wsbpel-comment message

Subject: Public Comment

From: comment-form@oasis-open.org
To: wsbpel-comment@lists.oasis-open.org
Date: 8 Apr 2004 21:33:33 -0000

Comment from: cindymcnally_6@hotmail.com

When a reply is nested within a faultHandler and the reply specifies a faultName, it is unclear as to how an implementation should validate this name or how the fault should be defined within the wsdl portType for the process itself. 

As an example suppose the faultHandler specifies no name, but instead just a faultVariable and faultMessageType.  To what do we compare the faultName in the reply?  The source of the fault could be from any invoke/throw within scope. 

The same difficulty applies when the reply is nested within a catchAll element, but is a bit worse.  With a catchAll there is no declared 'fault variable' which the reply may reference.  Should faultName be dissallowed on a reply in this situation?