Hi, I assume this is a known bug (or is it a feature), that the atan2-object takes its arguments twisted. like [3 2( --> [atan2] is not the same as [expr atan2(3,2)]. what lovely source of error... marius.