[Fwd: Re: [ws-caf] My email for Greg and my AI [Fwd: Our AI on Anish'sWS-Context comment #18]]

[Greg Pavlik <greg.pavlik@oracle.com>]

-------- Original Message --------
Subject: 	Re: [ws-caf] My email for Greg and my AI [Fwd: Our AI on 
Anish's WS-Context comment #18]
Date: 	Mon, 11 Apr 2005 10:53:31 -0400
From: 	Greg Pavlik <greg.pavlik@oracle.com>
Organization: 	Oracle Corporation
To: 	Doug Bunting <Doug.Bunting@Sun.COM>
References: 	<425A3208.3090603@Sun.com>



I suggest we remove any discussions around what addressing headers and 
MU semantics; let's assume this is covered by the WS-Addressing 
specification.

There is a larger problem with SOAP mus and context. To stimulate 
discussion, see attached strawman.

Greg

Doug Bunting wrote:

> sorry for the delay...
>
> ------------------------------------------------------------------------
>
> Subject:
> Our AI on Anish's WS-Context comment #18
> From:
> Doug Bunting <Doug.Bunting@Sun.COM>
> Date:
> Wed, 16 Feb 2005 13:02:12 -0800
> To:
> Greg Pavlik <greg.pavlik@oracle.com>
>
> To:
> Greg Pavlik <greg.pavlik@oracle.com>
>
>
> Greg,
>
> In an attempt to get to a single approach prior to sending something 
> out to the TC list, I outline two possibilities below. Let me know 
> what you think.
>
> Anish said:
> "
> 18) Page 7, para right below figure 4 says:
> "A service that requires a service reference
> element MUST use the mustUnderstand attribute for the SOAP header
> element within which it is enclosed and MUST return a mustUnderstand
> SOAP fault if the reference element isn’t present and understood."
>
> I don't understand what this means. How does MU header help here?
> "
>
> My comment during the call was that it appeared the first MUST was a 
> client requirement and the second a service requirement.
>
> Someone else asked if these requirements belonged in WS-Context at 
> all. I assume their point was the addressing specification (or, 
> possibly, a referencing specification) would decide when and where 
> mustUnderstand was useful or required.
>
> In context (pun intended), things get a bit more confusing:
> "
> Messages sent to referenced services MUST use the addressing scheme 
> defined by the specification indicated by the value of the 
> reference-scheme element if present. Otherwise, the namespace URI 
> associated with the Web service reference element MUST be used to 
> determine the required addressing scheme. A service that requires a 
> service reference element MUST use the mustUnderstand attribute for 
> the SOAP header element within which it is enclosed and MUST return a 
> mustUnderstand SOAP fault if the reference element isn’t present and 
> understood.
> "
>
> Just translating this into active voice and correcting to match the 
> schema might help:
> "
> If present, the value of the reference-scheme attribute indicates the 
> specification defining the addressing scheme in use for this 
> reference. Otherwise, the namespace URI associated with the immediate 
> descendant of the element of type ServiceRefType provides this 
> indication. A sender MUST use the identified addressing scheme when 
> sending messages to the referenced address.
>
> The reference itself MAY be required in some messages. In this case, 
> message sender MUST use the SOAP mustUnderstand attribute on the 
> header containing the required reference. A service receiving such a 
> message MUST return a mustUnderstand SOAP fault if the reference is 
> not present and understood.
> "
>
> Another alternative may be simply removing this entire paragraph. It 
> seems redundant with material earlier. For example, have a look at the 
> paragraph below figure 2.
>
> thanx,
> doug

mustunderstand.rtf