Re: RFC: implementation of package pools
On Mon, Oct 23, 2000 at 08:36:46PM +0300, Eray Ozkural wrote:
> Yeah, I noticed that. Also in many cases it seemed to me that the
> sentence becomes something simple like
> [notation: ! is negation, /\ conjunction, \/ disjunction ]
> P /\ Q /\ !R .... /\ Z
> (I know you don't like my ASCII art, but no logic symbols on the keyboard)
> which is just conjunction of propositional symbols or their negations.
> In this case, it's already known which values the symbols would take.
> In many other cases we have simple conjunctive forms of the sort
> (P \/ Q) /\ (S \/ !T) /\ X
> in which there's little or no overlap of propositional symbols among
> conjunctions. [makes the instance quite easy to solve]
> and correct me if I'm wrong, but the propositional sentences are
> unrestricted CNF (conjunctive normal form), so it really is SAT.
> the sentences in the database might be easy instances on the other hand.
[suggestion to use AI algorithms for this sort of thing]
This stuff sounds much like logic programming to me. Not that anyone
would listen to a suggestion involving advanced languages, but it may
well be possible to reuse some sort of logic programming engine, or,
even easier, write this in Prolog (or some other logic language).
Vast amounts of additional generality could also be implemented this way.
Cheers,
Bill
--
"... the most important thing in the programming language is the name.
A language will not succeed without a good name. I have recently
invented a very good name and now I am looking for a suitable language."
-- D. E. Knuth, 1967
Reply to: