Re: Narrowing vs. rewriting

From: Sergio Antoy <antoy_at_cs.pdx.edu>
Date: Thu, 3 Jul 1997 15:16:41 -0700

> By the way, did you take my point about laziness?

Yes. What I meant was that a client of the Haskell permute is not
served a permutation, but a list of permutations. Of course, only
the portions of this list that are needed by the client are
actually computed. This would be the same using Needed
Narrwoing. The implications of the Haskell version is that the
client must implement code to traverse the list to find a
permutation that satisfies a give requirement, e.g., the safety of
n-queens. Writing this extra code is no big deal, but the Curry
version has the minor advantage of not requiring this code in the
first place.

Rgds,
Sergio
Received on Fr Jul 04 1997 - 02:19:00 CEST