Re: ImmediateSender ambiguity

A related thread on the public WS-Addr discussion list starts here:

http://www.w3.org/mid/20050818152135.80151.qmail@web51606.mail.yahoo.com

Worth a read for the context.

Marc.

On Aug 24, 2005, at 2:33 PM, Mark Baker wrote:

>
> Greetings,
>
> I have a concern about the WS-Addressing specifications that boils  
> down
> to an interpretation of part of the SOAP 1.2 default HTTP binding.
>
> The WS-Addressing WG appears to believe that the Request-URI in a
> SOAP/HTTP message is hop-by-hop, while I maintain it's end-to-end.   
> As a
> result of this belief, WS-Addressing created a wsa:To header whose  
> role
> was to carry the end-to-end address for a SOAP message.  I believe  
> this
> interpretation is inconsistent with the SOAP 1.2 default HTTP binding
> (as well as HTTP itself).
>
> As evidence of this claim, I referred to the following part of the  
> spec
> where the value of the ImmediateDestination property is declared;
>
>  "An identifier (URI) that denotes the responding SOAP node"
>  -- http://www.w3.org/TR/2003/REC-soap12-part2-20030624/#tabreqcon
>
> Although the actual definition of the property seems ambiguous, and
> potentially inconsistent with the above;
>
>  "The identifier of the immediate destination of an outbound message."
>  -- http://www.w3.org/TR/2003/REC-soap12-part2-20030624/ 
> #tabreqresprops
>
> I'd like to request that the WG clarify which interpretation is
> intended.
>
> Thanks!
>
> P.S. I'll be out of email range until Sept 1.
>
> Mark.
> -- 
> Mark Baker.  Ottawa, Ontario, CANADA.          http://www.markbaker.ca
> Coactus; Web-inspired integration strategies   http://www.coactus.com
>
>

---
Marc Hadley <marc.hadley at sun.com>
Business Alliances, CTO Office, Sun Microsystems.