- From: Drew McDermott <drew.mcdermott@yale.edu>
- Date: Fri, 26 Nov 2004 23:52:24 +0000 (GMT)
- To: public-sws-ig@w3.org
> [Daniel Elenius]
>
> Are InputBinding and OutputBinding supposed to be an exhaustive
> partition of Binding, i.e. are all Binding either InputBinding or
> OutputBinding?
>
> Binding has a toParam property, with Parameter as its range.
> InputBinding restricts the property value to Input, and OutputBinding
> restricts it to Output. But Parameter also has subclasses Local and
> ResultVar. Can we create Bindings for these? I *think* the answer should
> be no, and we should make this explicit by defining Binding as the union
> of InputBinding and OutputBinding.
I agree
> Especially since the comment in Process.owl says:
> "Bindings are used in two ways in this process modeling ontology. They
> are used to specify how output parameters are specified in different
> result conditions for Atomic Processes, and they are used to specify
> how input parameters acquire values when invoked by Perform's in
> composite process descriptions."
The comment should be changed to reflect the fact that Produces can
also specify how output parameters are to be set.
>
> I need to know this for the tool we're building. Specifically, when
> adding data flow (hasDataFrom declarations), should the user only get to
> choose from Inputs and Outputs (I hope so), or also Locals and ResultVars?
Locals and ResultVars are bound in preconditions and result
conditions, so you're right.
-- Drew
--
-- Drew McDermott
Yale University
Computer Science Department
Some see the glass as half empty, some as half full. Surely the
truth lies somewhere in between.
Received on Friday, 26 November 2004 23:53:03 UTC