Known bugs

Main window and database

Multiple screens: does not work in all the configurations
Cause: Too many possible configuration possible (each version of Linux and Matlab handle this in a different way)
Workaround: work only with your main screen
Tree is not updated well and some nodes are missing
Cause: Unpredictible order of execution of Java/Matlab operations
Workaround: Press F5 to refresh
Colormap sliders for contrast / brightness do not work with versions of Matlab < 7.4
Workaround: Manipulate the colormap by clicking on the colorbar and moving up/down (brightness) and left/right (contrast)
Some user preferences are not saved in compiled versions
Tree menu: "File > XTerm" does not work with all the linux distributions
Tree menu: "Group conditions" does not update all the filename links (in results files, linkresults files, etc.)
Display time series t=[0,end] + Display time series t=[begin,end]: time bounds are not good for the second figure
Load simultaneously datasets with slightly different sample rates: does not work
Save figures as AVI movies: only available on Windows operating system

Stat: t-values cannot be displayed on 3D MRI slices
Figure menu "Set colormap max": does not work from the 3D figures with some versions of Matlab (eg. R2007a)
Cloned figures do not behave well
Example: Clone + scouts + moving in time : currents on cortex are not updated
Display topography + display channels from the tree : crashes if there are bad channels
Channel selection on 2D Residuals : don't work properly for magnetometers only (indices issue)

Tree menu "Project on default anatomy" : does not work for kernel / absolute values
Tree menu "Project on default anatomy" : does not work for kernels if there are some bad channels
OS_MEG crashes Matlab when scalp in not very well aligned with sensors
Workaround: align your them well...
OS_MEG (overlapping spheres) fails when the scalp surface is not perfect (some garbage inside the head)
Workaround: do your scalp extraction with FreeSurfer
"Project on default anatomy" cause Out of memory errors when processing FULL results (=> proceed by time blocks)

Grand average computation does not weight the different subjects with the number of trials for each subject
Rounding errors with time values: sometimes there are +/- 1 samples selected than expected