I admit that I did evaluate the mapping toolbox from Mathworks under a
trial license. However, other than putting all the outputs into structure
arrays (an idea I think I would have arrived at on my own anyway), nothing
changed hands here. In fact I looked at no code at all (I only evaluated
the OUTPUT). If you don't believe me, just look at the C code for the
mex file. I defy anyone at Mathworks to write code as sloppy as that.