Thomas Lord wrote:
I'd also like to see an attempt made to eliminate the store and
continuation passing style aspects of the semantic mapping because I
suspect that this poses additional problems.
I share a strong sense that the construction of a store, in the
familiar way, is not the best way to model things.
In particular, btw, it seems like there should be a kind of "coordinate
free" way to model mutation in which, instead of locations, we talk about
procedures that implicitly accept a parameter which represents an
equivalence class of possible points on a presumed world-line of a
computation selected from a space-time lattice of observable events
implied by a program's execution.
And, damn, as I write that I find myself now wondering if that
doesn't also give an alternative to continuations.
If you know what I mean.... :-) [then perhaps we are onto something].
-t
_______________________________________________
r6rs-discuss mailing list
[email protected]
http://lists.r6rs.org/cgi-bin/mailman/listinfo/r6rs-discuss
