[04:45] <hyperair> Sarvatt: okay maybe i was wrong. for some reason, locking the screen for long periods of time raises the memory consumption grealy
[04:45] <hyperair> greatly*
[05:06] <Sarvatt> screen locking, cryptloop disks, i'm convinced you're just trying to break stuff now :D
[05:09] <Sarvatt> darn, need polkit-gobject-1 for a few packages i want to update but it looks like we dont ship that from policykit yet
[05:12] <hyperair> hey it's dm-crypt stuff not cryptloop you know?
[06:08] <Sarvatt> oo, didn't notice .31 had ext4 online defragmentation in it
[06:08] <Sarvatt> wrong channel, sorry
[06:56] <hyperair> eh it does?
[06:56] <hyperair> O_O
[06:56] <hyperair> how do you turn it no?
[06:56] <hyperair> on*
[10:29] <RAOF> Alright.  Let's look at updating Karmic's nouveau drivers.
[11:47] <hyperair> Sarvatt: just booted into the kernel you compiled. for some reason, eventhough KMS is on, there seems to be issues changing VT O_o
[11:48] <hyperair> when i switch to a tty, it has the image rendered on X, but with funny rainbow like colours 
[11:48] <Sarvatt> fun!
[11:49] <hyperair> heheh
[14:37]  * hyperair is noticing some strange artifacts in compiz animations
[19:02] <Duke`> can someone try this program and tell me if it segfault: http://pastebin.com/mf5f60dc ? compilation instruction are in file header (at line 2)
[19:05] <Sarvatt> yep segfaults here
[19:06] <Duke`> and with LIBGL_ALWAYS_SOFTWARE=1 ./vbo ?
[19:07] <Sarvatt> Mesa: User error: GL_INVALID_OPERATION in glMapBufferARB(already mapped)
[19:07] <Sarvatt> thats with MESA_DEBUG=1
[19:08] <Duke`> hu ok
[19:08] <Sarvatt> well it started but isnt doing anything with LIBGL_ALWAYS_SOFTWARE=1
[19:08] <Duke`> you can move the cube with mouse
[19:08] <Sarvatt> oh had to interact with it :D
[19:09] <Duke`> :)
[19:09] <Duke`> I don't have GL_INVALID_OPERATION in software mode
[19:22] <Duke`> the driver seems to be less smart than before, in regard to VBO mapping/drawing :(
[19:34] <Sarvatt> doubt it helps but http://sarvatt.com/downloads/glut_bt.txt
[19:36] <Duke`> in fact, the first time glMapBufferARB is called, it returns a valid pointer, but at the next cycle it returns NULL (because it seems that the GPU is still reading this memory) and it crashes...
[19:36] <Duke`> hum I suspect a bug
[20:10] <Sarvatt> Duke`: are you on a 965
[20:11] <Sarvatt> err sorry, 965+?
[20:15] <Duke`> 945
[22:33] <Sarvatt> hyperair: seen this? http://bugs.freedesktop.org/show_bug.cgi?id=22383
[22:39] <Sarvatt> hyperair: you should file a bug on fdo about screen locking with compiz enabled leading to a huge gem_objects memory leak that you were talking about