🚀 go-pugleaf

>In article <1171@tuhold> thom@tuhold (Thom Fruehwirth) writes:
>>As Jamie Andrews in <628@etive.ed.ac.uk> pointed out, O'Keefe probably
>>intended to have a cut between random_permutation/2 and member/2.

     Actually, the point I was making was that
random_permutation/2 would either return 7! results (not good)
or else introduce additional incompleteness into the system
(returns only one of the possible results) which random clause
selection would not.  If we accept the incompleteness arising
from allowing cut, though, it wouldn't make much difference that
random_permutation/2 returned only one result, since the effect
of it would be the same as if it were followed by a cut.

In article <308@quintus.UUCP> ok@quintus.UUCP (Richard A. O'Keefe) writes:
<random(-X) generates *ONE* pseudo-random float uniformly distributed
<between 0.0 and 1.0.  You wouldn't expect it to generate another if you
<backtrack into it.  Why should random_permutation/2 be any different?

     I agree, from a practical point of view.  But I would
rather have the whole thing hidden, or the randomization
explained by a hidden seed parameter, or something like that,
just because I'm perversely theoretical.  (Trilogy anyone? :-)) 
BTW, (referring to a previous ok article) I also agree that
having random clause selection might not help very much for some
tasks requiring randomization (such as selecting a random
element of an arbitrary-length list) and that specific solutions
might need to be found for them.

--Jamie.
  jha@lfcs.ed.ac.uk
"I always find mathematics so difficult" -- David Hilbert