FW: [wsrf] Identity

["Vambenepe, William N" <vbp@hp.com>]

Hi all,

Forwarding an email from WSRF about our identity capability. Sam
explains some of his requirements and wonders whether this belongs in
WSRF or WSDM. Any thoughts from the group?

William

-----Original Message-----
From: Samuel Meder [mailto:meder@mcs.anl.gov] 
Sent: Monday, October 04, 2004 8:57 AM
To: wsrf-oasis
Cc: Steve Tuecke; Karl Czajkowski; Frank Siebenlist
Subject: [wsrf] Identity


I brought up the issue of resource identity at the f2f (for e.g. writing
policy about a resource) and was pointed at the work done in the WSDM
TC. I've now taken a look at the WSDM spec and while it solves some of
the issues I see (at least) a couple of problems:

      * Optimally this work should be factored out of WSDM. I believe
        that the solution to the problem of resource identity has wider
        applicability than just the management space so would benefit
        from separate treatment. Would this be possible within the
        charter of the WSRF TC? Any thoughts from WSDM?
        Another somewhat related issue is that support for identities
        should be more pervasive in the base specs, if only just to
        avoid chatty interfaces (e.g. a reply from any of the factory
        methods in the WSRF/WSN specs should also optionally return an
        identity). 
      * The WSDM spec does not address the problem of how to resolve a
        identity to an address. A possible solution here is to to use
        servicegroups for both a discovery and resolution mechanism, but
        there are likely to be scalability issues with this approach.
        Another solution might be to leverage the web's caching
        infrastructure for a more scalable resolution mechanism.
        Is this something that the WSDM group is planning to address in
        the future?

/Sam 
	
--
Sam Meder <meder@mcs.anl.gov>
The Globus Alliance - University of Chicago
630-252-1752