Re: Text related to fault handling discussion

MessageTony,

The same solution applies to multiple responses as it does to multiple fault messages. The issue in terms of fault handling is simply how the faults are uniquely identified. In the same way as my solution to your issue could be achieved (i.e. separating the different named responses into different interactions, as separate elements in a choice), the same could be done for faults - just by placing each separate fault exchange into separate interactions which are elements of a choice. Higher level model checking could provide the coupling of the different components for an MEP.

There are a lot of similarities between faults and the potential for multiple response messages - both need to be uniquely identified by a name - which means that I think the exchange needs to reference this name.

Regards
Gary
  ----- Original Message ----- 
  From: Tony Fletcher 
  To: 'Gary Brown' ; public-ws-chor@w3.org 
  Sent: Tuesday, March 01, 2005 6:08 PM
  Subject: RE: Text related to fault handling discussion


  Dear Gary,

  But in your discussion with me about multiple responses to a request, your solution, which I came to agree with was to allow different parts of the same operation to be in different interactions (but within the same workunit?  We need to state how far spread the operation can be) so we can take advantage of the current implicit sequencing and the explicit choice construct.  So does not this mean that CDL does have the same problem now?

  Unconsciously, perhaps that is why I was preferring the solution of having a slightly more complex syntax allowed within an interaction bracket so that the grouping of the request, response(s) and fault message(s) was clear.

  Best Regards     Tony
  A M Fletcher

  Cohesions  (TM)

  Business transaction management software for application coordination       www.choreology.com

  Choreology Ltd., 68 Lombard Street, London EC3V 9LJ     UK
  Tel: +44 (0) 1473 729537   Mobile: +44 (0) 7801 948219
  tony.fletcher@choreology.com     (Home: amfletcher@iee.org)
  -----Original Message-----
  From: public-ws-chor-request@w3.org [mailto:public-ws-chor-request@w3.org] On Behalf Of Gary Brown
  Sent: 28 February 2005 17:41
  To: public-ws-chor@w3.org
  Subject: Text related to fault handling discussion


  This text is from a recent WS-BPEL draft:

  "Because WSDL does not require that fault names be unique within the namespace where the service operation is defined, all faults sharing a common name and defined in the same namespace are indistinguishable in BPEL4WS. In WSDL 1.1 it is necessary to specify a portType name, an operation name, and the fault name to uniquely identify a fault. This limits the ability to use fault-handling mechanisms to deal with invocation faults. This is an important shortcoming of the WSDL fault model that will be removed in future versions of WSDL."

  From my understanding, BPEL is interested in being able to handle a fault message independently from any specific operation, and that is why they have raised this as an issue.

  However, from a CDL perspective, fault messages are always handled within an interaction that does have the relevant context (i.e. the port type and operation name) - and therefore adding the faultName to a respond exchange would "uniquely identify a fault". Unlike BPEL, CDL does not require faults to be globally unique, as they are alway handled in the context of an operation.

  Regards
  Gary