> However, ruling out unifications between two partial applications means that we would have to rule out unifications between a variable and a partial application too;
Yes, I would prefer to rule out every unification between arguments of
functional type and only support those unifications that are specified
in the Curry Report, that is, between data terms and logic variables
and data terms.
If it is not possible to rule out unification of two logic variables
of functional type without type information, I think it is fine to let
them succeed as, in practice, entually such variables will eventually
get unified with actual functions (which would then fail.)
> But generating a runtime error (or failure) on unifications like x =:= map id when x is a free variable IMHO is a too high price to pay
Can you give an example that demonstrates how high the price is?
Sebastian
_______________________________________________
curry mailing list
curry_at_lists.RWTH-Aachen.DE
http://MailMan.RWTH-Aachen.DE/mailman/listinfo/curry
Received on Mi Jun 20 2012 - 17:30:35 CEST