Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!samsung!munnari.oz.au!lee
From: lee@munnari.oz.au (Lee Naish)
Newsgroups: comp.lang.prolog
Subject: Re: delayed computation challenge
Keywords: NU-Prolog delays quicksort
Message-ID: <2950@munnari.oz.au>
Date: 21 Dec 89 07:31:57 GMT
References: <2235@cs-spool.calgary.UUCP> <2948@munnari.oz.au>
Sender: news@cs.mu.oz.au
Reply-To: lee@munmurra.UUCP (Lee Naish)
Organization: Comp Sci, University of Melbourne
Lines: 34

In article <2948@munnari.oz.au> lee@munmurra.UUCP (Lee Naish) writes:
>10?- qsort(X, [1, 3, 2]).
>Warning: Goal floundered.
>X = X 
>
>It would be nice to run the goal backwards inside the debugging
>environment, which has a facility for testing goals using a fair search
>strategy.  Unfortunately it doesn't seem to work for this example

It seems I was using an out of date version of the debugging
environment.  It actually does work:

8?- test qsort(X, [1, 2, 3]).
Do static analysis? n
Test known valid instances? 
Test known satisfiable atoms? 
Test known unsatisfiable atoms? 
Generate solutions to goal? 
Solution found: qsort([3, 2, 1], [1, 2, 3]).
qsort([3, 2, 1], [1, 2, 3])     Valid? 
Solution found: qsort([3, 1, 2], [1, 2, 3]).
qsort([3, 1, 2], [1, 2, 3])     Valid? 
Solution found: qsort([2, 1, 3], [1, 2, 3]).
qsort([2, 1, 3], [1, 2, 3])     Valid? 
Solution found: qsort([2, 3, 1], [1, 2, 3]).
qsort([2, 3, 1], [1, 2, 3])     Valid? 
Solution found: qsort([1, 3, 2], [1, 2, 3]).
qsort([1, 3, 2], [1, 2, 3])     Valid? 
Solution found: qsort([1, 2, 3], [1, 2, 3]).
qsort([1, 2, 3], [1, 2, 3])     Valid? 

It gets into the (unavoidable) infinite loop after this.

	lee