Re: Encapsulated search does not encapsulate (all)non-determinism

From: Wolfgang Lux <lux_at_wi.uni-muenster.de>
Date: Tue, 15 Jan 2002 11:13:42 +0100

Michael Hanus wrote

> If this is the case (and it actually seems to be the case),
> one has to decide what is more important: sharing or encapsulation?
> As Frank already pointed out, there must be a possibility to
> encapsulate all non-determinism in order to write reliable
> interactive programs. Thus, if my option 3 is the only possibility
> to do this, I am still in favor to commit to it, even we loose
> sharing or, as you pointed out, the results depend on the evaluation
> order. Note that this is in some sense already be the case:
> if you change the order of defining equations (which does not change
> the declarative meaning of a function), you obtain the answers
> in a findall in a different order. This is due to the fact that
> "try" is a meta-level construct which depend on the evaluation order
> of the underlying program.

This is true. And in fact the situation is even worse. You cannot even
rely on findall returning the solutions of the simple goal

  findall (\x -> x=:=coin)

in the textual order of the definition of coin, i.e. [0,1]. An implementation
might choose to return the solutions in any order it finds suitable (though
probably all existing implementations return the solutions in their textual
order).

But IMHO this is still something different than returning completly different
solutions depending on the order of evaluation. And I cannot imagine how to
write "reliable" programs, if the insertion of a flip operation (e.g. using --
probably indirectly -- flip (==) instead of (==) to compare two values) will
not only change the order of results but also whether some results are
produced at all.

> It is unclear to me what do you mean by "sound"?
> "Try" is related to the meta-level, so it might be difficult to
> talk about soundness for "try".

By unsound in this case I mean that y /= y can be true under option 3. I find
this unacceptable even for a meta-level construct in the language. Therefore
IMHO option 3 is not an option.

Regards
Wolfgang



--
Wolfgang Lux				  Phone: +49-251-83-38263
Institut fuer Wirtschaftinformatik	    FAX: +49-251-83-38259
Universitaet Muenster		      Email: wlux_at_uni-muenster.de
Received on Wed Jan 16 2002 - 02:10:46 CET

This archive was generated by hypermail 2.3.0 : Thu Sep 19 2019 - 07:15:05 CEST