BPEL-42: proposed resolution

[Anish Karmarkar <Anish.Karmarkar@oracle.com>]

Based on the discussion [1] on this issue, I would like to propose that 
we resolve issue 42 with the following resolution:

Replace SBPEL3004 with the following --
"In order to have an SCA multi-valued reference in a WS-BPEL process, a 
partnerLink that uses the sca-bpel:multiRefFrom extension attribute is 
declared in the process. The value of this attribute points to a 
variable declaration in the process that contains the 
sca-bpel:multiReference extension element. The multiplicity of the SCA 
reference that corresponds to such a partnerLink depends on the value of 
the multiplicity attribute of the sca-bpel:multiReference element in the 
variable declaration."

While coming up with a proposal for this issue, I discovered that:

1) there is no required in the spec that if there is a 
sca-bpel:multiRefFrom extension attribute on the partnerLink then it 
must result in a sca reference in the introspected CT.
2) the default value for sca-bpel:multiReference/@multiplicity (which is 
1..n) is not considered in the algorithm that generates the CT.

I'll raise separate issues for this.

-Anish
--

[1] http://lists.oasis-open.org/archives/sca-bpel/200905/maillist.html