Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!lll-winken!uunet!munnari!murtoa.cs.mu.oz.au!munnari.oz!lee
From: lee@munnari.oz (Lee Naish)
Newsgroups: comp.lang.prolog
Subject: Re: Constraint satisfaction in LOG(F)
Message-ID: <1353@murtoa.cs.mu.oz.au>
Date: 10 Apr 89 01:29:30 GMT
References: <1632@kulcs.kulcs.uucp> <1942@randvax.UUCP>
Sender: news@cs.mu.oz.au
Reply-To: lee@munmurra.UUCP (Lee Naish)
Organization: University of Melbourne, Comp Sci Dept
Lines: 37

In article <1942@randvax.UUCP> narain@randvax.UUCP (Sanjai Narain) writes:
>It is interesting that the NAC compiler can generate coroutining
>information automatically.  I would be interested in knowing how general
>it is.  Does the user ever have to specify coroutining annotations, or
>does the compiler always do the right thing?  Can it handle arbitrary
>symbolic generate and test predicates, or did it do a good job because it
>knew about +?

It works fairly well for programs which manipulate recursive data
structures (lists in this example).  It doesn't know anything about
numbers, which is a limitation.  See

%A Lee Naish
%T Automating control of logic programs
%J Journal of Logic Programming
%K jlp
%V 2
%N 3
%D October 1985
%P 167-183

It might be easier to automate this kind of coroutining in a functional
environment because more information is known about modes.  However, one
of the prices to pay is that tests must explicitly return true or false
(rather than just succeeding or failing).  Predicates seem nicer than
functions for expressing "generate and test" problems.

Although coroutining can simulate some uses of lazy evaluation (perhaps
more efficiently also), lazy evaluation seems more powerful in general.
This is something that functions have over predicates (in my opinion).
However, I think logic programming is a good implementation vehicle for
mixed "logic"/functional programming.  Quite a while back I wrote a
preprocessor which compiles NU-Prolog + functions defined by equations
down to NU-Prolog.  It has (optional) lazy evaluation and fits very
nicely with Parallel NU-Prolog, but thats another story...

	lee