[caret-users] Nvidia-glx and drawing borders/cuts (description+workaround)

[Roland Marcus Rutschmann]

Hi,

I came up with something funny when using Nvidia-glx and caret. I can do 
mostly everything including clicking on surfaces and getting the correct 
nodes but I cannot draw borders or cuts (didn't try paints).

Workaround: Use libGL.so.1 by mesa eg the included libGL.so of caret by using 
LD_LIBRARY_PATH eg:

export CARET_HOME=/opt/caret-5.51
export PATH=$CARET_HOME/bin:$PATH
export LD_LIBRARY_PATH=$CARET_HOME/lib:$LD_LIBRARY_PATH
caret5 $*

The problem: The borders are drawn but somewhere else. Caret debug gives the 
following:
Window: (223, 215) Model: (1.08007e+25, 1.08172e+21, 7.29698e+15)
Button State: 0
Button 1
Window: (223, 217) Model: (1.08007e+25, 1.01991e+21, 7.29698e+15)

These numbers are "a bit" higher then when using the mesa libraries:
Window: (273, 161) Model: (-7.15888, 57.9219, 6508.08)
Button State: 0
Button 1
Window: (275, 161) Model: (-5.85727, 57.9219, 6508.08)
Button State: 0
Button 1

The problem occurs at 2 computers at university both with debian etch and by 
now xorg-7.2. One of the computers had xorg-7.1 until recently and the same 
problems. I tested nvidia-glx versions 1.0.8776-4, 1.0.9746-2, and 
100.14.19-1.

BUT it works on my home computer which runs etch with xorg-7.1 and nvidia-glx 
1.0.9746-2.

Other parameters I tested:
In remote use the problem depends on the computer caret runs on, not the 
computer where its window appears.
It works with all other graphics drivers available to me.
The problem is independent of the use of one or 2 monitors.
I tried all caret versions since 5.3->same results.

The only difference between my home and university computer I can think of is  
the kernel.

I don't think this is a bug in caret but more a bug in my system. I just 
thought I report it just in case someone can replicate this and/or needs a 
work around.

I searched the archives and found some drawing border probs but I think those 
were different. If I overlooked something -> sorry.

Regards,

Roland

-- 
Dr. Roland Marcus Rutschmann <Roland.Rutschmann at psychologie.uni-regensburg.de>
Institute for Experimental Psychology, University of Regensburg
Universitätsstraße 31, 93053 Regensburg, Germany
Tel: +49 941 943 2533, Fax: +49 941 943 3233
http://www.psychologie.uni-regensburg.de/Rutschmann
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://brainvis.wustl.edu/pipermail/caret-users/attachments/20071204/62baac7b/attachment.bin