Path: utzoo!utgpu!water!watmath!clyde!att!osu-cis!tut.cis.ohio-state.edu!mailrus!ames!decwrl!sun!quintus!ok
From: ok@quintus.uucp (Richard A. O'Keefe)
Newsgroups: comp.lang.prolog
Subject: Re: Stupid Question about UNH Prolog
Keywords: UNH,cut,tarsiers
Message-ID: <606@quintus.UUCP>
Date: 31 Oct 88 04:15:08 GMT
References: <7505@well.UUCP>
Sender: news@quintus.UUCP
Reply-To: ok@quintus.UUCP (Richard A. O'Keefe)
Organization: Quintus Computer Systems, Inc.
Lines: 13

In article <7505@well.UUCP> jax@well.UUCP (Jack J. Woehr) reports that
>foo(1).  foo(2).  foo(3).  foobar(X,Y) :- foo(X), !, foo(Y).
yields the answers (X,Y) = (3,3), (3,1), (3,2), (3,3), but that
>goobar(X,Y) :- goo(X), !, goo(Y).  goo(1).  goo(2).  goo(3).
yields the answers (X,Y) = (1,1), (1,2), (1,3) and asks
>	Is UNH Prolog broken or am I?

You should get the same answers from both.  However, the "foo" version
looks suspiciously as though there is an additional clause foo(3) at
the front.  Are you sure you tried this in an empty Prolog system?
If starting a fresh copy of Prolog, entering those clauses, and verifying
with the aid of listing/1 that those are the _only_ clauses for foo/1 and
foobaz/2 produces the result claimed, there is indeed a mistake.