RE: [i95, i22] - Proposal for clarifying use of SOAPAction

>A. It seems to me that we should establish a rule or guideline that all
>information derivable from the SOAPAction URI also be derivable from the
>contents of the Envelope itself.  This ensures that the SOAPAction field is
>indeed a hint, and that any use of it (as opposed to use of the envelope)
>is indeed (a) just an optimization per point #2, and (b) can indeed be cost
>checked per #3.

Is the XMLP node required to check the consistency between the original 
information within the message header and the SOAPAction?
What happens if they differ?

>can be automatically provided
What would be be benefit of automatically providing the SOAPAction field?
If it's derived from the message's content also the content could be inspected 
directly.

I second on the idea mentioned in this weeks telcon that the SOAPAction should 
only contain information which is also stored in the XMLP message. Especially 
when considering other protocol bindings than HTTP which do not offer such a 
header mechanism this will be the only way to ensure equal treatment of them.

But when the SOAPAction as HTTP header field is required to contain only 
information which is derivable from the message's content, why not directly 
looking into the content ...?

This seems to back the idea of dropping the SOAPAction field at all.

Mario

Received on Friday, 4 May 2001 09:13:05 UTC