Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!rutgers!netnews.upenn.edu!linc.cis.upenn.edu!dowding
From: dowding@linc.cis.upenn.edu (John Dowding)
Newsgroups: comp.lang.prolog
Subject: Re: incrementing values
Message-ID: <19930@netnews.upenn.edu>
Date: 3 Feb 90 17:51:37 GMT
References: <17467@megaron.cs.arizona.edu> <31462@shemp.CS.UCLA.EDU>
Sender: news@netnews.upenn.edu
Reply-To: dowding@linc.cis.upenn.edu (John Dowding)
Organization: University of Pennsylvania
Lines: 28

In article <31462@shemp.CS.UCLA.EDU> coleman@cs.ucla.edu (Michael Coleman) writes:
>In article <17467@megaron.cs.arizona.edu> debray@cs.arizona.edu (Saumya K. Debray) writes:
>>This is an excerpt from an article posted recently to another newsgroup:
>>| 	increment(Variable) :-
>>| 		retract(value(Variable, OldValue)),
>>| 		NewValue is OldValue + 1,
>>| 		assert(value(Variable, NewValue)).
>>Comments?
>
>I saw this in the other newsgroup.  My jaw dropped when I read it.
>Incredible!  Kind of makes you wonder how this poster would implement, say,
>quicksort, or nested "for" loops in Prolog.
>

Just to set the record straight, Saumya Debray did not properly attribute the
excerpt to Dr. David Matuszek at Unisys (dave@prc.unisys.com).  Michael 
Coleman took this excerpt completely out of context.  Dr. Matuszek is an
exceptional computer scientist and programmer whose skills are beyond 
question.  David was arguing against the notion proposed in comp.lang.misc
that some varient of the  construct 'x = x + 1' was most common expression
in any programming language.  He offered as an example Prolog, saying that
he rarely had use for the equivalent of that construct in his 5 years of
professional Prolog programming.  His experiences are consistent with  
mine, that this construct is rarely useful in Prolog.

As an aside, to implement 'x = x + 1' in Prolog in full generality (in
the case where x is a global variable) does require some sort of 
side-effecting construct in most Prolog implementations.