and know it is running
[08:43] <Laney> I'm finding that gnoe-terminal i dopping some ofmy kepresses
[08:43] <Laney> is tis a known bug?
[08:43] <Amaranth> seb128: Right, unless it was the beta
[08:47] <Amaranth> Who is handling xorg bugs right now?
[08:47] <seb128> Amaranth, tseliot
[08:47] <Amaranth> ah, good timing then :)
[08:47] <tseliot> Amaranth: what's up?
[08:47] <Amaranth> tseliot: bug 422669 should probably have it's importance set to high
[08:48] <Amaranth> I'd like to help with getting it fixed but I don't even know where to start on such things
[08:48] <Amaranth> step 1: change the title :)
[08:48] <Amaranth> it affects compiz too
[08:48] <seb128> Amaranth, do you know about compiz decoration going away after suspend, resume?
[08:48] <seb128> is that this issue?
[08:48] <Amaranth> seb128: That's that bug
[08:48] <Amaranth> All windows with RGBA visuals go invisible
[08:48] <seb128> +1 to get it fixed for karmic ;-)
[08:49] <tseliot> Amaranth: the latest mesa + apw's kernel should fix that
[08:49] <Amaranth> I did about 20 suspends trying to reproduce it with various compiz package versions
[08:49] <chrisccoulson> good morning everybody!
[08:49] <seb128> hey chrisccoulson!
[08:49] <tseliot> that bug is duplicate
[08:49] <chrisccoulson> hey seb128
[08:49] <Amaranth> Amusingly I couldn't reproduce it at all when I was trying but then did so right after I stopped trying
[08:49] <Amaranth> mvo: good morning
[08:50] <Amaranth> mvo: some bad news: between the time I told you about the invisible panel fix and you making a new snapshot my fix was reverted and an alternative was tried that doesn't work :/
[08:51] <Amaranth> tseliot: Do you have a bug number for the original bug?
[08:51] <tseliot> Amaranth: bug 429241
[08:51] <pitti> tseliot: good morning
[08:51] <tseliot> pitti: good morning to you
[08:51] <mvo> hey Amaranth - hrm, thats bad
[08:51] <pitti> tseliot: I'm running latest mesa with andy's kernel now, suspended in the morning, and will now watch out for corruptions and freezes
[08:52] <tseliot> pitti: thanks a lot. Let me know if something goes wrong
[08:52] <Amaranth> tseliot: Are you sure that's the same issue? People in the other bug were talking about crashes and such but that was not the original issue the bug was about
[08:54] <tseliot> Amaranth: yes, the two problems (which could show up separately) were part of the same problem
[08:54] <Amaranth> tseliot: alright, will mark these as dupes of that one
[08:55] <Amaranth> I have never had the GPU hang and cannot reliably reproduce even the missing windows so I'll take your word for it
[08:55] <chrisccoulson> everybody managed to boot ok this morning then after the updates last night?
[08:55] <tseliot> Amaranth: thanks
[08:55] <Amaranth> It was nice to see that my system survived 20 rapid suspend/resume cycles though :)
[08:56] <tseliot> :-)
[08:57] <Amaranth> hrm, using two computers at the same desk at the same time is confusing
[08:57] <Laney> the quick brown fox jumps over telzy dg
[08:57] <Laney> I wonderhowIwould dbug this
[09:08] <chrisccoulson> seb128 - would devhelp fall under the standing freeze-exception that exists for other gnome packages?
[09:08] <pitti> chrisccoulson: bit of a borderline case, I think; how intrusive are the changes?
[09:09] <chrisccoulson> pitti - not sure. i just noticed that its quite out of date
[09:09] <chrisccoulson> i was just looking for stuff to update:)
[09:09] <pitti> oh, we synced that from Debian
[09:10] <pitti> they just have 0.23.1-1
[09:10] <pitti> so that shuold probably be safe to sync
[09:10] <seb128> that being?
[09:10] <pitti> seb128: devhelp
[09:10] <seb128> oh, right
[09:11] <Amaranth> chrisccoulson: it's not really that far out of date, 0.23.1 is the latest stable release and 0.27 is a development release, I guess
[09:11] <Amaranth> err, 2.27
[09:12] <Amaranth> so definitely development
[09:16] <chrisccoulson> ah, ok. 0.23 isn't that old then. it's just the version numbers that are misleading now upstream have a 2.27.92 release
[09:17] <Amaranth> maybe if they stick to the GNOME release schedule we can get development versions in karmic+1 :)
[09:19] <seb128> james_w, pitti: is bug #425792 a duplicate of bug #403192?
[09:20] <seb128> see bug #422392 from keescook too
[09:21] <kklimonda> chrisccoulson: why have you regenerated aclocal.m4 for transmission?
[09:21] <asac> hmm ... took me 20 resumes to get my computer up
[09:21] <chrisccoulson> kklimonda - it's regenerated automatically by autoreconf
[09:22] <chrisccoulson> because there is a patch that touches the build system
[09:22] <kklimonda> chrisccoulson: maybe we could just inject launchpad-integration into Makefile.in instead?
[09:22] <kklimonda> ya, I've made it myself and that's why I'm asking ;)
[09:23] <chrisccoulson> kklimonda - we prefer not to manually edit Makefile.in, because it can get difficult to maintain doing that
[09:23] <kklimonda> chrisccoulson: what happens if we don't apply 99_autoreconf.patch? spare me a free rebuilt :)
[09:23] <chrisccoulson> it's much easier to patch Makefile.am and then regenerate Makefile.in
[09:23] <kklimonda> ya, that was my understanding
[09:24] <chrisccoulson> kklimonda - if you don't apply that patch, the launchpad-integration won't work
[09:25] <chrisccoulson> it only worked before because there was a debian/rules hook to run autoreconf in the build system. but with desktop packages, we tend to regenerate it before uploading
[09:25] <chrisccoulson> some people have different opinions on how to do it though ;)
[09:25] <kklimonda> I see, i see
[09:25] <kklimonda> so instead of using autoreconf I should regenerate it myself?
[09:26] <kklimonda> it seems to be more work but maybe that's just me
[09:26] <asac> Keybuk: any ideas how i can best upstream the upstart init script? can we detect during build time
[09:26] <asac> if upstart version is high enough?
[09:28] <asac> anyone already rebooted after upstart upgrade?
[09:28] <asac> ;)
[09:28]  * asac scared about turning off this system
[09:28] <seb128> I did yesterday
[09:29] <asac> seb128: greasemonkey is still dead?
[09:29] <chrisccoulson> kklimonda - you still use autoreconf, but the changes go in to a patch (99_autoreconf) in this case
[09:29] <asac> seb128: thought it worked after you installed it in profile ;)
[09:29] <seb128> asac, the one installed via firefox says that it doesn't work with the current version
[09:29] <seb128> asac, well the error I got before restart came back a bit after restart and disabled it
[09:29] <seb128> the version not supported one
[09:29] <seb128> which is weird because it was working for a bit
[09:30] <asac> seb128: did you uninstall the package?
[09:30] <seb128> yes
[09:30] <seb128> I tried to reinstall it yesterday after uninstalling the web one
[09:30] <seb128> but it's not listed
[09:30] <asac> hmm
[09:30] <kklimonda> chrisccoulson: ~1mb 99_autoreconf is considered normal?
[09:30] <seb128> which is weird because the bug has been closed saying the new version should be working
[09:30] <asac> seb128: so after uninstalling package and uninstalling from firefox you were in a state where greasemonkey was completely gone?
[09:31] <seb128> kklimonda, diffstat it to be sure there is no cache there
[09:31] <seb128> asac, yes
[09:31] <kklimonda> :*
[09:31] <chrisccoulson> kklimonda - did you delete all the cruft from it? (autom4te.cache)
[09:31] <seb128> asac, then I did sudo apt-get install greasemonkey but that didn't make it listed again
[09:31] <asac> seb128: so you have the package installed again?
[09:31] <asac> hmm
[09:32] <asac> poor seb128
[09:32] <seb128> asac, I'm back to epiphany meanwhile
[09:32] <seb128> I need stock replies to triage bugs
[09:32] <asac> yeah
[09:32] <seb128> dunno how you handle those without stock replies ;-)
[09:32] <seb128> it's so quicker to click on a "need info" button
[09:32] <seb128> or "need stacktrace"
[09:33] <asac> you know my approach to bugs ;)
[09:33] <seb128> hehe
[09:33] <asac> i had script ones ... those were even better than stock replies
[09:33] <seb128> anyway I would appreciate if you bumped this greasemonkey fixing if that's not only me having the issue
[09:33] <asac> seb128: but actually. since we debugged it i always have your stockreplies ;)
[09:33] <seb128> I expect that will impact on the team efficiency
[09:33] <seb128> I know pedro use those a lot too
[09:33] <asac> seb128: i havent heard of anyone else having issues to be honest
[09:34] <seb128> ok, weird
[09:34] <asac> anyway. at least the package should get a regular maxversion bump
[09:34] <kklimonda> chrisccoulson: yes.. I get lots of changes in configure scripts..
[09:34] <asac> will prioritize this now that main is frozen
[09:35] <kklimonda> chrisccoulson: 20 files changed, 6387 insertions(+), 14382 deletions(-) vs 19 files changed, 1241 insertions(+), 2324 deletions(-) in your patch
[09:36] <chrisccoulson> that seems quite a lot :-/
[09:37] <seb128> happens easily when using different versions than upstream to do the update
[09:37] <seb128> that's not an issue
[09:37] <seb128> you can try running the same version of automake etc for the update to lower that
[09:41] <Amaranth> You guys have stock replies in a greasemonkey script?
[09:41]  * Amaranth dies
[09:41] <Amaranth> gimme!
[09:42] <kklimonda> btw, is it just me or is xplash similar to fedora's previous bootsplash that used X? I mean - won't we have similar problems with that (for example long start) ?
[09:42] <Amaranth> I went through 200 bugs once copy/pasting stock replies from the wiki before giving up on the idea
[09:42] <kklimonda> :0
[09:42] <kklimonda> :)
[09:42] <asac> Amaranth: http://people.canonical.com/~seb128/bugtriage.user.js ;)
[09:43] <Amaranth> kklimonda: No, rhgb started an X server just for the boot splash then killed it and started another one for you to login
[09:43] <pitti> Amaranth: they are awesome
[09:43] <Amaranth> kklimonda: xsplash runs as a part of gdm so when it finishes you'll looking at a gdm login window
[09:43] <pitti> Amaranth: https://launchpad.net/launchpad-gm-scripts
[09:43] <pitti> there's also a PPA of a packaged version, etc.
[09:43] <kklimonda> Amaranth: ach, that's why I haven't seen so much flickering :)
[09:43] <seb128> tseliot, where should crashes in nvidia libGL in karmic be reassigned?
[09:44]  * Amaranth has been pushing them to nvidia-glx-180
[09:44] <Amaranth> although I probably should have remembered which ones I did so I could mark dupes, I'm sure there were a few
[09:44] <seb128> Amaranth, https://edge.launchpad.net/launchpad-gm-scripts has a better version
[09:44] <seb128> mine is a old custom one
[09:45] <seb128> their one has button to add stock replies, etc on the webpage
[09:45] <seb128> so you can edit those easily
[09:45] <seb128> but it didn't work in epiphany so I kept using my old one ;-)
[09:45] <Amaranth> is software-sources supposed to add the PPA gpg key when you add a PPA?
[09:46] <seb128> mvo_, ^
[09:46] <seb128> Amaranth, http://bazaar.launchpad.net/%7Egm-dev-launchpad/launchpad-gm-scripts/master/files
[09:47] <Amaranth> yeah, I'm going to have to grab the file since my greasemonkey is from amo
[09:47] <kklimonda> cccCCCCC
[09:47] <kklimonda> oops, sorry
[09:47] <Amaranth> oh, it doesn't depend on the greasemonkey package
[09:48] <mvo_> Amaranth: yes, add-apt-repository ppa:compiz should add the key as well
[09:49] <Amaranth> mvo_: alright, was about to start looking up how to add it manually again :)
[09:49] <Amaranth> and this greasemonkey launchpad stuff is _so awesome_
[09:49] <Amaranth> I suddenly feel like doing a bunch of bug triage
[09:50] <Amaranth> wow, it even has one for valgrind
[10:05] <mac_v> pitti: hi... how do i request a UI freeze exception? i just subscribe the ~ubuntu-release team to the bug and add a comment about why this needs an exception, right?
[10:06] <pitti> mac_v: right, and also tell the doc team, to update screenshots etc.
[10:06] <mac_v> hmm.. , ok
[10:06] <pitti> (if they have some)
[10:07] <pitti> mac_v: usually it should be enough to subscribe them to the bug and get their ack there
[10:07] <mac_v> pitti: i dont think this  , Bug #409828 , needs the doc team..
[10:07] <mac_v> well anycase , i'll just subscribe them :)
[10:09] <mvo_> Amaranth: have you talked to maniac yet about the revert?
[10:09] <pitti> mac_v: should be fine, yes; but still better to sub them
[10:09] <Amaranth> mvo_: just via email, no reply yet
[10:09] <mvo_> Amaranth: cool, thanks
[10:10] <Amaranth> mvo_: since it won't get into alpha 6 anyway I figure we can afford to wait a bit and see if he can figure out what he would consider a proper fix
[10:13]  * mvo_ nods
[10:31] <seb128> does anybody get nothing on screen when booting with the newer linux version in karmic?
[10:34] <tseliot> seb128, Amaranth: nvidia-graphics-drivers-180
[10:34] <seb128> tseliot, thanks
[10:35] <seb128> I've reassigned to the glx package for now
[10:35] <seb128> will do that for the next one
[10:35] <tseliot> seb128: ok, thanks
[10:35] <Amaranth> seb128: if it's following what the ubuntu-boot packages were doing you're not supposed to see anything until xsplash
[10:35] <seb128> I tried -190 and -185 but launchpad said those were not published in ubuntu
[10:35] <Amaranth> usplash will only show up if you need to put in a password or a fsck is running
[10:35] <seb128> Amaranth, not doing something for over 30 seconds = very confusing
[10:36] <Amaranth> seb128: well X is supposed to start faster than that :)
[10:36] <seb128> I wonder how people can think that's a good idea
[10:36] <Amaranth> it's more like 10 seconds or so for me
[10:36] <seb128> well it's not
[10:36] <seb128> I've a slow laptop drive
[10:36] <Amaranth> I think Keybuk probably wants to make sure it is working like planned but will probably turn usplash back on for karmic if it doesn't get fast enough
[10:36] <Amaranth> seb128: I have a 5400rpm drive too
[10:36] <seb128> it's weird that it happens only with the new linux image
[10:37] <seb128> ie if I pick the previous one usplash is used
[10:37] <Amaranth> seb128: update-initramfs only updates the current kernel
[10:37] <seb128> shouldn't the initramfs be the same or updated for all images on upgrade?
[10:37] <Amaranth> seb128: if you force it to update all of them usplash will go away there too
[10:37] <seb128> hum ok
[10:37] <Amaranth> I think an update to initramfs updates all of them but updates to things in the initramfs only update the current one
[10:38] <seb128> I'm wondering how many people will think their machine crash and do ctrl-alt-del before getting xsplash
[10:38] <Amaranth> but maybe they all just update the current one which would make sense if you want to use the old one if something goes wrong
[10:39]  * tseliot can't even see the gnome-panel if he enters GNOME with nvidia. The xterm session works well though
[10:39] <Amaranth> tseliot: do you see the shadow for the panel?
[10:39] <tseliot> Amaranth: no, sometimes I see an empty panel at the top
[10:40] <Amaranth> ah
[10:40] <tseliot> but that's it
[10:40] <Amaranth> I tend to get an empty panel when gpm is acting up
[10:40] <Amaranth> Until gnome-panel either gives up on gpm or gpm starts working
[10:40] <seb128> xsplash starts around 32.8 seconds on this bootchart there
[10:40] <tseliot> I don't get anything else though
[10:40] <Amaranth> seb128: ouch
[10:40] <seb128> and that's a 7200rpm drive
[10:40] <seb128> ie not the slowest disk you can get
[10:41] <Amaranth> seb128: that sounds like a bug
[10:41] <seb128> hum compiz decorator crashed again right now
[10:41] <Amaranth> seb128: it's more like 15 seconds for me with a 5400rpm drive and sreadahead not actually working
[10:42]  * Amaranth wonders if anyone ever figured out what "exe" was in bootchart
[10:43] <Amaranth> stalls my boot about 4 seconds doing a crap load of IO before sreadahead can even start
[10:44] <Amaranth> seb128: is this something you can reproduce or get a lot?
[10:44] <seb128> Amaranth, what? the compiz issue? or the boot one?
[10:44] <Amaranth> compiz
[10:44] <seb128> looking at bootchart I've seen only it's as slow for other people
[10:44] <seb128> ie on pitti's one xorg starts after 35 seconds
[10:45] <seb128> xsplash is around 45s in boot
[10:45] <Amaranth> pitti seems to have an overall terrible boot time as well
[10:45] <Amaranth> heck with a clean install of jaunty my boot time was floating between 14 and 17 seconds so I thought my karmic times were terrible
[10:46] <Amaranth> then I see yours :P
[10:46] <Amaranth> jcastro is booting in 4 seconds though
[10:46] <pitti> yeah, grub->gdm is 65 seconds
[10:46] <pitti> before the upstartification it was 50
[10:47] <Amaranth> If you want better boot times steal his laptop at UDS :P
[10:47] <seb128> one minute of no screen = computer crashed for most users
[10:47] <seb128> http://people.canonical.com/~seb128/seb128-laptop-karmic-20090916.png
[10:47] <seb128> is my current bootchart
[10:47] <Amaranth> seb128: Like I said, I'm sure Keybuk won't leave usplash disabled for karmic, it's just a part of testing
[10:47] <pitti> seb128: I think it's now missing to load the fbcon module
[10:47] <pitti> so that you don't have text consoles with KMS
[10:48] <seb128> going through the udev calls and disk mounts takes almost 15 seconds
[10:48] <Amaranth> Yeah, it should at least show the couple of errors messages from the kernel everyone seems to get
[10:48] <seb128> pitti, ah ok
[10:48] <pitti> seb128: I asked Scott about it
[10:48] <seb128> alsactl = 5 seconds
[10:48] <pitti> for me the screen goes to standby during boot
[10:48] <seb128> xorg = 7 seconds
[10:49] <Amaranth> wow, you must have some crazy hardware
[10:49] <seb128> pitti, seems to do the same there
[10:49] <pitti> with everything starting in parallel, stuff just trashes
[10:49] <seb128> Amaranth, dell d630
[10:49] <Amaranth> that's a lot of calls to blkid and modprobe at the start
[10:49] <pitti> hard disks just aren't fit to load things in parallel
[10:49] <Amaranth> and you don't seem to have sreadahead
[10:49] <seb128> with a 7200rpm drive and some gigas of ram
[10:49] <pitti> Amaranth: I do
[10:49] <seb128> ii  sreadahead                                 1.0-3~boot3                                      Read required files in advance during boot
[10:50] <Amaranth> seb128: Oh, I see it now
[10:50] <Amaranth> seb128: but it must not have any profile data because it only runs for about half a second
[10:50] <Amaranth> probably less than that
[10:51] <seb128> Amaranth, it's installed for ages so that's weird
[10:51] <seb128> and I did reboots without changing packages today
[10:51] <Amaranth> pitti: ideally with sreadahead and careful ordering of startup tasks you could minimize the disk abuse
[10:51] <Amaranth> seb128: yeah, sreadahead seems broken in general so it's not surprising
[10:51] <Amaranth> for me sreadahead runs for the entire boot but never shows any IO or CPU usage
[10:51] <seb128> still I'm surprised you manage to get so fast loading
[10:52] <Amaranth> seb128: http://img193.imageshack.us/img193/9544/travislaptopkarmic20090.png
[10:53] <Amaranth> in that boot "exe" stalls things even longer than my latest chart actually
[10:53] <Amaranth> oh, that was before the "make it faster" bits landed
[10:54] <seb128> Amaranth, it takes you around 25 seconds to have xorg running though
[10:54] <seb128> not 15
[10:54] <Amaranth> right, that's out of date
[10:54] <Amaranth> that was before usplash went away
[10:55] <Amaranth> seb128: http://img34.imageshack.us/img34/9544/travislaptopkarmic20090.png
[10:55] <Amaranth> hmm, you know what, xsplash didn't even load
[10:55] <seb128> you have a fast disk
[10:56] <seb128> it tops to 37meg/s on the chart
[10:56] <seb128> mine is a 7200rpm and top at 15meg/s
[10:56] <Amaranth> guess I assumed it was just broken when I saw my mouse cursor
[10:56] <Amaranth> seb128: 7200rpm should be more like 50MB/s in a benchmark
[10:56] <Amaranth> my disk is only 5400rpm
[10:56] <seb128> well it's not
[10:56] <seb128> see my chart
[10:56] <seb128> 15mb/s
[10:56] <Amaranth> right, that's a problem with the boot, not your disk
[10:57] <Amaranth> seb128: sudo hdparm -t /dev/sda
[10:58] <Amaranth> actually mine is saying 50MB/s...
[10:58] <Amaranth> \
[10:58] <seb128>  Timing buffered disk reads:  160 MB in  3.03 seconds =  52.88 MB/sec
[10:59] <Amaranth> so it should be about as good assuming your disk has a good cache and decent seek times
[10:59] <Amaranth> but it is a dell so it probably has the worst you can get for both of those... :P
[10:59] <seb128> lol
[11:00] <seb128> in any case let's see where it goes
[11:00] <pitti> mine is 22 MB/s :-(
[11:00] <seb128> pitti, do you know if there is a bug about the screen being cut on current karmic?
[11:00] <pitti> seb128: "cut"?
 for me the screen goes to standby during boot
[11:01] <seb128> ^ that
[11:01] <pitti> ah
[11:01] <seb128> it does the same there, I did boot an old image
[11:01] <pitti> seb128: I asked Scott about it, waiting for his response
[11:01] <seb128> ok
[11:01] <seb128> at least he fixed the upstart restarting dbus which restarted all sessions this night ;-)
[11:02] <Amaranth> seb128: it looks like if you delete /var/lib/sreadahead/pack and reboot it should generate a new profile
[11:02] <Amaranth> that's what the monthly cron job does anyway
[11:03] <seb128> the pack there has been updated yesterday
[11:05] <Amaranth> weird
[11:05] <Amaranth> heh, it also looks like all 6 bugs open against sreadahead are the same bug
[11:07]  * tseliot 's gnome-session is busted. Xfce works great...
[11:08] <seb128> tseliot, how busted?
[11:08] <pitti> tseliot: ATI?
[11:08] <tseliot> pitti: nvidia
[11:08]  * pitti just stumbled over bug 423415 when testing the live system on wife's computer
[11:08] <Ng> sweet, suspend works again \o/
[11:08] <Ng> didn't lock my screen though, hmm
[11:08] <tseliot> seb128: I log in and I can see Ubuntu's default background after the splash screen
[11:08] <tseliot> seb128: nothing else
[11:08] <pitti> Ng: the mesa fix?
[11:09] <seb128> tseliot, gconftool-2 -R /desktop/gnome/session
[11:09] <Ng> pitti: I think it may have been, yeah. on resume I'd mostly just have a locked display
[11:09] <seb128> tseliot, could you run that and copy on pastebin?
[11:09] <Ng> but I couldn't always even persuade it to actually suspend, which was odd, sometimes it would just do... nothing
[11:09] <tseliot> seb128: sure
[11:10] <Ng> err, s/locked/hung/
[11:10] <tseliot> seb128: http://pastebin.ubuntu.com/271994/
[11:11] <seb128> tseliot, that looks correct, not this issue then
[11:11] <tseliot> pitti: the mesa fix simply corrects a calculation where the size of the delta of the buffer object was bigger than the whole object
[11:11] <seb128> tseliot, if you run gnome-session by hand is an anon GNOME login does it work?
[11:12] <Ng> did upstream remove the ability to disable bluetooth adapters from the applet?
[11:13] <tseliot> seb128: I can try
[11:13] <Amaranth> Ng: iirc it now requires something else
[11:13] <Amaranth> in the kernel or a dbus service or something, don't remember
[11:13] <Ng> Amaranth: oh, could be, bluetooth-properties says "Could not open RFKILL control device, please verify your installation"
[11:14] <Amaranth> ah, yep
[11:14] <tseliot> pitti: oh and the fix affected only intel (/src/mesa/drivers/dri/i965/brw_wm_surface_state.c)
[11:15] <Amaranth> Ng: http://mail.gnome.org/archives/desktop-devel-list/2009-July/msg00212.html
[11:15] <Ng> Amaranth: aha, so this is a permissions issue. I have /dev/rfkill, it's just root:root 644
[11:15] <Amaranth> Ng: udev to the rescue!
[11:16] <Ng> hehe
[11:16] <Amaranth> (please file a bug :)
[11:16]  * Ng looks fo... :)
[11:17] <Amaranth> Ng: thanks for reminding me about that, should save me some battery life
[11:18] <Ng> heh. I'd slightly forgotten about it because I'm using bluetooth a lot for internet access atm, but when I have a real connection again I'll want the battery life too ;)
[11:18]  * Ng can't find a bug, so files one.... against udev?
[11:20] <Amaranth> yay 55 wakeups a second
[11:21] <Amaranth> and most of that is wireless
[11:21] <Ng> nice
[11:21] <Ng> ok, filed as bug #430622
[11:23] <tseliot> seb128: I think you can ignore xfce warnings. I launched gnome-session from the xterm session: http://pastebin.ubuntu.com/272000/
[11:24] <seb128> tseliot, it seems a "compiz is crashing xorg" issue
[11:24] <seb128> it does the same on my ati card
[11:24] <Amaranth> down to 18 wakeups a second if I turn off wifi and unplug my mouse :)
[11:24] <seb128> does your session crash? ie back to gdm?
[11:24] <Amaranth> tseliot: nvidia bug
[11:24] <tseliot> seb128: is there a gconf key that I can use to make sure that compiz is disabled?
[11:24] <Amaranth> we used to have a failsafe GNOME session for just that reason but...
[11:25] <pitti> seb128: do you know a better bug for the ati crash than 423415?
[11:25] <seb128> tseliot, change /desktop/gnome/session/required_components/windowmanager
[11:25] <tseliot> seb128: no, it doesn't. It doesn't allow me to do anything else and I have to restart X
[11:25] <tseliot> Amaranth: it used to work with the same driver
[11:26] <Ng> Amaranth: huh, I didn't realise that gnome power manager can display wakup info now too
[11:26] <Amaranth> I don't even see compiz giving any errors
[11:26] <Amaranth> tseliot: If X crashes it's X or the driver
[11:26] <seb128> tseliot, could be bug #419126
[11:26] <tseliot> Amaranth: it doesn't crash X
[11:27] <tseliot> seb128: and it also happened before the upgrade to the latest mesa
[11:27] <davmor2> pitti: all the logs in the bug are the ones that bryce asked for
[11:27] <Amaranth> tseliot: the only other reason I can think of everything would suddenly fail to connect is if something used up all the available X connections
[11:27] <seb128> tseliot, right, I'm looking for a bug, it was already there a week ago
[11:27] <Amaranth> but that something would have to spawn like 500 windows
[11:28] <tseliot> Amaranth: heh, there's no reason why that should happen when the session starts
[11:28] <Amaranth> compiz opens one as a child of the root window so apps can get the WM name and one for each window you have open so...
[11:28] <davmor2> seb128: the ati bug start the day the ati driver was update.  Just before the release of alpha 5 for me
[11:28] <Amaranth> yeah
[11:29]  * tseliot tries to log in with metacity instead of compiz
[11:30] <Amaranth> Ng: ha, gnome-power-manager causes most of the wakeups when I have it showing wakeups :P
[11:30] <seb128> tseliot, bug #426159
[11:30] <Ng> Amaranth: haha
[11:30] <Amaranth> due to updating the screen so much
[11:30] <tseliot> seb128, Amaranth: using metacity seems to solve the problem. So it has to be compiz
[11:31] <seb128> tseliot, that was the same issue
[11:31] <Amaranth> tseliot: X crash can't be a compiz issue
[11:31] <seb128> tseliot, well, compiz crashes the x server, but that's a driver or xorg issue
[11:31] <seb128> whatever compiz is doing xorg should not crash
[11:32] <tseliot> Amaranth: well, X didn't crash. I could still move the mouse pointer and even see the xsplash animation. No trace of it in the log
[11:32] <seb128> tseliot, use #419126
[11:32] <Amaranth> as far as what compiz is doing there was only one core change I can think of that would even affect rendering (and thus the drivers) and that was to fix a memory leak with DRI2
[11:32] <Amaranth> tseliot: but the log you posted shows X crashing...
[11:32] <seb128> tseliot, are you sure? can you put your /var/log/Xorg.0.log.old online?
[11:32] <Amaranth> or getting stuck in an infinite loop
[11:32] <tseliot> unless X is trapping useful information
[11:32] <Amaranth> but if that was the cause I don't think the mouse would move
[11:33] <Amaranth> no, that's a lie, there are times X gets wedged like that and the mouse still moves
[11:33] <cassidy> seb128, hi. I'm aware that's probably a hot topic, but is there discussion to revert to the old behaviour regardind icons in menu and buttons in Ubuntu ?
[11:33] <cassidy> Empathy looks completely shit without icons
[11:33] <chrisccoulson> seb128 - do you want to subscribe desktop-bugs to libgdata bugs? one of the issues raised in the MIR was that it doesn't have anybody looking at that package yet?
[11:33] <Amaranth> cassidy: empathy should be fixed to show icons where they are appropriate
[11:33] <seb128> cassidy, that's a GNOME decision I'm not aware of any move to revert since mpt was in favor of that
[11:34] <seb128> chrisccoulson, sure
[11:34] <cassidy> Amaranth, it doesn't display them. But that's fugly
[11:34] <chrisccoulson> seb128 - thanks
[11:34] <cassidy> for example in context menu there is still the space of the icon
[11:34] <cassidy> but nothing in it
[11:34] <seb128> chrisccoulson, I doubt anybody read all the desktop bugs though so that makes virtually no difference but if some people feel better
[11:34] <Amaranth> I'd really like to see some icons in the System menu though...
[11:34] <seb128> cassidy, bug #407621
[11:35] <Zdra> seb128, the problem is nobody made fuzz about it, I was warned about Empathy being shit yesterday...
[11:35] <seb128> cassidy, feel free to comment there or to talk to mpt
[11:35] <cassidy> seb128, I know, was just wondering about the Ubuntu position regarding this decision
[11:35] <Zdra> seb128, really, ubuntu should revert to default
[11:35] <Amaranth> cassidy: yeah, there are plans to change GTK+ to not leave that gap when it is configured to hide most icons
[11:35] <seb128> cassidy, mpt was one of the drivers for the change
[11:35] <tseliot> seb128: http://paste.ubuntu.com/272005/ (the ddxSigGiveUp took place when I restarted X)
[11:35] <Amaranth> Zdra: The default is no icons
[11:35] <Zdra> seb128, that something that needs at least one cycle with all maintainers warned
[11:35] <seb128> ie the canonical design team seels to agree
[11:35] <Zdra> Amaranth, I meant old default
[11:36] <seb128> Zdra, right I agree they didn't communicate properly about that and I told it to vuntz and some other rt people by then too
[11:36] <Zdra> seb128, IMO it's not a question of good or bad, it is just that they made that without telling anyone and that result to crap applications because we didn't had time to fix them
[11:36] <seb128> Zdra, we might change back before karmic to fix it for next cycle
[11:36] <Amaranth> the only missing icon that annoys me all the time is totem not showing the icon for the show button
[11:36] <Zdra> seb128, that would be wize
[11:37] <cassidy> yeah if the change is good or not it's not the problem (for now). But it has been done in a shitty way in GNOME
[11:37] <Amaranth> Zdra: It was talked about a lot on d-d-l
[11:37] <cassidy> I discovered the result when switching to Karmic this we
[11:37] <seb128> I told so to andre and vuntz
[11:37] <cassidy> Empathy was looking related
[11:37] <seb128> Amaranth, there was no announce describing what to do though
[11:37] <cassidy> retarded :p
[11:37] <Zdra> Amaranth, Some maintainers have a live outside GNOME and don't read all of ddl
[11:37] <Amaranth> seb128: there was, about 20 replies into the thread about it on d-d-l :P
[11:38] <seb128> I asked mccann to write an email about that to d-d-l, he said he would be never did
[11:38] <seb128> Amaranth, the discussion was a flame about the change
[11:38] <Amaranth> I know, I was joking with that one
[11:39] <seb128> Amaranth, not a "applications should set the always show property for object, objects being ..."
[11:39] <Amaranth> although there was a discussion about the change too
[11:39] <Amaranth> I believe the metacity maintainer started the whole thing
[11:39] <seb128> well, nobody announced maintainers they have to review their icon use
[11:39] <seb128> that was just discussed as a setting change
[11:39] <Zdra> seb128, see http://mail.gnome.org/archives/desktop-devel-list/2009-September/msg00086.html
[11:39] <seb128> Zdra, thanks
[11:40] <seb128> tseliot, oh, you are using an nvidia card?
[11:40] <seb128> tseliot, so it's not the same as the ati bug ;-)
[11:40] <seb128> tseliot, do you get the issue using nv?
[11:41] <Amaranth> r600 reports it isn't using the software renderer?
[11:41] <Amaranth> didn't think r600 3d support was that far along
[11:41] <tseliot> seb128: no, it works well with nv (which doesn't work with compiz)
[11:42] <Amaranth> at beta time someone should open a bug for each pciid that compiz makes X crash for so we can fill up the blacklist
[11:42] <Amaranth> crappy fix but we've had to do it before
[11:42] <tseliot> Amaranth: weird, maybe it's because of the latest mesa
[11:42] <tseliot> mesa/drm
[11:43] <Amaranth> tseliot: yeah, seems we have enough new mesa/drm to get 3d support that instantly crashes X
[11:43] <Amaranth> going to be a PITA blacklisting every single Radeon HD card
[11:43] <Amaranth> probably easy to make mesa/drm claim to not support any acceleration on those
[11:44] <Amaranth> easier*
[11:44] <tseliot> Amaranth, pitti, Ng: I think the -ati driver is still using xaa. Maybe I should have a look at some X logs
[11:44] <Amaranth> oh man, I hope not
[11:45] <Amaranth> I think the X patch to disable XaaNoOffscreenPixmaps was dropped when intel stopped using XAA
[11:45] <Amaranth> I was about to drop the corresponding patch to trigger that from compiz
[11:45] <tseliot> weird, apw's card was using xaa. The one in the bug report is using exa though
[11:46] <Amaranth> he probably did that manually
[11:47] <tseliot> it could be
[11:47] <Amaranth> yeah, the patch to make compiz and XAA work well together was dropped from X so if you use XAA you either get weird rendering glitches with some (mostly Qt) apps or you enable XaaNoOffscreenPixmaps which basically disables XAA and does pure software 2D rendering
[11:48] <pitti> tseliot: I just started kubuntu-netbook-remix, and it works there; I guess compiz triggers something which makes it crash?
[11:48] <Amaranth> pitti: is that a Radeon HD?
[11:48] <pitti> Amaranth: yes, 6xx
[11:48] <seb128> same here
[11:48] <pitti> i. e. not "fully" supported by -ati yet
[11:48] <Amaranth> yes, compiz tries to use the completely broken 3D acceleration for r600 we seem to have picked up
[11:49] <Amaranth> I thought the r600 3D stuff was a part of the radeon KMS stuff
[11:49] <Amaranth> oh, we have that, don't we
[11:49] <seb128> Ng, bug #430240 btw
[11:49]  * tseliot wonders if 3D can be disabled (in the driver) for such cards
[11:49] <soren> I'm seeing the same symptoms as bug #412619. Does anyone want me to try something?
[11:50] <Amaranth> tseliot: whatever glxinfo checks to get the OpenGL renderer string needs to return "Software Renderer" for those cards
[11:50] <Amaranth> tseliot: I guess that'd be mesa
[11:50] <seb128> soren, no
[11:50] <tseliot> pitti, seb128: do 3D apps work?
[11:50] <soren> seb128: Alright.
[11:50] <Amaranth> the bug they linked to says glxgears crashes X too
[11:50] <tseliot> Amaranth: I think it should be there. I need to check
[11:51] <Amaranth> I didn't think the r600 stuff was much past making glxgears work in any branch of the code
[11:51] <pitti> tseliot: no hw acceleration, I think; but I just tested video
[11:51] <Amaranth> pitti: try glxgears
[11:51] <pitti> tseliot: if nothing else helps we can blacklist the chips in compiz
[11:51] <Ng> seb128: oh bah, sorry. Can I dupe that to mine?
[11:51] <seb128> Ng, sure
[11:51] <Amaranth> pitti: There are a _lot_ of Radeon HD chips...
[11:52] <pitti> Amaranth: will do next time I boot a test system on that one (my wife currently needs her computer, sorry)
[11:52] <pitti> Amaranth: I guess r5xx and below will work
[11:52] <Amaranth> pitti: we'd have to blacklist every pciid fglrx claims to support
[11:52] <seb128> will try too later, I'm on my laptop right now and I've to go for lunch
[11:52] <Amaranth> but we'd have to implement per-driver blacklists first
[11:52] <pitti> Amaranth: oh, that's not fglrx, it's the free one
[11:52] <seb128> why fglrx?
[11:52] <pitti> (live system)
[11:52] <seb128> I'm using ati
[11:52] <Amaranth> pitti: fglrx only supports Radeon HD
[11:53] <Amaranth> it's an easy source of the pciids we need to blacklist
[11:53] <seb128> oh ok
[11:53] <seb128> lunch, bbl
[11:53] <Amaranth> but if we blacklist them you won't get compiz when you install fglrx either
[11:53] <Amaranth> since the blacklist is currently global, not per-driver
[11:53] <tseliot> it looks like r600 was enabled in mesa 7.6.0~git20090817.7c422387-0ubuntu1
[11:53] <Amaranth> like I said, probably easier to disable 3D acceleration in mesa
[11:53] <pitti> Amaranth: eww
[11:54] <Amaranth> pitti: we've only had one other time we needed blacklists per-driver
[11:54] <pitti> Amaranth: ok, I'll boot it again to confirm; I was pretty sure that in alpha-3 and jaunty it just used software 3D
[11:54] <Amaranth> back when -ati had problems with laptops and installing fglrx made it work right
[11:54] <tseliot> Amaranth: yes, that should be easier
[11:55] <Amaranth> iirc we ended up just blocking all laptops using the -ati driver instead
[12:00] <pitti_kne> Amaranth: ok, I'm in Kubuntu Netbook, on ATI
[12:00] <Amaranth> pitti_kne: alright, try glxinfo | grep renderer
[12:00] <pitti_kne> it's an RV630
[12:00] <pitti_kne> meh, no glxinfo
[12:01] <Amaranth> hehe
[12:01] <Amaranth> you'll need glxgears too while you're installing things
[12:01]  * pitti_kne installs mesa-utils
[12:01] <Amaranth> ah, right, all in one package
[12:02] <pitti_kne> OpenGL renderer string: Mesa DRI R600 (RV630 9587) 20090101 AGP 8x x86/MMX+/3DNow!+/SSE2 TCL
[12:02] <pitti_kne> right, so it's 3D now
[12:02] <Amaranth> pitti_kne: ok, that's our problem right there
[12:02] <Amaranth> pitti_kne: try glxgears
[12:02] <pitti_kne>  glxgears
[12:02] <pitti_kne> IRQ's not enabled, falling back to busy waits: 2 16
[12:02] <pitti_kne> no rrb
[12:02] <pitti_kne> drmRadeonCmdBuffer: -22
[12:03] <pitti_kne> and immediately quits
[12:03] <Amaranth> alright, so it didn't crash X but it also failed somewhat spectacularly
[12:03] <Amaranth> that's not surprising, glxgears uses almost no 3D stuff compared to compiz
[12:03] <pitti_kne> Amaranth: it might crash compiz, though
[12:03] <Amaranth> I don't think we've found an OpenGL extension we don't like yet ;)
[12:03] <pitti_kne> if only I had VTs on the Ubuntu live system..
[12:04] <pitti_kne> then I could switch to metacity and get some logging
[12:04] <pitti_kne> but it's completely busted
[12:04] <pitti_kne> X keeps restarting itself, and VTs are broken
[12:04] <Amaranth> wouldn't help anyway, if gdb stops compiz your X is completely frozen
[12:04] <Amaranth> you have to ssh in and run gdb remotely
[12:04] <pitti_kne> Amaranth: well, I could at least check .xsession-errors and the like
[12:05] <pitti_kne> Amaranth: no ssh on the live system
[12:05] <pitti_kne> hm, I could try to boot it with "text"
[12:05] <Amaranth> you could install openssh-server
[12:05]  * pitti_kne checks if the upstart job still respects that
[12:05] <pitti_kne> Amaranth: if I could type that command, I wouldn't need it in the first place :-)
[12:06] <Amaranth> you don't get a terminal in the netbook remix?
[12:06] <pitti_kne> Amaranth: I do; I mean in UBuntu with compiz
[12:06] <pitti_kne> I cuold try installing compiz onKubuntu, of course
[12:06] <Amaranth> oh, just install compiz and ssh on the netbook remix
[12:06] <Amaranth> if it's an actual netbook RAM could get a bit tight
[12:07] <kwwii> pitti_kne: hey, I wanted to install karmic so I can show my team-mates the xsplash and such...is the artwork already included?
[12:10] <Laney> is zenity in bzr?
[12:11] <pitti_kne> kwwii: just the old one that we carry for some weeks
[12:12] <pitti_kne> Amaranth: nah, it's 3 GB RAM, Athon 3000, pretty beefy machine :)
[12:13] <davmor2> pitti_kne: if you want vt on ubuntu live try removing the quiet and usplash from the grub line and then as soon as the desktop appears hit your vt of choice that's how I got the info for bryce
[12:14] <pitti> meh, and of course switching to text VTs broke X
[12:14] <davmor2> not desktop xsplash sorry
[12:14] <pitti> davmor2: ok, will try that; easier than installing everything all over again in the next KNE boot
[12:15] <pitti> meh, /etc/init/gdm.conf doesn't check for "text" any more
[12:18] <kwwii> pitti_kne: hrm, is there any way for me to show the team the current status of karmic, only with the new artwork?
[12:27] <pitti> kwwii: sorry, pitti_kne is dead
[12:27] <pitti> kwwii: I don't have the new artwork
[12:27] <pitti> kwwii: for xsplash you mean? I just have some new icons for usplash
[12:27] <pitti> Amaranth: ok, I reproduced the compiz failure in a controlled fashion
[12:27] <kwwii> pitti: yes
[12:28] <pitti> Amaranth: using metacity, compiz --replace, got X crash
[12:28] <Amaranth> pitti: alright, to see what compiz is doing right before X crashes you'll need to ssh in
[12:28] <Amaranth> wait, compiz isn't crashing so gdb won't help...
[12:28] <pitti> Amaranth: I redirected out and err to a file
[12:29]  * pitti walks over and pastes it into the bug
[12:29] <Amaranth> I don't think it spits out any useful messages either
[12:34] <pitti> Amaranth: sent to the bug
[12:34] <pitti> Amaranth: right, it doesn't, but for teh record
[12:35] <pitti> Amaranth: want me to try anything else?
[12:35] <pitti> the "gconftool" is an issue
[12:35] <Amaranth> pitti: does your Xorg log have any useful info?
[12:35] <pitti> seb128: was that compat name dropped deliberately? (only gconftool-2 now), or was that an accident?
[12:35] <pitti> Amaranth: no, nothing; no new entries when it's crashing
[12:36] <Amaranth> hrm
[12:36] <pitti> Amaranth: https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/423415/comments/16
[12:36] <seb128> pitti, looking
[12:36] <Amaranth> dbgsym packages for X would probably be too much even if you do have 3GB RAM
[12:36] <pitti> seb128: it's easy to fix compiz to use gconftool-2, but I already fixed it somewhere else, too
[12:36] <pitti> Amaranth: what would that help?
[12:36] <pitti> it's not a crash in the SEGV sense
[12:37] <pitti> otherwise it should be in the X log, and apport should catch it, too
[12:37] <Amaranth> what else could it be?
[12:37] <seb128> pitti, the postinst still has code to install the alternative
[12:37] <pitti> Amaranth: I don't know..
[12:37] <Amaranth> I can't think of why it'd decide to X normally
[12:37] <seb128> seems to be a bug if it doesn't work
[12:37] <Amaranth> err, exit
[12:37] <pitti> seb128: do you have it?
[12:37] <Amaranth> and if it did there would be something in the log about that
[12:37] <seb128> pitti, yes
[12:37] <Amaranth> I don't have gconftool anymore
[12:38] <pitti> Amaranth: so I thought; unfortunately the old ~/xsession-errors gets killed with the respawn
[12:38] <Amaranth> I didn't even know it existed, I've always used gconftool-2
[12:38] <seb128> /usr/bin/gconftool -> /etc/alternatives/gconftool
[12:38] <seb128> /usr/bin/gconftool-2
[12:38] <pitti> Amaranth: let me try without gdm and startx, that should retain .xsession-errors
[12:38] <seb128> in gconf2 binary
[12:38] <pitti> I don't have /usr/bin/gconftool
[12:38] <pitti> brb
[12:42] <seb128> pitti, the gconf2 postinst has
[12:42] <seb128> "if [ "$1" = install ]; then
[12:42] <seb128>     install_alternative
[12:42] <seb128> fi
[12:42] <seb128> if [ "$1" = configure ] && dpkg --compare-versions "$2" lt-nl 2.19.1-2; then
[12:42] <seb128>     install_alternative
[12:42] <seb128> fi
[12:42] <seb128> "
[12:43] <pitti> Amaranth: ah, https://bugs.edge.launchpad.net/ubuntu/+source/xserver-xorg-video-ati/+bug/423415/comments/17
[12:45]  * davmor2 hugs pitti
[12:45] <davmor2> bugs been doing my head in for a fortnight :)
[12:45] <pitti> seb128: that looks weird
[12:46] <pitti> seb128: "install" isn't even a legal postinst argument, is it?
[12:46] <pitti> seb128: and why woudl it install the alternative only for these old packages? (except if it really was deliberate)
[12:47] <seb128> pitti, that comes from debian and I'm not sure
[12:57] <Amaranth> pitti: that appears to die as soon as it calls glxinfo
[12:59] <Amaranth> first error is on `/usr/bin/glxinfo | grep -c GLX_EXT_texture_from_pixmap`
[12:59] <pitti> Amaranth: I'll start it again and try that manually; it's still running
[13:00]  * pitti walks over
[13:00] <Amaranth> but then on the second round with LIBGL_ALWAYS_INDIRECT set it correctly says there is no support
[13:00] <pitti> ah, sorry, not running any more; froze again after vt switch
[13:00] <Amaranth> pitti: don't bother
[13:00] <pitti> Amaranth: ok, then I'll bother about lunch :)
[13:01] <Amaranth> no matter what information we get the answer is to disable this acceleration code in mesa so compiz doesn't even try to run
[13:13] <seb128> hey MacSlow
[13:13] <seb128> MacSlow, any luck on those notify-osd crashers?
[13:13] <MacSlow> seb128, still on the mem-leaks
[13:13] <MacSlow> seb128, but the crasher are on my plate for this week too
[13:14] <seb128> MacSlow, don't you think crashes should be fixed first?
[13:20] <Amaranth> mvo_: the invisible panel bug seems to go away if I do a local rebuild of compiz
[13:22] <Amaranth> What exactly was the problem with the buildds?
[13:23]  * tseliot is building mesa without r600
[13:23] <mvo_> Amaranth: uh, that is strange
[13:23] <Amaranth> very
[13:23] <Amaranth> me and maniac just went over the code and couldn't figure out why it was not working right
[13:24] <Amaranth> everything I suggested was shot down except for "stack corruption"
[13:24] <Amaranth> because building with a printf made the bug go away
[13:24] <Amaranth> so I tried again with a fresh apt-get source compiz and built it without making any changes
[13:26] <mvo_> hmmm
[13:26] <mvo_> and now its gone everytime you try?
[13:34] <Amaranth> mvo_: well all 5 times I tried so far anyway
[13:34] <Amaranth> with the package from the repo it happens every single time
[13:34] <mvo_> how odd
[13:34] <Amaranth> yeah
[13:35] <Amaranth> I don't know if it was just something weird with the buildd and rebuilding will help or if it's an environment issue
[13:35] <Amaranth> guess I can setup pbuilder
[13:37] <seb128> try a ppa it's faster ;-)
[13:38] <Amaranth> not if someone is doing OOo uploads
[13:38] <cassidy> seb128, would be cool to fix https://bugs.edge.launchpad.net/ubuntu/+source/farsight2/+bug/429927  bigon did the merge
[13:38] <seb128> cassidy, it's planned but we are frozen for alpha6 today
[13:38] <seb128> I will upload tomorrow
[13:39] <cassidy> thanks
[13:47] <pitti> tseliot: mesa without r600> in a PPA?
[13:48] <tseliot> pitti, seb128: mesa without r600 should be here (as soon as it's built): https://launchpad.net/~albertomilone/+archive/x-testing
[13:48] <pitti> tseliot: ok, I'll try that on the live system then
[13:48]  * pitti hugs tseliot
[13:48]  * tseliot hugs pitti (thanks for testing)
[13:48] <pitti> Amaranth, tseliot: do you know of anything which we could grab as debug information to fix that properly before disabling and forgetting it?
[13:49] <Amaranth> I suppose it should be saying no, there is no tfp instead of crashing X...
[13:49] <pitti> Amaranth: you said that it crashed on calling glxinfo? that worked on KNE, but perhaps the ubuntu build has a different set of package versions; I'll try it again
[13:50] <Amaranth> but since the r600 code isn't even supposed to run glxgears yet I don't think we should care
[13:50] <tseliot> pitti: maybe updating drm, mesa (and kernel for KMS) should fix it (but might cause regressions to any other driver)
[13:51] <Amaranth> not worth it since the 3d support for r600 is worthless right now
[13:51]  * tseliot would rather not run the risk
[13:52] <Amaranth> mvo_: it works fine if I build with pbuilder too
[13:52] <Amaranth> perhaps it is related to the buildd problems
[13:52] <Amaranth> I guess after the alpha 6 freeze a no change rebuild should fix it
[13:52] <pitti> tseliot: I think I'll mark bug 423415 as a dupe of bug 419126
[13:53] <tseliot> Amaranth: yes, I meant updating drm, mesa, etc. to experimental code. In other words, I was saying no ;)
[13:53] <Amaranth> tseliot: experimental code that is most likely out of date compared to the git snapshot of mesa we have now :P
[13:54] <tseliot> heh
[13:54] <Amaranth> so either merge the branch before making a snapshot or lose all fixes we've gotten
[13:54] <Amaranth> yeah, not a good idea
[13:54] <tseliot> pitti: yes, it's the same bug
[13:55]  * tseliot nods
[13:55]  * pitti -> AFK for some more CD testing
[14:34] <tseliot> pitti, seb128: the mesa packages were built and are now ready for testing
[14:35] <tseliot> Ng: ^^
[14:35] <pitti> tseliot: ah, will do; great!
[14:35] <rickspencer3> seb128, pitti, tseliot, kenvandine,  what's the word on the street?
[14:35] <pitti> rickspencer3: b0rkage
[14:35] <rickspencer3> how's A6 looking?
[14:35] <seb128> rickspencer3, desktop or ubuntu?
[14:35] <rickspencer3> pitti, I was afraid of that
[14:35] <kenvandine> yo  yo
[14:35] <seb128> rickspencer3, the boot changes broke a lot of things
[14:35] <seb128> rickspencer3, desktop looks alright...
[14:36] <rickspencer3> seb128, does it boot?
[14:36]  * rickspencer3 is downloading Kubuntu Netbook ISO right now
[14:36] <seb128> it does for me yes
[14:36] <tseliot> rickspencer3: 419126 doesn't help but hopefully my change in mesa will prevent it from happening
[14:36] <seb128> not sure of other people
[14:36] <rickspencer3> bug #419126
[14:36] <rickspencer3> nice
[14:36] <rickspencer3> I have been worried that Ubuntu would fall over after all the A6 updates :)
[14:37] <tseliot> hehe
[14:37] <seb128> brb
[14:37] <pitti> rickspencer3: bug 430611 is currently the biggest bug, I think
[14:37] <rickspencer3> pitti, my, that is problematic indeed
[14:38] <rickspencer3> I suppose Keybuk is working feverishly at bug whack-a-mole
[14:38] <pitti> right
[14:41] <Keybuk> nobody mentioned that bug to me ;)
[14:41] <kenvandine> bratsche, that is the bug you hit
[14:41] <Keybuk> oh
[14:41] <Keybuk> maybe they did, just not by number
[14:41] <Keybuk> is that the /var on separate partition one?
[14:41] <bratsche> kenvandine: Thanks.
[14:42]  * kenvandine avoids updating this morning
[14:42] <dobey> pitti: care to sponsor a couple packages for me? :)
[14:42] <pitti> dobey: please just sub ubuntu-main-sponsors, I'll upload them after a6
[14:42] <bratsche> Yeah, I did an update yesterday and then the startup failed after it did a battery check or something.  So later last night I updated again and it got further, but then dbus wasn't starting.  That's where it's still at today.
[14:42] <davmor2> rickspencer3: doesn't ubuntu need to get up before a6 before it can fall over :D
[14:43] <dobey> pitti: after tomorrow? :(
[14:43] <pitti> tseliot: damn, doesn't seem to help :(
[14:43] <kenvandine> dobey, yes... frozen for a6 now
[14:43] <pitti> tseliot: I updated from the PPA, and still get the same loop
[14:43] <dobey> :(
[14:44] <pitti> tseliot: and froze again (tried to switch VT); anything I should check with those?
[14:45] <tseliot> pitti: make sure you don't have r600_dri.so
[14:45] <pitti> tseliot: where is that?
[14:45] <tseliot> pitti: /usr/lib/dri/r600_dri.so
[14:46] <pitti> tseliot: can I just rm that on the live system? or does your mesa package do other changes?
[14:46] <tseliot> pitti: try to remove it
[14:48] <pitti> tseliot: now compiz falls back to metacity
[14:49] <tseliot> pitti: which is exactly what we want. Now let me check why that file was still there
[14:50] <pitti> tseliot: followed up
[14:50] <pitti> tseliot: well, jaunty did have compiz on r6xx, I think?
[14:50] <dobey> these milestones make no sense to me then
[14:50] <pitti> dobey: why not?
[14:51] <tseliot> pitti: support for r600 and r700 is still experimental. Are you sure about that?
[14:51] <dobey> because the milestone's date doesn't match the freeze, which means there's a significant portion of the milestone that doesn't make sense because it's targetted at the release time, and not the freeze time
[14:51] <pitti> tseliot: not entirely; I can boot jaunty later to be sure, but I thought it had composite
[14:52] <pitti> dobey: well, you can't do a release in a snap of a finger; if the release is on Thursday, you need to freeze on Tuesday to be able to fix the breakage, build CDs, test them, etc.
[14:52] <tseliot> pitti: ok, please test it with jaunty and let me know. thanks
[14:53] <kwwii> seb128: on the latest karmic update the xdg places icons in the "places" menu shows some funky icons...all of them exist at 24x24 but it seems to pick two of them from svg....any idea why?
[14:54] <seb128> no
[14:54] <seb128> the icons seem fine there
[14:54] <seb128> do you have a screenshot?
[14:55] <dobey> pitti: yes. but the launchpad milestones should have their dates set to match the freeze times. it's confusing (and disconcerting) to me (as a new contributor), that i can't get major fixes in, because i missed the freeze which isn't set as a milestone in launchpad. i have to go look somewhere completely different for the freeze information.
[14:55] <kwwii> seb128: no worries, I just figured it out...hard to see but there is a naming problem with two of them, sorry to bother
[14:55] <dobey> pitti: and it makes it very difficult to target milestones correctly for other projects on launchpad, that we want to be in ubuntu
[14:55] <seb128> ok
[14:56] <seb128> dobey, btw about your sponsoring request from this week
[14:56] <seb128> dobey, when you ask for testing on the bug should sponsoring be blocked on testing results?
[14:56] <seb128> it was not clear to me so I didn't upload
[14:57] <dobey> seb128: yes. though that's probably not the best way to do things. not exactly sure at the moment what a better way to do it is though.
[14:58] <seb128> dobey, could you subscribe sponsors after getting testing feedback then?
[14:59] <dobey> seb128: sure. sorry for confusion. very stressed/in a hurry lately :-/
[15:00] <seb128> no problem don't worry it will just make things clearer next time
[15:01] <tseliot> pitti: ah, I've just found this in the debian/rules: 	DRI_DRIVERS += mach64 mga r128 r200 r300 r600 radeon s3v savage tdfx trident
[15:03] <tseliot> pitti: shall I bump the revision to 0ubuntu6 in the ppa or shall I use some ~xtesting?
[15:04] <pitti> tseliot: ok, I misremembered; no compiz love in jaunty
[15:04] <Zdra> I installed epiphany-webkit, but I can't find plugins
[15:04] <tseliot> :-/
[15:04] <Zdra> is there a package for them?
[15:04] <Zdra> epiphany-extensions seems to be for the mozilla based epiphany
[15:04] <pitti> tseliot: doesn't matter much; but 0ubuntu5+ppa1 would be better
[15:04] <pitti> tseliot: since the next regular ubuntu upload will be ubuntu5, the PPA should be smaller
[15:05] <tseliot> pitti: yes, I meant "+" instead of "~"
[15:05] <pitti> tseliot: I recommend using ubuntu5~ppa1, ~ppa2, etc.
[15:05] <pitti> but of course that's too late now for this ppa :)
[15:05] <tseliot> oh
[15:05] <tseliot> ok
[15:25] <seb128> Zdra, not yet I think, asac was going to look at it iirc but I'm not sure
[15:25] <Zdra> seb128, mozilla based is going to be the default?
[15:26] <seb128> Zdra, no, firefox is the default
[15:26] <Zdra> seb128, I mean for epiphany-browser
[15:26] <seb128> Zdra, I think asac wants to migrate epiphany users to epiphany-webkit for karmic but didn't yet
[15:27] <Zdra> IMO epiphany-browser should be wekbit, and add a epiphany-mozilla
[15:27] <asac> seb128: it breaks my heart
[15:27] <asac> seb128: i know we already said: yes. but here is another chance to say no :)
[15:27] <asac> no would come with universe demotion though
[15:28] <seb128> Zdra, do we need an epiphany-mozilla?
[15:28] <seb128> asac, I'm not sure why we have epiphany in main to be honest
[15:28] <Zdra> seb128, probably not
[15:28] <seb128> out of "it's GNOME"
[15:28] <asac> seb128: me neither ;)
[15:28] <seb128> but that add extra work for no benefit
[15:28] <seb128> and makes harder for motus to work on it
[15:28] <asac> seb128: can you demote it?
[15:29] <asac> Zdra: the transition is just adding the transitional packages to the webkit source and removing webkit
[15:29] <asac> from archive.
[15:29] <dobey> pitti, seb128: who's responsible for setting up the Ubuntu milestones/dates in LP for the development cycle? i'd like to involve whomever it is in discussion about improving them a bit, once i collect my thoughts and write something reasonable :)
[15:30] <Zdra> asac, makes sense
[15:30] <Zdra> asac, I hope it will be done for karmic :)
[15:30] <asac> Zdra: you really hope?
[15:30] <seb128> asac, I prefer not doing anything right now in case it's on DVD images or something, need to be checked with pitti or slangasek first I would say
[15:30] <asac> Zdra: are you a user speaking for webkit or epiphany?
[15:30] <asac> seb128: sure
[15:31] <seb128> dobey, the ubuntu drivers probably, pitti or slangasek should know
[15:31] <Zdra> asac, I like both :)
[15:31] <asac> Zdra: are you running the webkitversion?
[15:31] <dobey> seb128: thanks
[15:31] <asac> isnt it a major problem that there is no password management?
[15:31] <Zdra> asac, I tried, but with no plugin package, it is not usable
[15:31] <asac> (for your personal use?)
[15:31] <pitti> dobey: so far cjwatson did it, but claiming that alpha 6 was due yesterday would be wrong, too
[15:31] <Zdra> asac, adblock is mendatory
[15:31] <seb128> never used that
[15:32] <pitti> dobey: we announce freezes on planet, ubuntu-devel-announce, #ubuntu-devel, and fridge (I think)
[15:32] <seb128> but I use greasemonkey ;-)
[15:32] <dobey> pitti: development on alpha6 was due yesterday, no?
[15:32] <Zdra> seb128, epiphany's adblock is wonderful
[15:32] <asac> Zdra: how does debian maintain extensions for the gecko/webkit split?
[15:32] <pitti> dobey: by and large, yes; we are still landing changes to fix release critical bugs
[15:32] <Zdra> I never had to change rules, it just work out of the box
[15:32] <pitti> dobey: where did you look for the dates?
[15:32] <Zdra> asac, I have no idea
[15:32] <seb128> asac, the way we do I think
[15:32] <dobey> pitti: i'm looking at the milestone dates in launchpad
[15:32] <seb128> they didn't package the webkit ones yet
[15:33] <asac> allright
[15:33] <asac> thats what i thought ... just wanted to be sure
[15:33] <pitti> dobey: ah, "Expected: in 9 hours"?
[15:33] <pitti> heh
[15:33] <dobey> pitti: yes
[15:34] <asac> Zdra: so unless upstream fixed this very recently epiphany-webkit has no password management. how bad would that feel for an epiphany user like you?
[15:35] <pitti> dobey: well, IMHO it's not that critical that the updates don't land on the CDs; you hardly will do large and complex U1 operations in the live system, and people will upgrade after installation anyway
[15:35] <asac> e.g. if we decided to drop gecko and move everyone to webkit
[15:35] <dobey> and i'm not sure what qualifies as 'release critical' but current ubuntuone-client in the archive doesn't work.
[15:35] <dobey> i guess
[15:35] <Zdra> asac, I heard pwd is fixed now upstream, but I'm not sure
[15:35] <pitti> dobey: bug 426804?
[15:36]  * asac  checks
[15:36] <Zdra> asac, youtube and facebook works
[15:36] <Zdra> asac, so I'm happy
[15:36] <dobey> pitti: that's one of the issues, yes
[15:36] <Zdra> guess those are good test, if they work, everything works
[15:36] <pitti> dobey: so well, I can upload it, but can't guarantee that it will land on the CDs
[15:37] <dobey> pitti: well i've already come to terms with it not getting on there
[15:37] <asac> Zdra: not sure why you think that youtube and facebook cover all corner cases. probably a too simple pov ;)
[15:38] <dobey> pitti: but regardless, i'd like to discuss improving the milestone situation with whoever it is best to discuss it with :)
[15:38] <asac> Zdra: or are you saying that passwords work there?
[15:38] <dobey> hell
[15:38] <dobey> 'everyhthing' doesn't even work in firefox
[15:39] <pitti> dobey: hm, I'm not sure; adding a "freeze" date to the +milestone page, or just a string saying "freeze is usually two days before this" or so?
[15:39] <Zdra> asac, I didn't test, but fredp said that xan said that it is done
[15:40] <dobey> pitti: i think it's better to set the milestone date as the freeze date in LP. LP already has a separate 'releases' feature, so one should add releases for alphas/betas with the appropriate dates set there
[15:40] <fredp> Zdra, asac: http://git.gnome.org/cgit/epiphany/commit/?id=eea4e5155ed3102ee21c57958a29603ed0f5166f & http://blogs.gnome.org/xan/2009/09/08/the-show-so-far/
[15:40] <Zdra> Hm, it also changed my home page to debian.org... but that's not big issue
[15:40] <fredp> perhaps the package is missing a nss dependency?
[15:41] <pitti> dobey: that sounds weird, though; we shouldn't set the release date for the final to "two weeks earlier" either
[15:41] <dobey> pitti: no no. the release date is not the milsetone date
[15:41] <dobey> pitti: these are already separate concepts in launchpad :)
[15:45] <seb128> asac,
[15:45] <seb128> "epiphany-webkit (2.27.92-2) experimental; urgency=low
[15:45] <seb128>   * debian/control.in:
[15:45] <seb128>   - Added missing build-dep on libnss3-dev (used in the password importer)"
[15:46] <asac> semi-nice
[15:46] <asac> ;)
[15:46] <asac> let me check what they did for real
[15:46] <seb128> asac, it's not built since it dep-wait on a newer webkit that the one we have though
[15:46] <seb128> I've seen that you were discussing webkit with dholbach and other people the other day
[15:47] <seb128> so I stayed away from this one
[15:47] <dobey> pitti: but as i said, i still need to collect my thoughts fully. just wondering who i should ping to discuss further once i do
[15:48] <asac> seb128: that was about the ABI/API changes introduced by us
[15:48] <pitti> dobey: I'm not actually sure who sets the milestones in LP, but I suspect it's either slangasek or cjwatson
[15:49] <seb128> asac, right it was sort of looking like we could drop it and sync
[15:49] <seb128> asac, so I waited for somebody to open a sync request ;-)
[15:49] <asac> seb128: but since ephy-webkit password management is fixed (somehow) and they even migrate profile data
[15:49] <asac> we should go for that i think
[15:49] <seb128> +1
[15:49] <asac> i was just in doubt because of that
[15:49] <asac> oki
[15:49] <seb128> epiphany-gecko is not maintained anyway and if you want a gecko browser use firefox
[15:49] <asac> lets hope webkit doesnt bust gwibber or so
[15:49] <seb128> asac, can you look at the webkit update?
[15:50] <dobey> pitti: ok. i'll ping them and you as well when i collect thoughts and forumlate a mail to express them :)
[15:50] <pitti> thanks
[15:50] <asac> seb128: right. but i would hav ehated to force current ephy users to use firefox for one cycle (like you ;))
[15:51] <asac> seb128: yes. i do the webkit update. in worst case we just break anjal and see if they complain.
[15:51] <seb128> asac, I'm using firefox now and I somewhat like it ;-)
[15:51] <asac> imo they should put that into ppa anyway
[15:51] <seb128> asac, it doesn't crash every second closing which is a good win :-p
[15:51] <asac> seb128: yes. but thats not how its supposed to be ... also you have a bad experience ;) with greasmonkey breaking etc.
[15:51] <asac> never seen before issues :)
[15:52] <seb128> asac, I did reinstall the internet addon greasemonkey and it's working again, weird
[15:52] <seb128> asac, let see if it decided after a while that it's not compatible with firefox 3.5.3 again now
[15:52] <seb128> decides
[15:55] <tseliot> pitti: does 0ubuntu5+ppa1 solve the problem (I see no trace of the r600 module in the package)?
[15:56]  * asac off some NM testing
[16:02] <pitti> tseliot: I suppose, but I didn't test it
[16:05] <tseliot> pitti: also it's likely that editing the debian/rules is enough (as opposed to patching the source). I'm building the packages locally and I'll attach a debdiff to the bug report in any case
[16:05] <pitti> tseliot: right, if it can be done with a configure switch, that's better
[16:07] <tseliot> pitti: yes, I think the debian/rules overrides the list of modules in the configure.ac of the source
[16:09] <hyperair> hmmm i wonder if gdu will allow me to hot-unplug my hard disks
[16:09] <hyperair> IDE hard disks i mean. =D
[16:09] <Laney> good luck with that
[16:11] <hyperair> hahaha
[16:12] <hyperair> well at least my SATA disks can be hot unplugged
[16:20]  * Laney needs to go home and fix the karmic mess
[16:20] <Laney> someone come round and do it for me please :(
[16:20] <Laney> I offer cake!
[16:21] <pitti> Laney: good luck
[16:21] <pitti> Laney: dist-upgrade should do it
[16:21] <pitti> seb128: btw, with most recent karmic I don't get the "powers off screen on boot" issue any more
[16:22] <Laney> pitti: Gotta get into the system though, which is a pain
[16:23] <Laney> chroot from a live cd?
[16:23] <seb128> pitti, what did you upgrade?
[16:23] <pitti> seb128: I don't know, everything
[16:24] <pitti> I didn't see it on today's CD live systems either
[16:24] <pitti> this morning I still had yesterday evening's packages
[16:24] <seb128> same here
[16:24] <seb128> I will upgrade later
[16:24] <seb128> I want a working machine for now
[16:25] <chrisccoulson> i upgraded my desktop last night and then spent until 3am trying to get it to boot again ;)
[16:26] <Laney> chrisccoulson: how did it fail/
[16:26]  * Laney is in busybox land
[16:26] <chrisccoulson> it got just past the initramfs and hung at mountall
[16:26] <chrisccoulson> it turns out that some bindfs mounts in my fstab screw it up, so i just disabled them for now
[16:27] <Laney> I hope a dist-upgrade fixes it... I did actually plan on doing some sponsoring yesterday
[16:27] <Laney> after my PC was down for a week
[16:27] <chrisccoulson> but it took me a long time to figure that out. my keyboard doesn't work that early on sometimes
[16:27] <Laney> gnome-terminal was losing keypresses too even when it was working
[16:27]  * Laney is a bit frustrated
[16:28] <chrisccoulson> is it your only machine?
[16:28] <Laney> nah, got a laptop
[16:28] <Laney> it's my only Ubuntu machine though
[16:28] <chrisccoulson> heh, i only have 1 machine here
[16:29] <chrisccoulson> i think it's probably time i invested in a laptop now
[16:29] <chrisccoulson> i used to have a company laptop, but my current employer is too tight to give me one
[16:30] <Laney> I bought my own... and then a few months later work said that if I hadn't they'd have bought me one then
[16:31] <chrisccoulson> i don't think there is any chance of my employer doing that. i'd be lucky to get a pen off them ;)
[16:32]  * chrisccoulson thinks he should probably not be using the company network for IRC now
[16:32] <Laney> heh heh
[16:42] <chrisccoulson> i should probably do some sponsoring tonight really. i havent done any for a little while now
[16:48] <pitti> ubuntu-bug...
[16:48] <pitti> Gdk-ERROR **: The program 'firefox' received an X Window System error.
[16:48] <pitti> This probably reflects a bug in the program.
[16:48] <pitti> The error was 'BadWindow (invalid Window parameter)'.
[16:48] <pitti>   (Details: serial 523 error_code 3 request_code 20 minor_code 0)
[16:48] <pitti> meh
[16:49] <pitti> asac: any idea about that? This seems to happen randomly
[16:50] <asac> yes
[16:50] <asac> one second
[16:50] <asac> bug 401055
[16:50] <asac> i have a bandaid for this. and upstream supposely landed a right fix on trunk
[16:51] <asac> pitti: do you still see this?
[16:51] <asac> could you verify that its fixed in firefox-3.7 from the daily ppa?
[16:51] <asac> https://edge.launchpad.net/~ubuntu-mozilla-daily/+archive/ppa
[16:52] <asac> i need a confirm before requesting approval for the 3.6 and 3.5 branche
[17:02] <pitti> asac: seems I can currently reproduce it with ubuntu-bug with every other try
[17:03] <asac> with 3.7?
[17:05] <pitti> asac: no, with current karmic (3.5)
[17:05] <pitti> asac: can I install 3.7 in parallel? or should I backup my .mozilla?
[17:06] <asac> pitti: parallel should work. it copies your profile if all goes well ;)
[17:06] <asac> (backup is always good)
[17:07] <asac> just have to teach ubuntu-bug to use the firefox-3.7 command
[17:12] <mac_v> asac: hm... why isnt the bandaid applied for that? ;p
[17:13] <mac_v> i lost several bugs to that error , even if i have FF open :(
[17:15] <seb128> mac_v, to limit bug flood on launchpad? ;-)
[17:15] <asac> tough question. you can argue for both ways. but it was alpha stage so i thought better not hide it
[17:15] <mac_v> seb128: hehe ,  i want to flood with all known bugs ;)
[17:16] <mac_v> *unknown
[17:37] <superm1> kenvandine, how do you set third party backgrounds/logo's for xsplash?
[17:38] <superm1> it looks like it supports command line arguments, but it's not obvious where to use them
[17:41] <pitti> superm1: he's just working on a package split
[17:42] <pitti> superm1: there will be an ubuntu-xsplash-artwork package
[17:42] <pitti> superm1: which derivatives can just replace (same files)
[17:42] <superm1> pitti, ah okay
[17:42] <pitti> kenvandine: please add a Conflicts:/Replaces: xsplash-artwork to it
[17:42] <pitti> let's not go through the alternatives mess again like in usplash
[18:02] <chrisccoulson> seb128 - i'm glad i'm not the only one who can't confirm bug 427464 ;)
[18:02] <seb128> chrisccoulson, hehe
[18:22] <kenvandine> pitti, pushed, how does that look?
[18:22] <pitti> kenvandine: pushed where?
[18:22] <pitti> (and what?)
[18:22] <kenvandine> oh... i thought you looked at the branch :)
[18:23] <kenvandine> ubuntu-xsplash-artwork
[18:23] <kenvandine> i had a Provides and Conflicts
[18:23] <kenvandine> i thought you wanted me to add a Replaces too :)
[18:23] <kenvandine> ~ubuntu-desktop/xsplash/ubuntu
[18:24] <kenvandine> pitti, i read your comment earlier and thought it was feedback :)
[18:24] <pitti> kenvandine: ah, xsplash
[18:24] <pitti> sorry, too many things going on today
[18:24] <kenvandine> understand
[18:24] <kenvandine> i thought you went out and checked up on my work :)
[18:25] <kenvandine> pitti, when you can, please do go look at it
[18:25] <kenvandine> it isn't ready for upload yet though
[18:25] <kenvandine> we are doing a 0.8 tomorrow
[18:26] <pitti> kenvandine: please depend on ubuntu-xsplash-artwork | xsplash-artwork
[18:26] <pitti> kenvandine: there always must be a real preferred alternative
[18:26] <kenvandine> ok
[18:27] <pitti> kenvandine: and -artwork needs to conflicts/replaces: xsplash (<< 0.7.1-0ubuntu2) for the moved files
[18:27] <kenvandine> so you can use a | for or?
[18:27] <kenvandine> ok
[18:27] <kenvandine> actually 0.8
[18:27] <pitti> kenvandine: FYI, having "etc/" in debian/xsplash.install will probably suffice
[18:27] <pitti> kenvandine: | is "or", yes
[18:28] <kenvandine> i'll go ahead and bump version too
[18:28] <pitti> kenvandine: 0.7.1-0ubuntu2 will do even if you do a 0.8 later (that's also bigger than 0.7.1-0ubuntu2 :) )
[18:28] <pitti> kenvandine: likewise, just for future laziness you probably just want /usr/bin/ into xsplash.install (just in case there'll be more in the future
[18:28] <kenvandine> ok
[18:28] <pitti> kenvandine: dh_install gets along with directories just fine
[18:29] <pitti> kenvandine: so, with the fixed C/R and alternative dep it looks fine, thanks
[18:30] <kenvandine> actually... the version conflicts/replaces isn't useful... that package has never existed
[18:30]  * kenvandine puts it in for good measure though
[18:35] <pitti> kenvandine: sure it is; xsplash exists
[18:36] <pitti> kenvandine: the *new* package c/r's to the *old* (existing one)
[18:36] <kenvandine> humm... oh...
[18:36] <pitti> kenvandine: since u-x-artwork now ships files which are already shipped in xsplash
[18:36] <kenvandine> duh...
[18:36] <pitti> so you have to upgrade xsplash first before you can unpack u-x-artwork
[18:36] <pitti> and the C/R ensures that
[18:37] <kenvandine> yeah
[18:37] <kenvandine> good point
[18:37] <kenvandine> thx!
[18:37] <kenvandine> fixed
[18:38] <pitti> need to leave for TKD, cu tomorrow!
[18:39] <kenvandine> have fun!
[18:39] <kenvandine> good night
[18:45] <mvo_> hey rugby471 - many thanks for your branch
[18:45] <rugby471> mvo_: no problem
[18:58] <chrisccoulson> mvo_ - just looking at the changelog for compiz. was "unredirect fullscreen windows" disabled in jaunty too?
[18:59] <chrisccoulson> unfortunately, disabling it breaks user switching for nvidia users :(
[18:59] <mac_v> chrisccoulson: no , actually it is being done to prevent flicker from notify-osd ;)
[19:00] <chrisccoulson> ah, ok. that is one side effect of it. but i have to re-enable it again anyway ;)
[19:01] <chrisccoulson> maybe compiz could do a run-time check for people like me who have crappy drivers ;)
[19:01] <mac_v> dont we all have crappy drivers ;p
[19:02] <bratsche> pitti: ping
[19:02] <chrisccoulson> mac_v - i don't know. i'm not sure what other drivers are like. but this white-screen bug has existed for eternity in the nvidia driver, and will probably never be fixed in my lifetime ;)
[19:03] <bratsche> pitti: Nevermind
[19:03] <mvo_> chrisccoulson: oh? hm, that is bad. I guess we need to re-enable it then again, that is a pretty big use-case
[19:03] <mac_v> chrisccoulson: hehe , as an ATI user i thought nvidia guys had the best time of their life , \o/ good to know
[19:04] <chrisccoulson> mvo_ - i'm not sure really. i suppose it depends on how many nvidia users want user switching or not
[19:04] <chrisccoulson> the issue with it disabled is that you end up with a completely white screen on user switching
[19:05] <mvo_> chrisccoulson: many thanks - is there a bug open about this yet?
[19:05] <mac_v> chrisccoulson: could you comment that on Bug 346187
[19:05] <chrisccoulson> mvo_ - i think so. i'd have to try and find it though.
[19:06] <mac_v> that was the bug which "fixed" this issue
[19:06] <chrisccoulson> it would be a shame to break notify-osd appearance though if only a small percentage of the user base experience the nvidia bug on a regular basis
[19:06] <chrisccoulson> so i'm quite fine with just re-enabling it again ;)
[19:06] <chrisccoulson> (i mean, manually re-enabling it locally on my machine)
[19:08] <chrisccoulson> mvo_ - bug 160264 is the nvidia driver issue
[19:09] <mvo_> chrisccoulson: thanks
[19:10] <chrisccoulson> enabling unredirect fullscreen windows is only a workaround for the issue though - it doesn't fix it, and nvidia don't seem particularly interested in it
[19:10] <chrisccoulson> you still end up with windows appearing as white boxes if they are created on an inactive X session
[19:41] <mac_v> anyone notice , after the recent compiz update, that the window jumps up when closing?
[19:45] <seb128> no
[19:45] <mac_v> hmm... only certain windows consistently do that
[19:48] <mac_v> seb128: you are also using ATI ,right?
[19:48] <mac_v> do you have azureus installed?
[19:48] <mac_v> or vuze rather
[19:48] <seb128> no, intel on this box
[19:49] <mac_v> :(
[19:49] <seb128> the ati config crashes on compiz start
[19:49] <seb128> on what applications do you get the issue? where do they jump if they are closing?
[19:51] <mac_v> i think its related to an animation [burn], and it happens to the vuze window when i use close button on the top right to minimize to the notification area
[19:52] <mac_v> i noticed this on another app also but i'm just not able to recall :(
[19:57] <chrisccoulson> i've just tried gnome-shell for the first time on real hardware. i'm amazed at how painfully slow and jerky the animations are
[20:02]  * mac_v couldnt get gnome-shell to start :(
[20:02] <chrisccoulson> it starts ok here, but it's not very nice to use
[20:04] <mac_v> could some one confirm that Bug #423355 , is fixed? lp is lying as commited
[20:06] <chrisccoulson> thats fixed for me
[20:06] <mac_v> yup me too , thanks
[20:09] <hyperair> huh what happened to compiz O-o
[20:10] <hyperair> compiz.real: ../../src/display.c:793: updatePlugins: Assertion `j == pListCount' failed.
[20:11] <mac_v> compiz is doing all sorts of weird stuff! , the modal window for the root prompt started using animation after the recent updates
[20:12] <hyperair> ⇧..what?
[20:12] <hyperair> modal window for the root prompt?
[20:12] <hyperair> what modal window?
[20:13] <mac_v> hyperair: the window you get for root password prompts
[20:13] <mac_v> damn the window are jumping!
[20:14] <mac_v> hyperair: could you test this , open up the history from synaptic
[20:15] <mac_v> now when closing[from the button on the top right] the window do you see a short jump in the window?
[20:15] <hyperair> mac_v: you want me to test with compiz?
[20:15] <mac_v> when closing the history window*
[20:15] <hyperair> mac_v: i can't even START it!
[20:15] <hyperair> said assertion happens, then metacity kicks in
[20:15] <mac_v> oh ,
[20:15] <mac_v> thats bad!
[20:15] <hyperair> yes, that IS bad
[20:15] <hyperair> VERY bad
[20:15] <mac_v> lol
[20:16] <mac_v> chrisccoulson: could you test that^ with compiz ?
[20:16] <hyperair> there seems to be a bug on it
[20:16] <hyperair> but it's private so i can't see it =.=
[20:16] <hyperair> 429858
[20:16] <hyperair> #429858
[20:16] <hyperair> bug #429858
[20:17] <hyperair> see
[20:17]  * mac_v kicks the bot
[20:18] <chrisccoulson> mac_v - i can't recreate that. but then, i havent restarted my session since i upgraded
[20:18] <chrisccoulson> actually, i must have restarted compiz when i ran gnome-shell
[20:18] <chrisccoulson> so, i can't recreate it then
[20:19] <mac_v> hm.. it happens only for certain windows! and for those windows its consistent :(
[20:20] <mac_v> argh , it happens to all the synaptic child windows!
[20:22] <mac_v> and to gedit about window!
[20:22] <chrisccoulson> hyperair - you should be able to see bug 429858 now
[20:23] <chrisccoulson> mac_v - actually, i can get the window to jump
[20:23] <mac_v> \o/ i'm not see things
[20:23] <mac_v> seeing*
[20:24] <mac_v> chrisccoulson: which app? , i'v now found the terminal's about window jumps too
[20:25] <chrisccoulson> for synaptic, the modal dialogs get the wrong animation (as if it were a normal window)
[20:25] <chrisccoulson> i wonder if it has the wrong type hint?
[20:25] <chrisccoulson> xprop to the rescue
[20:26] <chrisccoulson> "_NET_WM_WINDOW_TYPE(ATOM) = _NET_WM_WINDOW_TYPE_NORMAL"
[20:26] <chrisccoulson> ^^^that is wrong, I think
[20:26] <mac_v> hmm...
[20:26] <chrisccoulson> it would be "_NET_WM_WINDOW_TYPE_DIALOG"
[20:27] <chrisccoulson> s/would/should
[20:27] <mac_v> which update did that change o.0
[20:27] <chrisccoulson> so, compiz gives it the wrong animation
[20:27] <chrisccoulson> mac_v - not sure. i don't open the synaptic preferences too often, so i don't know
[20:28] <chrisccoulson> what other windows do you see it with?
[20:29] <mac_v> the jump? > vuze main window , all synaptics child windows , gedit/terminal about windows
[20:29] <mac_v> and compiz's child windows
[20:30] <mac_v> Amaranth: is this a known issue ^ ?
[20:30] <chrisccoulson> yeah, i can trigger it with the about windows, but they get the correct animation, so the 2 issues are unrelated. i just noticed that the synaptic preferences gets the wrong animation by chance
[20:32] <mac_v> yeah the modal getting animations and the jumps are different issue
[20:36] <chrisccoulson> mac_v - bug 430955
[20:37] <Amaranth> jump?
[20:37] <mac_v> chrisccoulson: when you say > "modal Preferences dialog in Synaptic" , you mean the modal windows for all the gksudo apps right?
[20:38] <chrisccoulson> mac_v - no, i just mean the preferences window in this case
[20:38] <chrisccoulson> the authentication window is generally not modal, and is spawned in another process
[20:38]  * Amaranth gets lost, goes have have morning caffeine injection before trying to follow
[20:39] <chrisccoulson> mac_v - in the case of policykit prompts, they're displayed by polkit-gnome-authentication-agent-1 now
[20:39] <mac_v> chrisccoulson: the authentication window also gets an animation here
[20:39] <mac_v> policy kit prompts are not modal windows for me.
[20:39] <chrisccoulson> mac_v - compiz can animate all windows, but by default, only normal windows get the glide animation
[20:40] <chrisccoulson> other window types should be faded
[20:40] <chrisccoulson> although, you can set them all to glide;)
[20:43] <Amaranth> wait, is the problem windows getting the wrong animation?
[20:43] <mac_v> chrisccoulson: hm.. those modal prompts dont cause a problem for me , the modal where the screen greys out and displays the authentication window for gksudo apps is now animating for me
[20:44] <mac_v> the grey background animates!
[20:44] <chrisccoulson> yeah, i get that too
[20:44] <mac_v> Amaranth: the jump is noticed with these windows >  all synaptics child windows , gedit/terminal about windows ,  compiz's child windows
[20:45] <mac_v> could you test it ?
[20:45] <Amaranth> "the jump"? I hope that isn't a bad as "the clap"
[20:45] <chrisccoulson> it's worse;)
[20:45] <hyperair> seb128: gnome-shell needs libclutter-1.0-dev to run.
[20:45] <mac_v> Amaranth: when you close the windows using the top right close button , the window jump up and animate
[20:45] <seb128> hyperair, yes, I know
[20:45] <hyperair> seb128: basically the symlink for libclutter-glx.so
[20:45] <hyperair> ah
[20:46] <seb128> I think that's a gnome-shell bug
[20:46] <seb128> but I didn't track it yet
[20:46] <seb128> I don't want to add a depends that would be wrong
[20:46] <seb128> it should use the .so.soname
[20:46] <seb128> not the .so
[20:46] <Amaranth> wow, I thought my compiz was really broken
[20:46] <Amaranth> turns out I was running metacity
[20:46] <mac_v> lol
[20:47] <Amaranth> mac_v: yeah, I see that little jump
[20:47] <Amaranth> mac_v: ccsm's windows do it too?
[20:47] <mac_v> yeah
[20:47] <mac_v> ccsm's child windows
[20:47]  * Amaranth puts this just above making the logout plugin work on his todo list
[20:48] <Amaranth> :P
[20:48] <mac_v> ;)
[20:48] <seb128> what is the logout thing doing?
[20:48] <Amaranth> seb128: basically the same thing we used to do, it fades out the rest of the screen
[20:49] <seb128> is anybody using that?
[20:49] <Amaranth> no, mac_v wants us to
[20:49] <mac_v> ;p
[20:49] <mac_v> seb128: Bug #429132
[20:50] <mac_v> it works , but Amaranth is noticing some problems
[20:50]  * hyperair continues grumbling about compiz not starting
[20:50] <seb128> we should stop shipping useless options
[20:50] <Amaranth> yeah, like your match rule not working at all :P
[20:50] <Amaranth> mac_v: because I don't use FUSA
[20:50] <Amaranth> I figured it out :)
[20:51] <mac_v> i dont use FUSA too ;p
[20:51] <seb128> we get tons of bugs about those and we don't have time to work on basic options already so who cares about those
[20:51] <Amaranth> well dang
[20:51] <Amaranth> seb128: it's more of a feature request
[20:51] <mac_v> Amaranth: or did you mean indicator session?
[20:51] <Amaranth> mac_v: I use System->Log Out
[20:51] <Amaranth> mac_v: and System->Shutdown
[20:52] <mac_v> ah! those are not available in the default setup ;p
[20:52] <Amaranth> not unless you remove the FUSA applet
[20:52] <hyperair> i used to use gnome-do for everything, but then hibernate and suspend are not working due to devicekit changes
[20:52] <Amaranth> mac_v: If you aren't using those dialogs that effect is worthless
[20:52] <seb128> I tried gnome-do once and I don't understand the point of it
[20:52] <Amaranth> mac_v: I'm going to mark it Won't Fix now that I know what you want
[20:53] <seb128> I find it unefficient and ugly and not obvious to trigger
[20:53] <mac_v> Amaranth: huh?
[20:53] <Amaranth> seb128: it's so you take your hands off the keyboard less often
[20:53] <mac_v> Amaranth: yeah , indicator session removes those options  ;)
[20:54] <Amaranth> mac_v: Every single dialog FUSA pops up you want to use with this is just a confirm/deny dialog
[20:54] <hyperair> seb128: that's just you. many others love gnome-do
[20:54] <hyperair> seb128: use your keyboard a little more
[20:54] <Amaranth> I only thought the logout effect would be cool with the window where you choose what you want to do
[20:54] <seb128> Amaranth, alt-f2 - command doesn't make me move my hands
[20:54] <hyperair> mm yeah, and you're going to alt+f2 suspend?
[20:54] <seb128> hyperair, I probably didn't spent enough time to get the grasp of it
[20:54] <mac_v> Amaranth: for every single fusa prompts
[20:54] <Amaranth> seb128: no but you move them awkwardly and that only helps for running apps
[20:54] <seb128> it just seems alt-f2 with ugly icons to me
[20:54] <Amaranth> seb128: alt-f2 won't let you skip to the next song
[20:55] <Amaranth> seb128: or post to twitter :)
[20:55] <Amaranth> mac_v: right, that's kind of silly
[20:55] <walters> yeah, there's two parts; application launching/switching, and application-specific plugins
[20:55] <Amaranth> mac_v: those dialogs are just yes/no
[20:55] <mac_v> Amaranth: ;p
[20:55] <seb128> Amaranth, that's the sort of things I didn't figure and don't do anyway
[20:55] <hyperair> i use it to load my files
[20:55] <seb128> and for skipping songs most keyboards have keys
[20:55] <mac_v> Amaranth: so why shouldnt they have the effect?
[20:56] <Amaranth> mac_v: It just looks...wrong
[20:56] <seb128> there were some people suggesting installing gnome-do by default
[20:56] <mac_v> seb128: oh pls dont
[20:56] <seb128> but I think the whole thing is just not obvious to any non-tech user
[20:56] <Amaranth> mac_v: that effect is currently being used right now for gksu to tell you you can't do anything else until you decide what to do here (put in password or cancel)
[20:56] <seb128> and even for tech users it's not really...
[20:57] <Amaranth> seb128: it took me a year to figure out the point of quicksilver on OS X, now I can't do without
[20:57] <Amaranth> hmm, that's not exactly an endorsement
[20:57] <Amaranth> (gnome-do is very similar to quicksilver)
[20:57] <mac_v> gnome-do is a quicksilver rip-off ;p
[20:58]  * chrisccoulson starts installing gnome-do
[20:58] <Amaranth> it's really meant for power users who are also keyboard junkies
[20:58] <seb128> I'm using the keyboard a lot
[20:58]  * mac_v tries to jedi unplug chrisccoulson's connection ;p
[20:58] <seb128> but I use alt-tab, command lines and alt-f2 usually
[20:58] <Amaranth> people who would use ratpoison if it was just a little more practical :P
[20:59] <seb128> and gnome-do didn't bring any value for what I do
[20:59] <seb128> it's just slow and ugly
[20:59] <mac_v> i hate it because i cant customize the dock ;p
[20:59] <hyperair> mac_v: then don't use the dock.
[20:59] <mac_v> cairo-dock is awesome
[20:59] <Amaranth> seb128: the people suggesting it be installed by default all seemed to actually want docky
[20:59] <hyperair> mac_v: anyway docky's pretty cool all the same, though it's going to be splitting off from Do
[21:00] <Amaranth> seb128: the dock part of gnome-do
[21:00] <hyperair> cairo dock is awesome, but rather crashy
[21:00] <hyperair> it lacks polish
[21:00] <seb128> dunno what docky is
[21:00] <Amaranth> hyperair: is it? DBO was talking about that some time ago but nothing ever seemed to happen
[21:00] <seb128> I install gnome-do
[21:00] <mac_v> the latest 2.0 is great
[21:00] <mac_v> hyperair: ^
[21:00] <seb128> and sort of never figured how to run it
[21:00] <Amaranth> seb128: yet another OS X style dock
[21:00] <hyperair> Amaranth: the code's split off already.
[21:00] <Amaranth> seb128: it's a mode in gnome-do
[21:00] <hyperair> mac_v: maybe i should try it sometime.
[21:00] <seb128> I managed to make it displays ugly scares with pixelized icons to launch things
[21:01] <seb128> and that's about it
[21:01] <Amaranth> it's a theme actually, which is so absurd it's scary
[21:01] <hyperair> mac_v: the reason i ditched cairo dock was mainly because docky was pretty much do. it allowed me to cut down some resources by using docky instead of cairo dock
[21:01]  * mac_v waits for a gnome-do plugin for cairo-dock ;)
[21:01] <Amaranth> well, they call it in "interface" but it's basically this dock with gnome-do latched onto the side
[21:02] <Amaranth> seb128: so our users want an OS X clone
[21:02] <mac_v> grr...^
[21:03] <seb128> not only them, the design team too apparently ;-)
[21:03] <mac_v> seb128: booo David ;p
[21:03] <Amaranth> hahahaha
[21:03] <seb128> half joking there
[21:03] <hyperair> seb128: if your icons are pixelly you blame your icon theme, not do.
[21:03] <Amaranth> yeah, kind of biased there
[21:03] <seb128> every time they try to sell a thing their justification is that macos do it too
[21:04] <mac_v> eew
[21:04] <Amaranth> seb128: well OS X is a good model to base your work on if you're worried about usability
[21:04] <seb128> hyperair, I do blame it but same result, gnome-do looks ugly
[21:04] <Amaranth> except in the rather huge cases where it isn't
[21:04] <hyperair> seb128: i've never had a problem with pixelly icons.
[21:04] <Amaranth> like, well, the dock!
[21:04] <hyperair> seb128: like i said, your icon theme is screwed. go toss it and get a new oen
[21:05] <seb128> hyperair, lucky you, the search one displayed by default is ugly there and I use the default theme
[21:05] <Amaranth> gnome-do developers are recommend you use gnome-colors so as to not make gnome-do look bad
[21:05] <hyperair> seb128: then file a bug and get the icon fixed.
[21:05] <Amaranth> s/are//
[21:05] <mac_v> seb128: sometimes they just say ,its done by OSX rather than why doing it so is good! , thats want is crazy o.0
[21:05] <mac_v> what*
[21:05] <seb128> hyperair, oh, there are bugs, the issue is to find icons drawers
[21:06] <seb128> seems less obvious that you think
[21:06] <hyperair> seb128: icons drawers?
[21:06] <Amaranth> gnome-colors has a human-style version ;)
[21:06] <seb128> hyperair, whoever is drawing icons
[21:06] <seb128> designers
[21:06] <mac_v> hehe ;p
[21:06] <hyperair> ah
[21:06] <seb128> dunno how you call those guys
[21:06] <hyperair> the word is artist
[21:06] <Amaranth> artists
[21:06] <mac_v> artists
[21:07] <seb128> but we don't have lot of them around
[21:07] <hyperair> *not* drawer
[21:07]  * hyperair facepalms
[21:07] <Amaranth> hehe
[21:07] <seb128> well, artists do not only do icons
[21:07] <seb128> I was speaking about people interested in doing icons
[21:07] <hyperair> then you call them icon artists
[21:07] <seb128> not backgrounds
[21:07] <seb128> ok ;-)
[21:07] <mac_v> hyperair: seb128 is french so anything he says will sound good ;p
[21:07] <seb128> lol
[21:07] <hyperair> haha
[21:07] <Amaranth> only if he says it in french and in person
[21:08] <seb128> you don't have a word made from drawing?
[21:08] <hyperair> also, compiz's code leaves me wondering wtf it's doing
[21:08] <mac_v> nope
[21:08] <seb128> ie people who draw?
[21:08] <mac_v> artists
[21:08] <seb128> english is weird sometime ;-)
[21:08] <Amaranth> seb128: we only stole part of french
[21:08] <seb128> artists are people doing art in french
[21:08] <seb128> could be music as well as drawing
[21:09] <Amaranth> hyperair: the code to comment ratio is something like 10000:1
[21:09] <Amaranth> hyperair: what part are you looking at or looking for?
[21:09] <hyperair> Amaranth: updatePlugins. there's a bloody assertion that's failing
[21:09] <hyperair> Amaranth: it's left me without compiz. and no compiz leaves me very annoyed
[21:10] <hyperair> Amaranth: bug #429858
[21:10] <hyperair> well i could just take the shortcut and comment out the assertion
[21:10] <hyperair> but what the hell is j supposed to be?!
[21:11] <hyperair> my goodness they've reused their iterators so many times there's no telling what each iterator is for any more =.=
[21:11] <Amaranth> oh, that code is rather tame actually
[21:11] <hyperair> then please fix it. it's not tame enough for me to handle
[21:12] <Amaranth> j is the number of plugins we added, it should be the size of the pList array
[21:12] <seb128> Amaranth, btw did you figure why your local compiz rebuilds don't have the transparency issue?
[21:12] <Amaranth> seb128: no, I'm guessing buildd weirdness and hoping a rebuild will help
[21:12] <hyperair> Amaranth: so what does j != pListCount imply?
[21:12] <seb128> seems not likely
[21:13] <Amaranth> hyperair: pListCount is the number of plugins we should have determined by the number in the original plugin list plus the number in the new plugin list minus the one in the new plugin list that were already in the old plugin list
[21:14] <Amaranth> so somehow we didn't add enough plugins
[21:14]  * hyperair groans
[21:14] <Amaranth> what triggers his?
[21:14] <Amaranth> this*
[21:14] <hyperair> starting compiz!
[21:14] <hyperair> i can't start compiz damnit!
[21:14] <mac_v> lol
[21:14]  * hyperair headdesks repeatedly
[21:15] <hyperair> if this isn't the worst regression i've ever seen, i don't know what is
[21:15] <Amaranth> hyperair: a bit dramatic
[21:15] <mac_v> hyperair: add tag regression to the bug ;)
[21:16] <hyperair> well of course. but you do realize that compiz is a core part of the majority of ubuntu users' desktop usage.
[21:16] <Amaranth> hyperair: gconftool-2 -g /apps/compiz/general/allscreens/options/active_plugins
[21:16] <Amaranth> hyperair: and you are the first person I've heard of having this bug :P
[21:16] <hyperair> [core,ccp,dbus,place,mousepoll,gnomecompat,move,resize,decoration,png,svg,imgjpeg,text,neg,video,wall,snap,animation,scale,scaleaddon,expo,staticswitcher,regex,resizeinfo,workarounds,ezoom,vpswitch,extrawm,fade,session
[21:17] <hyperair> Amaranth: i didn't report it. there are three duplicates.
[21:17] <Amaranth> hrm, the sort is different
[21:17] <hyperair> i remember recursive-unsetting it
[21:17] <Amaranth> hyperair: bug number?
[21:17] <hyperair> https://bugs.launchpad.net/ubuntu/+source/compiz/+bug/429858
[21:17] <Amaranth> ah, you said that earlier
[21:18]  * Amaranth gets more caffeine
[21:18] <hyperair> indeed
[21:18]  * hyperair whacks gnome-shell for not having anywhere near compiz functionality
[21:18] <hyperair> i don't know what i'm going to do when gnome-panel finally gets deprecated in favour of shell
[21:19] <Amaranth> hyperair: feel like installing every dev package for the entire GNOME and KDE desktops and trying to build compiz with some debug printfs?
[21:19] <hyperair> Amaranth: hell no.
[21:19] <hyperair> Amaranth: i've got pbuilder though
[21:19] <Amaranth> ok, amd64 then?
[21:19] <hyperair> mmhmm
[21:19] <hyperair> amd64.
[21:19] <Amaranth> hyperair: it takes like 20 minutes to download all the dependencies in pbuilder if you've got a decent connection
[21:20] <Amaranth> not something we want to do over and over
[21:20] <hyperair> there's /var/cache/pbuilder/aptcache
[21:20] <Amaranth> just installing them takes 5 minutes
[21:21] <hyperair> heh maybe for you
[21:21] <hyperair> i can fall asleep waiting for my pbuilder to finish installing dependencies sometimes
[21:26]  * Amaranth tries to find wiki page for people.ubuntu.com usage
[21:27] <seb128> hyperair, don't worry gnome-panel will not be deprecated any time soon
[21:28] <hyperair> seb128: hmm that's nice to hear, but it will eventually, when shell takes over, right?
[21:28] <seb128> well, gnome-shell requires 3d to be working
[21:28] <seb128> so GNOME needs a fallback solution for all users who don't have that
[21:28] <seb128> so gnome-panel will still be around for a while
[21:30] <Amaranth> seb128: I heard it'll be around until GNOME 3.2
[21:30] <Amaranth> The GNOME experience is going to be so different between people with tfp and people without
[21:30] <Amaranth> tfp = GLX_EXT_texture_from_pixmap
[21:31] <Amaranth> hmm, sure does take a while to upload to people.ubuntu.com
[21:31] <seb128> the thing let to know is which ones will get the best experience ;-)
[21:32] <Amaranth> seb128: people without, from what I've seen
[21:32] <seb128> ;-)
[21:33] <hyperair> Amaranth: agree.
[21:34] <Amaranth> hyperair: http://www.realistanew.com/random/compiz-core_0.8.3+git20090915-0ubuntu1_amd64.deb
[21:34] <hyperair> is that all that needs to be reinstalled?
[21:34] <Amaranth> install that and run `compiz &`
[21:34] <Amaranth> pastebin the output
[21:34] <hyperair> alright
[21:35] <Amaranth> I gave up on people.ubuntu.com, it is still supposedly doing something
[21:35] <Amaranth> took about 2 seconds to upload to my website instead
[21:37] <hyperair> dupPluginCount=2, pListCount=3, initial plugins=2, final plugin count=2
[21:37] <hyperair> Amaranth: ^
[21:38] <Amaranth> hyperair: have you modified compiz-wrapper at all?
[21:38] <hyperair> lemme stare at it
[21:38] <Amaranth> it should be shoving 6 plugins in there from the start
[21:38] <Amaranth> it's only doing 2
[21:38] <hyperair> hmm
[21:39] <hyperair> ah i have a dpkg-divert i forgot to remove
[21:39] <Amaranth> hyperair: CM_DRY=yes compiz
[21:39] <Amaranth> don't remove the divert yet
[21:40] <Amaranth> from the dry run paste the Execute line
[21:41] <hyperair> ..shit i removed it already
[21:41] <hyperair> /usr/bin/compiz.real --ignore-desktop-hints --replace  core ccp
[21:41] <hyperair> that's what it says
[21:41] <Amaranth> ah, core
[21:41] <hyperair> ?
[21:42] <Amaranth> yeah, you can't have core in there anymore
[21:42] <hyperair> i see
[21:42] <Amaranth> we did in older compiz-wrapper packages but the changes in git to make plugins added on the command line stay loaded no matter what made things blow up if you added core on the command line
[21:42] <Amaranth> although that was supposed to be fixed too, guess it failed
[21:43] <Amaranth> sure enough, if I run that line I get the same error
[21:43] <hyperair> i see
[21:44] <Amaranth> apparently you and 4 other people are doing things to compiz-wrapper they shouldn't be :P
[21:44] <hyperair> so why weren't you seeing it in the first place?
[21:44] <Amaranth> because once I got that crash I fixed compiz-wrapper to not add core before I pushed the changes to bzr
[21:44] <Amaranth> i just forgot about it
[21:44] <Amaranth> actually I think before they tried to "fix" it you got an infinite loop of loading/unloading core
[21:45] <hyperair> hmm
[21:45] <hyperair> well what i was doing to compiz-wrapper was unblacklisting my gpu
[21:46] <hyperair> because someone had the great idea to blacklist my pci id
[21:46] <Amaranth> http://git.compiz.org/compiz/core/commit/?h=compiz-0.8&id=20cdf69d096616104cd26fc30c8efd7e4e4e164d
[21:46] <Amaranth> hyperair: what chip is that?
[21:46] <Amaranth> and there are better ways to work around the blacklist
[21:48] <hyperair> Amaranth: i965
[21:48] <Amaranth> hyperair: http://ubuntuforums.org/showthread.php?t=582112
[21:48] <Amaranth> I wrote that post when we blacklisted i965 :P
[21:48] <hyperair> wait, lemme fix my gconf then log back in
[21:49] <hyperair> ...this is fun. xsplash won't stop
[21:50] <hyperair> now i'm staring at an xsplash screen + panel
[21:50] <hyperair> ah it's finally back to normal
[21:50] <seb128> xsplash should have a  25 seconds timeout
[21:50] <seb128> weird that it doesn't go away for you
[21:50] <hyperair> i see
[21:51] <hyperair> er it did timeout in the end
[21:51] <hyperair> it's just that i managed to finish logging in before 25 seconds was up
[21:51] <seb128> ok, the issue about the timeout being reached when it should not should be fixed after the alpha freeze I think
[21:52] <hyperair> i see
[21:52] <hyperair> now then.. is there a way i can tell gdu to go away, and that my 6 bad sectors are here to stay and i'm lazy to do anything (more) with it?
[21:53] <seb128> the git versions allows that I think
[21:53] <seb128> chrisccoulson was going to look at backporting the change for karmic
[21:54] <hyperair> hmm that's nice
[21:54] <hyperair> it drove me nuts enough to attempt to get those 6 sectors reallocated
[21:54] <hyperair> i ended up forgetting to include count=1 in my dd invocation and lost 4G of data after a good fsck
[21:54]  * hyperair has now developed a phobia of dd
[21:55] <chrisccoulson> yeah, i'm going to look at that gdu change. but it currently blocks on a devicekit-disks update, which will need a FF exception
[21:55] <seb128> you can also uninstall gnome-disk-utility
[21:56] <hyperair> hmm true..
[21:56] <hyperair> or i could prevent it from starting up
[21:56] <hyperair> i'm interested to know about any *more* issue with my hard disk
[21:57] <hyperair> like i know about my 6 bad sectors, but if it deteriorates any further, i'd like to know
[21:57] <mac_v> hyperair: just turn off disk notifications from startup list
[21:57] <seb128> right, makes sense
[21:57] <chrisccoulson> here's a screenshot of the new smart status window:
[21:57] <chrisccoulson> http://people.freedesktop.org/~david/nsui-2.png
[21:57] <hyperair> mac_v: please read what i said again, more carefully this time ;)
[21:57] <mac_v> heh , i type slow ;p
[21:58] <hyperair> chrisccoulson: ooh nice
[21:58] <chrisccoulson> the notifier is used for more than failure notification too, so turning it off is not a solution;)
[21:58] <Amaranth> hyperair: If you have 6 bad sectors that means it is going to (probably somewhat quickly) deteriorate further
[21:58] <hyperair> Amaranth: i know :(
[21:58] <chrisccoulson> Amaranth - i have 24 bad sectors;)
[21:58] <chrisccoulson> it's been like that for ages
[21:58] <hyperair> chrisccoulson: fun. =p how old is it?
[21:59] <Amaranth> chrisccoulson: just under the threshold for "disk has failed"
[21:59] <mac_v> i'v had 3 for nearly 1 yr!
[21:59] <chrisccoulson> it's about 4 years old now
[21:59] <hyperair> Amaranth: but 24 out of millions of sectors is nothing, right?
[21:59] <Amaranth> I suppose you could get it like that from the factory due to a manufacturing defect
[21:59] <hyperair> i wonder how many bad sectors my desktop has..
[22:00] <hyperair> either way my "bad" sectors aren't bad, but more of half-bad half-ok
[22:00] <hyperair> sometimes they read
[22:00] <hyperair> sometimes they don't
[22:00] <Amaranth> Although if I got a disk with any bad sectors right after buying it I'd be taking it back
[22:00] <hyperair> same goes for writing
[22:01] <chrisccoulson> it might be a good thing if my disk fails
[22:01] <hyperair> my disk failing is a bad thing. /me has backups but they're all over the place
[22:01] <chrisccoulson> i can then justify buying a new computer to my girlfriend
[22:02] <seb128> lol
[22:02] <hyperair> heheh
[22:02] <chrisccoulson> "i have to buy a new computer because it has a terminal fault and you won't be able to view our photos again"
[22:02] <chrisccoulson> that should do it ;)
[22:02] <mac_v> lmao
[22:02] <chrisccoulson> then i'd just hide this one in another room and use one of the spare disks
[22:03]  * hyperair prints out the irc logs and files them up for future blackmailing purposes
[22:03] <Amaranth> chrisccoulson: Too bad you didn't get caught in the recent "karmic is completely broken" stuff
[22:03] <Amaranth> You could have used that as an excuse, blamed it on the hardware :P
[22:03] <seb128> oh he did
[22:03] <chrisccoulson> Amaranth - i got caught up in that;)_
[22:04] <seb128> he just decided to spend his night to fix it rather than use it as en excuse apparently ;-)
[22:04] <chrisccoulson> i spent 5 hours last night trying to get my machine to boot ;)
[22:04] <chrisccoulson> heh
[22:04] <Amaranth> No one is ever sneaky when their computer is broken
[22:04] <hyperair> what was the whole karmic is completely broken issue about?
[22:04] <hyperair> was it the repeated fscks?
[22:04] <hyperair> due to a bad timestamp or something
[22:04] <Amaranth> hyperair: something to do with upstart, I guess
[22:04] <hyperair> huh O_o
[22:04] <seb128> some people have dbus not starting apparently
[22:04] <hyperair> i feel lucky i didn't get to see that one
[22:05] <Amaranth> at least with no dbus you can still plug in an ethernet cable and upgrade to fixed packages
[22:05] <hyperair> i still have to find some way to hijack my friend's notebook for debugging. 9.10 won't boot for him
[22:05] <Amaranth> the libc fun a release or two back was real painful
[22:05] <chrisccoulson> i had no issues with dbus or anything like that, but couldn't boot because i use bindfs for sharing files
[22:05] <hyperair> Amaranth: oh yeah i got caught up in that. i think it was hardy.
[22:05] <hyperair> or before that
[22:05] <Amaranth> hardy sounds right
[22:06] <hyperair> yeah hardy.
[22:06] <seb128> chrisccoulson, did you let Keybuk about that issue?
[22:06] <seb128> chrisccoulson, or opened a bug about it?
[22:06] <Amaranth> About a month before I had decided to start checking ubuntu forums if any package that would make the system unbootable got updated
[22:06] <seb128> I'm not addict enough to dist-upgrade to fall into all those issues usually
[22:06] <Amaranth> that's what saved me
[22:07]  * hyperair groans
[22:07] <hyperair> now stackswitch isn't working..
[22:07] <seb128> by the time I join IRC and want to upgrade there is already enough noise about breakages to let me stay away from those
[22:07] <Amaranth> seb128: I run a dist-upgrade when I wake up, a dist-upgrade when I go to lunch, and a dist-upgrade before I get off the computer
[22:07] <chrisccoulson> seb128 - yeah - bug 430880
[22:07] <seb128> chrisccoulson, ok, cool
[22:07]  * hyperair wonders if stackswitch will ever enter ubuntu's archives
[22:07]  * Keybuk points at the status ;)
[22:07] <chrisccoulson> right, i'm just going to do that now:)
[22:08] <Amaranth> Keybuk must have upstart in his highlights :P
[22:08] <Amaranth> hyperair: stackswitch has never made it into compiz
[22:08] <Amaranth> hyperair: we don't diverge that far from upstream
[22:08] <hyperair> oh it hasn't?
[22:08] <hyperair> can't it be packaged separately though?
[22:09] <mac_v> Keybuk: are those issues fixed? so its safe to restart? ;)
[22:09] <Amaranth> no because it isn't really maintained
[22:09] <hyperair> ah damn
[22:09] <Amaranth> onestone wrote it when someone told me about the idea in prague and it has sit since
[22:09] <Amaranth> s/sit/sat/
[22:10] <Amaranth> actually I think it was Ng but that doesn't seem right
[22:10] <Keybuk> mac_v: some issues are fixed
[22:10] <Keybuk> some issues are not fixed
[22:11] <Amaranth> mac_v: if you don't do any funky networking stuff you should be fine
[22:11] <mac_v> aw... i'm using the boot ppa ,
[22:11] <Ng> what did I do? ;)
[22:11] <Amaranth> Ng: proposed the idea for stackswitch in compiz
[22:11] <Amaranth> I remember someone sitting with my downstairs and telling me about it but the idea was so exciting I forgot who
[22:11] <Amaranth> s/my/me/
[22:11] <Amaranth> too many typos today
[22:11] <Ng> Amaranth: I haven't suggested random compiz things to you since the UDS in Boston, and I wasn't in Prague ;)
[22:12] <Amaranth> ok then, wrong person
[22:12] <Ng> innocent \o/
[22:12] <chrisccoulson> Keybuk - i'm just trying to get the info for bug 430880 - should mountall actually return eventually?
[22:12] <Amaranth> oh, I know, it must have been mpt
[22:12] <Keybuk> no, it doesn't
[22:12] <Keybuk> well
[22:12] <Keybuk> it should
[22:12] <Keybuk> but clearly it doesn't return in your case ;)
[22:12] <Keybuk> so it's ok to ^C it
[22:13] <Keybuk> the log up to that point should be reasonably revealing
[22:13] <chrisccoulson> cool, thanks
[22:13] <chrisccoulson> Keybuk - http://launchpadlibrarian.net/31922472/mountall.log
[22:15] <seb128> chrisccoulson, you probably want to change the bug back the new too
[22:15] <seb128> to new
[22:15] <chrisccoulson> yeah, i should do really
[22:15] <chrisccoulson> done!
[22:15] <Amaranth> hyperair: did dropping the revert fix the problem?
[22:15]  * Amaranth cries at his inbox
[22:15] <Keybuk> chrisccoulson: that's weird, your log is missing the important bit
[22:16] <hyperair> Amaranth: yeah, thanks
[22:16] <chrisccoulson> hmmm, what's the important bit?
[22:16] <Keybuk> chrisccoulson: did you ^C it?
[22:16] <chrisccoulson> i did, there was no extra info when i did that
[22:16] <Keybuk> chrisccoulson: the dump of what mountall thinks each filesystem is
[22:18] <chrisccoulson> Keybuk - isn't the last statement there incomplete too? ("dropping unknown")
[22:18] <Keybuk> yeah
[22:18]  * Keybuk hates UNIX
[22:18] <chrisccoulson> Keybuk - i'll attach gdb too it. i wonder if nih_debug chokes on any wierd characters in that entry?
[22:19] <Keybuk> nah
[22:19] <Keybuk> it's just silly stdout buffering because you're piping it
[22:19] <Amaranth> I hate when that happens
[22:19] <mac_v> Amaranth: i didnt understand the reason for the FUSA not having the logout effect? , its a shutdown or a logout process why should it be treated the same way as the upstream dialogue?
[22:20] <hyperair> is anyone familiar with gnome-power-manager not locking the screen when suspending/hibernating?
[22:20] <mac_v> when a user is getting out of the system , the screen darkens
[22:20] <Amaranth> mac_v: the upstream dialog has choices you need to make before it'll do anything
[22:20] <Amaranth> well, it'll shutdown if you don't make any choice but that isn't always what you want so you want to make sure they are paying attention to pick the right task
[22:21] <Amaranth> mac_v: upstream dialog is like windows, fusa dialog is like OS X ;)
[22:21] <mac_v> Amaranth: yes  ,i understand that , but lack of choices doesnt make the dialogue less important , the present fusa confirmation dialogue is miniscule and doesn grab enough attention
[22:22] <Amaranth> mac_v: it doesn't have to
[22:22] <mac_v> it gets lost in the back ground
[22:22] <Amaranth> if you chose to shutdown and they ignore the window that comes up the system will still shutdown
[22:22] <Amaranth> pronoun problems...
[22:23] <Amaranth> mac_v: meanwhile with the upstream dialog if you chose to shutdown but you actually want to suspend if you don't see the dialog it'll shutdown instead
[22:23] <mac_v> Amaranth: that is the same behavior in the upstream dialogue too
[22:23] <mac_v> hmm...
[22:23] <Amaranth> but since we don't use the upstream dialog and you can't make a match rule for the upstream dialog there is nothing that can be done for it
[22:24] <chrisccoulson> Keybuk - http://launchpadlibrarian.net/31922802/mountall-gdb.log
[22:24] <chrisccoulson> i interrupted it twice in GDB and got the same trace
[22:24] <mac_v> Amaranth: why do we have to worry about the upstream dialogue?
[22:24] <Amaranth> any match rule for the upstream dialog also includes gnome-session running in general
[22:24] <mac_v> its something like gnome-strati... or what ever the name is ;)
[22:25] <Amaranth> mac_v: we don't, that's why I closed it instead of adding a gnome-session task to make the upstream dialog distinct
[22:26] <Amaranth> mac_v: I'm going to pull a Design Team and say "that's how OS X does it"
[22:26] <Keybuk> chrisccoulson: trace is uninteresting - it's the debug log I want to see
[22:26] <Keybuk> chrisccoulson: in gdb "run --debug" then capture that
[22:26] <Keybuk> that *would* be useful
[22:26] <mac_v> Amaranth: ;p , instead of that we could just use it for FUSA alone, ;) the present dialogue is miniscule ... let me ge tback to  you on that ;)
[22:27] <mac_v> OSX is not always right
[22:27] <Amaranth> no but in this case there is no reason to give more attention to the dialog
[22:27] <Amaranth> it is going to do what you want even if you ignore it
[22:27] <chrisccoulson> Keybuk - i get the same debug output as before though
[22:27] <mac_v> Amaranth: indeed there is , since its a countdown , if the user misses it he would be booted out :(
[22:28] <Amaranth> mac_v: he did ask the system to shut down
[22:28] <mac_v> hrm...
[22:29] <Keybuk> chrisccoulson: then try "p fflush (stdout)"
[22:30] <mac_v> Amaranth: i disagree , but lets get an option from the design team ;)
[22:31] <mac_v> Amaranth: why is find that its necessary is , since its a process thats closing the session , so the greying out signifies that .... But maybe its just me ;p
[22:31] <chrisccoulson> Keybuk - http://paste.ubuntu.com/272367/
[22:32] <Keybuk> that's better
[22:33] <Keybuk> chrisccoulson: ok, it thinks they're virtual filesystems
[22:33] <Keybuk> that's ok
[22:33]  * Keybuk has a fix pending for this
[22:33] <chrisccoulson> Keybuk - thanks:)
[22:35] <mac_v> Amaranth: BTW , which plugin does gksu use?
[22:37] <Amaranth> mac_v: none, I added code to it to do a composited fade
[22:37] <Amaranth> way before that compiz plugin even existed
[22:37] <mac_v> the grey background ok the password prompt , started using "Close" animations after the recent compiz update
[22:38] <mac_v> s/ok/of
[22:38] <Amaranth> yeah, I see that
[22:38] <mac_v> \o/
[22:38] <Amaranth> it avoids that on open by making a window 2x2 or so then resizing it to the size of the screen before starting the fade
[22:39] <Amaranth> any settings change we make in compiz for this will not take effect if users have changed their animation settings on their own
[22:40] <mac_v> huh? you mean if we fix the gksu to not having the close animation ... it will break for customized settings?
[22:41] <mac_v> rather > it will not be fixed for customized settings
[22:41] <Amaranth> right because the fix would have to be in compiz settings
[22:41] <Amaranth> hmm, I suppose we could screw with the window type for gksu's fade window
[22:41] <mac_v> aw!
[22:41] <Amaranth> but that may cause other problems
[22:45] <mac_v> what is the window match of that grey background? couldnt we add it to the exclusions...? ;p i guess its not so simple
[22:46] <Amaranth> mac_v: you can't find it normally since gksu prevents all input while it is running
[22:46] <Amaranth> and again, if you make such a rule change in compiz only people running stock will see it
[22:46] <mac_v> aw :(
[22:49] <Amaranth> since it's a GDK window and not a GTK one it is harder to change the window type
[22:49] <Amaranth> gdk_window_set_opacity only works on TOPLEVEL windows
[22:51] <Amaranth> nevermind, figured it out
[22:55] <Amaranth> mac_v: fixed :)
[22:55] <mac_v> \o/
[22:55] <mac_v> Amaranth: how?
[22:56] <Amaranth> mac_v: gdk_window_set_type_hint (fadeout->window, GDK_WINDOW_TYPE_HINT_DIALOG)
[22:56] <Amaranth> dialogs get the fade animation
[22:57] <mac_v> nice :)
[23:12] <Amaranth> wow, OS X fail
[23:12] <Amaranth> that's rate
[23:12] <Amaranth> err, rare
[23:20] <chrisccoulson> wow, users are really having a rant about tap-to-click being disabled by default