Possible mouse mapping architecture quad-mesh -> quad-mesh transformations?

[Alan Cox]

Ar Mer, 2006-06-28 am 13:17 +0200, ysgrifennodd Keith Packard:
> We've explored a number of possible architectures to handle the mapping
> of mouse coordinates in a transformed desktop environment. The constant
> assumption in each of these architectures has been that a user
> application must be involved in the transformation.

What does this mean for physical feedback devices. There is a separate
mapping problem of going from a mouse movement to feedback (whether that
is the mouse braking as you 'hit' a wall, or rolling "around" an object)

Alan