[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: RE: [wsbpel] Issue R25 - Proposal to vote
The parallel <forEach> case from the
submitter would not fault with a conflictingReceive because each iteration of a
parallel <forEach> is a separate copy of the <scope>. As such, the
conflcitingReceive test passes since they are different partner links and
different correlation sets. The description of a parallel <forEach>
describes the <forEach> as being equivalent to a <flow> with one
copy of the <scope> for each iteration. The test for conflictingReceives is not
sufficient to prevent all cases of ambiguity during message dispatch but it
serves to remove one well defined case. In order to handle any other cases, we
introduced a separate fault called ambiguousReceive which is thrown if an
ambiguity is detected during the message dispatch. The proposal below expands the test for conflictingReceive
to include a check of the values of a partner link and correlationSet(s). As I
mentioned during the call, I am concerned about using the term “values”
without defining it, especially with a normative MUST. For example, what’s
the value of a partner link? I’m also opposed to the possible amendment
since it’s not portable. I would prefer that the behavior is consistent
across all implementations. I think the existing text is fine as it is. From: Diane Jordan
[mailto:drj@us.ibm.com]
|
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]