Aloha fellow Surfers,
while testing the dev20060224 centos4_x86_64 release I stumbled upon some issues with tkregister2. That is tkregister2 produces X errors leading to a black screen once the image window is clicked into. Since I try to cheat monkey data into the freesurfer chain I use tkregister2 to register my structurals with some atlas based assumptions (basically rotating the volume slightly around three axis). Then I use mri_vol2vol to reslice a new rotated volume. And this is where tkregister2 chokes: Received X error ! Error code : 10 Request code : 89 Minor code : 0
Error test : ´BadAccess (attempt to access private resource denied)´
There are plenty of these errors, and once the image window is clicked into the screen turns black, only the mouse survives. Remotly login in and killing the tkregister2 process gives back control over the machine. Any hints besides "don't do that then" ;)?
ahoi Sebastian
Hmmm, I have not seen anything like this around here. Does it only happen with your monkey data? What are the sizes of the volumes you are trying to run? Can you send all your command lines?
doug
Sebastian Moeller wrote:
Aloha fellow Surfers,
while testing the dev20060224 centos4_x86_64 release I stumbled upon some issues with tkregister2. That is tkregister2 produces X errors leading to a black screen once the image window is clicked into. Since I try to cheat monkey data into the freesurfer chain I use tkregister2 to register my structurals with some atlas based assumptions (basically rotating the volume slightly around three axis). Then I use mri_vol2vol to reslice a new rotated volume. And this is where tkregister2 chokes: Received X error ! Error code : 10 Request code : 89 Minor code : 0
Error test : ´BadAccess (attempt to access private resource denied)´There are plenty of these errors, and once the image window is clicked into the screen turns black, only the mouse survives. Remotly login in and killing the tkregister2 process gives back control over the machine. Any hints besides "don't do that then" ;)?
ahoi Sebastian
Sebastian,
Which platform are you running the centos4_x86_64 release upon?
Also, can you send me the output of this command:
ldd `which tkregister2`
Also, if not running on a 64bit Centos4 platform, then you could attempt the following:
cd $FREESURFER_HOME/lib mv tcltktixblt backup-tcltktixblt
open a new terminal and source freesurfer and retry tkregister2
The freesurfer distribution contains Tcl/Tk/Tix/BLT libs which are tested to work with the freesurfer gui apps, but maybe you are experiencing some compatibility issue. By moving this libs dir, freesurfer will use whatever is installed already on your system (which may or may not work with tkregister2).
Nick
On Mon, 2006-02-27 at 12:05 -0500, Doug Greve wrote:
Hmmm, I have not seen anything like this around here. Does it only happen with your monkey data? What are the sizes of the volumes you are trying to run? Can you send all your command lines?
doug
Sebastian Moeller wrote:
Aloha fellow Surfers,
while testing the dev20060224 centos4_x86_64 release I stumbled upon some issues with tkregister2. That is tkregister2 produces X errors leading to a black screen once the image window is clicked into. Since I try to cheat monkey data into the freesurfer chain I use tkregister2 to register my structurals with some atlas based assumptions (basically rotating the volume slightly around three axis). Then I use mri_vol2vol to reslice a new rotated volume. And this is where tkregister2 chokes: Received X error ! Error code : 10 Request code : 89 Minor code : 0
Error test : ´BadAccess (attempt to access private resource denied)´There are plenty of these errors, and once the image window is clicked into the screen turns black, only the mouse survives. Remotly login in and killing the tkregister2 process gives back control over the machine. Any hints besides "don't do that then" ;)?
ahoi Sebastian
freesurfer@nmr.mgh.harvard.edu
-
Doug Greve -
Nick Schmansky -
Sebastian Moeller