From: Assaf Arkin <arkin@intalio.com>

Date: Wed, 16 Apr 2003 11:27:34 -0700

Message-ID: <3E9DA096.8000706@intalio.com>

To: "Burdett, David" <david.burdett@commerceone.com>

CC: "'Howard N Smith'" <howard.smith@ontology.org>, public-ws-chor@w3.org

Date: Wed, 16 Apr 2003 11:27:34 -0700

Message-ID: <3E9DA096.8000706@intalio.com>

To: "Burdett, David" <david.burdett@commerceone.com>

CC: "'Howard N Smith'" <howard.smith@ontology.org>, public-ws-chor@w3.org

Burdett, David wrote: >If your "formalism" is PI Calculus (which is what I think you are >suggesting), then does it impose any constraints on the types of tools you >can use with it? > Let's change perspective for a second. pi-calculus is mathematical model for looking for looking at certain things and drawing conclusions about them. It can cover a wider definition of processes than say CCS. And it has appeal because it can cover a wide spectrum of definitions from the more complex scenarios involving calculations down to the level of a single unit of processing (pi-caclulus also covers lambda calculus). Mathematical models provide forms of analysis and proofs, in this particular way through a form of calculation (hence the name calculus). By themselves they do not provide ways to construct things. If you want to construct a process defintion that you can exchange, monitor and otherwise manage, you would have to look at languages like BPEL, BPML and WSCI. If you want to construct a process definition that you can depict visually - and this is certainly an interesting thing to do - you would look at visual notations like BPMN and UML. There is no constrain on the things you can do or the type of tools you can use. What you get is the ability to pick the right tool for the job and use the formalism to draw conclusions about it. For example, you may decide on a tool that can only define bi-lateral choreographies because that happens to be your main interest at some point in time. The formalism will address that. At some other point in time you may want to go further and add multi-party collaboration, or you may want a tool that can show you the n-party collaboration along with your implementation process that takes part in it. The tool needs to be more sophisticated, and you need some formalism that goes both ways. pi-calculus is interesting because it allows that formalism to exist, and at the same time allow these variety of tools and languages to use that formalism. arkin >David > > >Received on Wednesday, 16 April 2003 14:29:49 UTC

*
This archive was generated by hypermail 2.3.1
: Tuesday, 6 January 2015 21:00:58 UTC
*