RE: OWL-S difficulty

Charlie,

Your question raises some quite basic questions about the semantics of
OWL-S. Does it describe (a) the ways in which 'a client may interact with a
service' or is it (b) an orchestration language to be followed by an agent
behind the scenes. In other words is it front-of-shop or back-of-shop. It
appears that the recent 'Modelling Services as Processes' paper from Drew
McDermott finally lays this to rest with answer (a).

In that case, it appears as though OWL-S 'Choice' captures exactly the
intended semantics. The choice is that of the client, so needn't be
expressed explicitly as a condition. From the point of view of the process
description the choice is non-deterministic, but we have to be clear about
who the non-determinism belongs to (in this case the client). The semantics
is that one of A, B, C will occur within the process context of D. Given
that, I don't see the need for an additional atomic service to make the
choice indirectly on behalf of the user.

Steve.

> -----Original Message-----
> From: public-sws-ig-request@w3.org
> [mailto:public-sws-ig-request@w3.org]On Behalf Of Charlie Abela
> Sent: Friday, July 02, 2004 5:36 PM
> To: public-sws-ig@w3.org
> Subject: OWL-S difficulty
> 
> 
> 
> 
> Hi,
> 
> I like to pose this difficulty with OWL-S. 
> I have a set of processes that can be combined into a single 
> composite one.
> Consider services, A, B and C that can be composed into service D. 
> The control construct that can be applied could be an 
> if-then-else or even a
> while. Consider also that the 3 subprocesses A, B and C are 
> independent of each
> other and the user has to make a choice on which one to 
> execute. Would the most
> suitable solution include some other atomic service that 
> given some user's
> choice invokes one of A, B or C? Or possibly there can exist 
> some OWL-S
> construct, some special choice, which has to be added to perform this?
> 
> regards
> 
> Charlie
> 
> -------------------------------------------------
> This mail sent through IMP: http://horde.org/imp/
>

Received on Monday, 5 July 2004 07:40:23 UTC