Choreography component meeting agenda (9th March):

1. Actions from last meeting:
	- Joe will try to get in touch with his contacts in IBM to ask about mapping WSDL bindings to UML
	- Koustubh will talk to contacts in CA that are already on the BPEL TC about representing TPTP

2. Discuss the design of enhancement 50076

3. Review of test execution architecture, choreography architecture and discussion of the possibility of a choreography based TPTP test execution


Attending:
	Antony Miguel (Scapa)
	George Christelis (Scapa)
	Scott Schneider (IBM Rational)
	George Din (Fokus)
	Diana Vega (Fokus)
	Koustubh Parwar (CA)
	Shobhit Maini (CA)
	Augustus Kalimuthu (CA)
	Tejas Patel (CA)

Absent:	
	Harm Sluiman (IBM)
	Serge Lucio (IBM)
	Jim Saliba (CA)
	Kent Siefkes (IBM Rational)
	Joe Toomey (IBM Rational)
	Srinivas P Doddapaneni (Intel)



Point 1 discussion:
Joe is not here to discuss WSDL bindings to UML etc

Koustubh has spoken to the Oasis members in CA but they are not involved in BPEL.  Koustubh could join as an observer and then participate in the BPEL Technical Committee or he can join as a full member of the TC which would place demands on his time to attend the TC calls regularly.  Koustubh will speak to Jim about his joining the TC and will get back to the group later.


Point 2 discussion:
joe not here so can't really discuss 50076


Point 3 discussion:
Scott isn't ready to review the execution architecture, maybe we can do this another day.  Scott would like review of the choreography architecture though.

Choreography component came about as a way to run generic TPTP modeled behaviours.  Idea was that the runtime would be written based on existing web services standards and that later on a UML2 interactions to BPEL mapping would be produced which would enable us to run any TPTP modeled behaviours as BPEL processes.

Scott is unsure if this is a good fit for the test execution language he had in mind - think the test execution language is more of a high level language to specify deployment etc of the test.  Language outside the behaviour of the test rather than inside the behaviour.

Antony thinks there may still be a match as choreography does deployment differently.  Choreography has concept of a deployment server which the running test speaks to to find out where it should run things.  Choreography may describe both the high level behaviour and the low level behaviour.

Scott will look more into BPEL etc and the choreography component and get back to us.


Other discussion:

The examples in CVS have been broken due to a mismatch between the examples and an update to the choreography engine code. Have the examples in CVS had been updated yet?  No.  Antony is finishing off some transport layer development and will drop the new code and updated examples in the next day or two.

What is the purpose of the clock synchronisation messages?  The clock synchronisation messages do actually sync relative clocks on the engine controller and subcontrollers but they are legacy code.  The java engine had clock synchronisation but the BPEL running on the engine will probably not use the clock sync code.

How should a program be run on multiple machines?  Should Controller Launch Runner be called multiple times, once for each machine?  Normally within the engine the Controller Launch Runner would be called once and then further runners would be spawned by the first runner in the program using the methods in the RunnerInterface class.  Having said that this is not an incorrect use of the API.



Actions:
	- Antony should put the choreography docs on website (rather than in the plugin CVS)
	- Joe will try to get in touch with his contacts in IBM to ask about mapping WSDL bindings to UML
	- Koustubh will talk to Jim about joining the BPEL TC