Re: the precondition property in OWL-S 1.0

   [Monika Solanki]
   > I think we should retain the Precondition class and the hasPrecondition
   > property. This is because, although Precondition is effectively a
   > Condition, however it is a "special" kind of Condition. In the process
   > model, Condition is a general thing, which  is also  used for
   > Conditional Effects and Conditional Outputs.

   [Sheila McIlraith]
   I agree. We may want to syntactically restrict the types of preconditions
   we have.  This affords us that flexibility.

That would make sense.

But what Monika actually said is a bit different.  In the quoted
paragraph, she suggested that Condition encompassed preconditions, plus
conditional-effect and conditional-output conditions ("secondary
preconditions").  I've been assuming that a Condition was an arbitrary
formula.  After all, it's not a subclass of anything (according to
Process.owl).

It might be reasonable to have a class Formula, and a subclass
Condition that was restricted syntactically somehow.  I'd be very
surprised if there were a syntactic difference between primary and
secondary preconditions.

                                             -- Drew

-- 
                                   -- Drew McDermott
                                      Yale Computer Science Department

Received on Monday, 10 November 2003 18:09:39 UTC