Re: the precondition property in OWL-S 1.0

At 11:42 AM 09/11/2003 +0000, Austin Tate wrote:
>I would favour a much simpler underlying idea that an activity always has:
>
>a) an (implicitly) associated begin and end time point
>
>b) can optionally can be broken down into 0 or more sub-activities - if 0 
>its considered as a primitive activity in the current model
>
>c) a set of constraints of various types on and between the activity, the 
>objects manipulated by that activity and the time-points of the activity 
>and any of its sub-activities.  The constraints then include the simplest 
>activity ordering constraint "(before end-of(activity-1) to 
>begin-of(activity-2))" and world state preconditions/effects as 
>specialisations that are frequently used - but they are not the only ones.


I could have added that this is compatible with the approach taken by NIST 
PSL which can give us a nice extendable framework for the process model in 
SWSL.  Austin

Received on Sunday, 9 November 2003 06:57:59 UTC