[03:04] <wgrant> tjaalton: Bug #267611 doesn't look good; they're using the latest versions of things...
[05:51] <tjaalton> wgrant: there are also bugs where lshal doesn't show the device but it can be manually configured in xorg.conf. maybe this is related
[05:52] <tjaalton> oh, the log shows that it's found
[05:52] <tjaalton> hum no, it's using evdev
[05:53] <tjaalton> that's because the fdi file doesn't know about SynPS/2 :)
[05:53] <tjaalton> so this bug is easy
[05:57] <wgrant> There's also a patch in upstream git that will probably fix some of the scrolling issues.
[06:00] <wgrant> tjaalton: What do you mean it doesn't know about SynPS/2? My device name is identical, and it works fine.
[06:02] <tjaalton> hum right, maybe lshal shows a different string
[06:02] <tjaalton> and maybe that user doesn't have hal running?
[06:02] <tjaalton> although that would cause more problems
[06:02] <wgrant> It finds the device through hal.
[06:02] <tjaalton> and since he got evdev it should be running..
[06:02] <tjaalton> yep
[06:02] <wgrant> (II) config/hal: Adding input device SynPS/2 Synaptics TouchPad
[06:03] <wgrant> It uses the right driver.
[06:03] <wgrant> But it doesn't work.
[06:03] <wgrant> This is why I am confused.
[06:03] <tjaalton> oh right, it does
[06:03] <tjaalton> bah
[06:04] <tjaalton> breakfast first, then off to work :) ->
[06:04] <wgrant> Hmm.
[06:04] <wgrant> xorg-server 2:1.4.99.906-2ubuntu3 (buildd@crested.buildd) 
[06:04] <wgrant> That's old.
[06:04] <wgrant> That would be the problem.
[06:05] <wgrant> I suppose.
[06:05] <tjaalton> it should support properties
[06:06] <wgrant> Hm, indeed.
[06:06] <wgrant> Anyway, eat.
[08:43] <tjaalton> wgrant: btw, let's keep all synaptics bugs in x-x-i-s, because I believe the source package will be renamed before too long
[08:44] <crevette> hello
[08:46] <bryce> heya
[08:48] <tjaalton> howdy
[08:51] <crevette> do you have any report on widget corruption on gtk with intel ?
[08:51] <crevette> I have that specifically on epiphany
[08:52] <crevette> http://www.monsterup.com/upload/1220860308.jpg
[08:52] <tjaalton> never seen that on my 965GM
[08:52] <tjaalton> don't use epiphany though
[09:12] <seb128> I do use it and I've no such issue
[10:32] <wgrant> tjaalton: Didn't we only rename it to match Debian a few months ago?
[11:20] <tjaalton> wgrant: yes, to enable easier merges, but since then the project has adopted the MIT license and is hosted on b.fd.o. I believe Mattia is ready to rename the source now :)
[11:22] <wgrant> Aha.
[11:22] <wgrant> tjaalton: I was pleased to see that it reawoke a couple of months back.
[11:22] <wgrant> It had been looking unpleasantly dead for too long.
[11:26] <tjaalton> yes, that's true.. at least it's maintained now
[11:32] <wgrant> It's unfortunate that float properties such as speed aren't yet able to be exposed through XI properties.
[11:35] <tjaalton> yeah
[11:35] <wgrant> Otherwise we could easily have a nice complete Synaptics GUI.
[11:35] <wgrant> That doesn't need SHMConfig...
[11:36] <tjaalton> the driver could use a different scale, like x1000 or so and use constants :)
[11:37] <wgrant> Integers, you mean?
[11:37] <tjaalton> uhm, yes :)
[11:37] <tjaalton> constant isn't that useful
[11:37] <wgrant> No, not really.
[11:37] <wgrant> But yes, that is one easy option that I'd thought of. But I presume upstream will have a better one.
[11:39] <wgrant> Do we know of any sane person who has hardware afflicted by the ultra-slowness bug? That's getting quite a few people annoyed on the forums.
[11:39] <wgrant> And Launchpad.
[11:39] <tjaalton> the bug has been updated
[11:40] <tjaalton> and upstream is aware of it
[11:40] <wgrant> Ah, good.
[11:40] <tjaalton> it can be quirked on the fdi file
[11:40] <tjaalton> but the real problem is that some code is missing from the new version that used to be there
[11:40] <wgrant> Useful, those FDI files.
[11:40] <wgrant> Ah.
[11:40] <tjaalton> so it's not a proper fix
[12:09] <tjaalton> wgrant: float support is coming, just not implemented yet
[12:14] <tjaalton> wgrant: something you'd like to try?-)
[12:19] <wgrant> tjaalton: I'd love to. But I've not looked into how the whole thing works in depth.
[12:20] <wgrant> I've remained blissfully ignorant of how things like X operate internally.
[12:21] <tjaalton> me too. I also lack the skill to fix anything
[12:21] <tjaalton> which would involve coding, and in C
[12:23] <wgrant> My C isn't bad, but X is formidable.
[12:23] <tjaalton> heh
[12:29] <jcristau> some parts are not too bad
[12:34] <wgrant> Encouraging.
[12:59] <tjaalton> bryce: looks like you deleted changelog.Debian.old from the xorg git on Aug 5th ;)
[13:00] <tjaalton> at least that merge deleted it
[14:45] <wgrant> tjaalton: Well, X doesn't seem to like 64-bit property values, but I've got it working rather well converting -synaptics' doubles to/from floats.
[14:53] <tjaalton> wgrant: whoa, excellent :)
[14:55] <wgrant> A few changes in xinput and -synaptics, and defining XA_FLOAT in Xatom.h (although there must be something like that already...).
[14:58] <tjaalton> you could send those to the list for a review
[14:59] <wgrant> Once they're less pathetic, perhaps.
[15:00] <tjaalton> heh, you could also ask peter directly (whot on irc)
[15:24] <jcristau> wgrant: sending floats over the wire?
[15:24] <jcristau> wgrant: might want to do fixed point instead
[15:26] <wgrant> jcristau: What is the benefit of that?
[15:27] <jcristau> probably less architecture dependent
[15:27] <wgrant> True.
[15:31] <wgrant> Grumble. But then one of course has the issue of differing ranges. And who knows how much precision is required...
[15:31] <jcristau> right. integer properties are easier
[15:32] <wgrant> Definitely.
[15:34] <wgrant> Unfortunately, some of the more useful -synaptics properties aren't integers.
[15:41]  * wgrant gives up and goes to bed.
[15:41] <wgrant> Night all.
[18:59] <Ampelbein> hi! what is the right package for bugs in xkb? like bug #224649 ?
[19:03] <jcristau> Ampelbein: "yes, there's a reason"
[19:10] <jcristau> might be fixed at some point, but, since it's xkb...
[19:31]  * jcristau wtfs at bug 267867
[19:37] <bryce> ideas on bug 267682?
[19:39] <bryce> ah hell, brb
[19:42] <tjaalton> jcristau: heh
[21:08] <bryce> tjaalton: you about?
[21:40] <tjaalton> bryce: yep
[22:08] <seb128> hi there
[22:09] <seb128> is current intrepid known to be broken on intel?
[22:09] <tjaalton> in what way? my laptop works
[22:10] <tjaalton> sure there are bugs too
[22:11] <seb128> xglinfo segfault on startup but I just noticed that nvidia-glx-173 got installed and diverts libGL, let me remove it and restart my session
[22:11] <seb128> that's an intel laptop why the heck does the binary nvidia get installed?
[22:12] <tjaalton> beats me :)
[22:14] <seb128> nvidia-173-kernel-source
[22:14] <seb128> Recommends: nvidia-glx-173 (>= 173.14.12)
[22:15] <seb128> doesn't explain why
[22:16] <crevette> seb128: doesn't crash here
[22:17] <seb128> crevette: do you have nvidia-glx-173 installed?
[22:17] <crevette> no
[22:24] <crevette> I have this message when running glxinfo -> Failed to initialize TTM buffer manager.  Falling back to classic.
[22:25] <tjaalton> ignore it
[22:29] <seb128> alright fixed now
[22:29] <seb128> I still would like to now why those nvidia packages got installed
[22:30] <seb128> that will teach me to use mvo's toys, I updated using update-manager ;-)
[22:31] <tjaalton> well, can't answer that one.. :)
[22:32] <crevette> tjaalton: framerate of glx* tools is synced on vertical freq of the screen?
[22:33] <tjaalton> crevette: something like that
[22:33] <crevette> tjaalton: sorry, I don't know much
[22:34] <tjaalton> crevette: no worries, me neither
[22:40] <jcristau> yes, sync to vblank is on by default on intel
[22:40] <jcristau> so gears fps addicts will be sad
[22:57]  * wgrant wtfs at the response to tjaalton's closing of bug #267867
[23:06] <bryce> hi tormod
[23:14] <tormod> hi bryce
[23:14] <tormod> you and brian have done some rocking -ati bug work lately :)
[23:20] <bryce> tormod: thanks, yeah I've been really focused on it
[23:21] <bryce> tormod: there were a number of bugs you had asked for updates 1 month ago, that have received no response, that could probably be closed as expired at this point
[23:22] <bryce> on the 14th we could do another review - that's the 1 month anniversary of bdmurray's request for updates, so I guess we can expire many of the bugs that still have no received word
[23:23] <bryce> tormod: also I don't know if you saw, but I did up a patch to allow us to quirk fixes for the AGPMode issue.  I think I caught all the cases where people were reporting applicable symptoms and requested they test, but please keep your eyes open for more such cases
[23:23] <bryce> I suspect the AGPMode bug is going to be fairly pervasive