> I feel very lucky to have stumbled on this interpretation of differentiation for datatypes with such potential for both utility and fascination, and has been intriguing me since the day I made the connection (whilst changing trains at Shrewsbury). This work has benefited greatly from hours spent on trains, [...]