[Prev][Next][Index][Thread]

Wraith on recursive sums and products



In-Reply-To: Your message of Wed, 22 Mar 89 12:58:31 EST.
	     <8903221758.AA03957@stork.LCS.MIT.EDU>
Date: 22 Mar 89 13:14:08 PST (Wed)


	From: Gavin Wraith <gavinw%cogs.sussex.ac.uk@nss.cs.ucl.ac.uk>
	...
        What I ought to have said is that values of a recursive sum
        (initial algebra) A(T) are well-founded over T (can be
        expressed as a term involving constructors and T-values)
        whereas values of a recursive product (final coalgebra) B(T)
        are not in general.

Is strictness still meant to fit in here?

        We would expect quite different representations for values of
        type (U+V)*W from representations of values of type
        (U*W)+(V*W).

This would seem to contradict both theory and practice:  if two objects
are isomorphic then surely the same representation can serve for both.
-v