#204 Inconsistent Matrix inversion interfaces

[Anonymous]

The interfaces for inversion in FMatrixC is different
from those in Matrix2|3dC. Charles's suggestion is to
use the implementation from Matrix2|3dC but with the
interface from FMatrixC, so that the faster Matrix2|3dC
ones overload the general FMatrixC one..

Also could add the Inverse(Matrix3dC &result) interface
to FMatrixC

To see the current mess, look at e.g.
http://www.ee.surrey.ac.uk/Research/VSSP/RAVL/share/doc/RAVL/Auto/Basic/Class/RavlN.Matrix3dC.html

Cheers,
Bill.