Path: utzoo!attcan!uunet!lll-winken!lll-tis!helios.ee.lbl.gov!pasteur!ucbvax!decwrl!labrea!sri-unix!quintus!ok
From: ok@quintus.uucp (Richard A. O'Keefe)
Newsgroups: comp.lang.prolog
Subject: Re: Determining order of argument unification
Message-ID: <647@quintus.UUCP>
Date: 8 Nov 88 09:04:20 GMT
References: <8192@burdvax.PRC.Unisys.COM> <629@quintus.UUCP> <8210@burdvax.PRC.Unisys.COM>
Sender: news@quintus.UUCP
Reply-To: ok@quintus.UUCP (Richard A. O'Keefe)
Organization: Quintus Computer Systems, Inc.
Lines: 21

In article <8210@burdvax.PRC.Unisys.COM> lang@zeta.PRC.Unisys.COM (Francois-Michel Lang) writes:
>> In article <8192@burdvax.PRC.Unisys.COM> lang@zeta.PRC.Unisys.COM (Francois-Michel Lang) writes:
>> >Is it possible to determine (by writing a Prolog program)
>> >which of these two orders is used for given a Prolog system,
>if we assume that the Prolog system one is running
>(1) does indeed use left -> right or right -> left order,
>(2) uses it consistently, and
>(3) can't handle cyclic terms.

Why not simply assume that it uses right->left order consistently and make
the problem _really_ easy?  My earlier reply did not exhaust all the things
that a Prolog system might sensibly do instead.  Imagine a Prolog system
where the compiler replaces all variables in the head by new distinct
variables, and then at the end unifies repeated variables pairwise, e.g.
	p([H|T], L, [H|R]) :-
=>	p([H|T], L, [H'|R]) :- H = H'
If the goal is acyclic, the (simplified) head unification is order-
independent; what matters is the order in which the repeated variables
are handled.  And that might quite plausibly depend on the spelling of
the source variables.

I tell you plainly that Quintus Prolog violates Lang's assumptions.