I will be starting soon to work again on a driver to control our new
robuCARTT. The only thing I'm interested in is sending speed and steering
commands. In particular, I'm not interested in maintaining odometric
positioning, partly because I don't need it, but mostly because with the kind
of terrain this will be driving on, such positioning will be very unreliable
As far as I know, position2d is the only interface that would allow me to send
speed and steering control to the robot but a complete implementation of the
interface would also require the positioning information.
I can obviously create a new interface (although the tutorial to do that is to
be written), but I was wondering if it wouls make sense in a more general
case to separate the actual control of robots from their positioning and
therefore have different interfaces for that.
What do people think?