Re: Stateful Web Services...

Yes, I think this is right. I generally find these discussions very
interesting, but I think the community needs concrete implementations and
experiences. We (University of Virginia) have created a WSRF implementation
on .NET that I encourage people to download and play around with
(http://www.ws-rf.net) . We have two papers that people might be interested
in that are relevant to these discussions:

"Exploiting WSRF and WSRF.NET for Remote Job Execution in Grid Environments"
http://www.cs.virginia.edu/~humphrey/papers/WSRFRemoteExecution.pdf

"An Early Evaluation of WSRF and WS-Notification via WSRF.NET"
http://www.cs.virginia.edu/~humphrey/papers/EarlyEvalWSRF.pdf

By the way, we have been working with the Globus team to ensure interop. I
think we (my team's implementation and the Globus team's implementation) are
on the right track - there are definitely new capabilities there. I
certainly think there's a value to standardizing such mechanisms re: state,
but I realize that the best evidence for this claim will be time-tested
deployments, which of course do not exist yet.

Regards,
Marty

Marty Humphrey
Assistant Professor
Department of Computer Science
University of Virginia

> Thanks, Geoff

> But I don't think I am looking for the one true way.  I am only looking 
> for progress.

> I absolutely agree about building on real world experience -- which is 
> the path being followed by GGF and Globus.

> Chalon Mullins
> Technical Director, Infrastructure Strategy and Architecture
> Charles Schwab & Co., Inc.
> SF211MN-08-472
> 101 Montgomery St
> San Francisco, CA 94104
> phone:  (415) 667-1117