Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!usc!samsung!rex!fs
From: fs@rex.cs.tulane.edu (Frank Silbermann)
Newsgroups: comp.lang.functional
Subject: Re: Patterns, Equations and Functions
Message-ID: <3486@rex.cs.tulane.edu>
Date: 4 Jun 90 17:24:03 GMT
References: <14921@dime.cs.umass.edu>
Organization: Computer Science Dept., Tulane Univ., New Orleans, LA
Lines: 30


In article <14921@dime.cs.umass.edu> pop@cs.umass.edu () writes:
>
>	Surely one important point of functional programming
>	is that it provides a basis for reasoning about programs,
>	and for program transformation.  E.g. ML and Haskell
>	do not feature a defined, manipulable internal representation
>	of programs, in the way that LISP and Prolog do.
>
>	Regarding the equational definitions of functions
>	as _more_ than syntactic sugar provides support
>	for reasoning about programs in a way that concurs
>	with people's basic mathematical training.

This would be nice, but unfortunately, the resulting languages
do not seem to be computable (not merely difficult, but impossible,
at least in the higher-order case).

This is why I consider pattern-matching to be a syntactic sugar;
a shorthand for programmers which provides a non-enforced documentation
of his intentions.

>	Thus it would seem healthy to treat such definitions
>	as potential inputs to a term-rewriting system.

What does term-rewriting have to do with any theory
of domains and mappings (e.g. programming with functions)?

	Frank Silbermann	fs@rex.cs.tulane.edu
	Tulane University, New Orleans,  Louisianna, USA