[ogsa-naming-wg] Question on WS-Naming

[Mark McKeown]

Hi folks,
         I have been reading the WS-Naming spec and was
wondering why the authors chose not to use WSRF?

If I have an EPR and I want to provide a resolver for
it I could use a WS-Resource to implement the resolver
for that EPR.

The client would only have to use the standard WSRF get*
operations to retrieve a new EPR - the EPR would be the
ResourceProperty document. The reuse of WSRF would
also add support for updating the resolver WS-Resource
through the standard WSRF operations.

The EPR of the resolver WS-Resource could be embeded in the
original EPR as is done already in the WS-Naming spec.

This approach seems to have a number of advantages to me:
reuse of WSRF and WSRF client tooling, no longer any need
to try and compare EPRs, de-couples the need for
naming from resolving EPRs (naming is of course important
in its own right), simplifies the WS-Naming spec
(no need for WSDL).


cheers
Mark

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Mark Mc Keown                            RSS
Mark.McKeown at man.ac.uk 	                 Manchester Computing
+44 161 275 0601     		         University of Manchester
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~