Ed -
That's how you compose in security, including non-repudiation
information (see e.g. lines 1245-1247).
The Alliance can put its information in the header; the EI payload
is the Payload in figure 6-1 line 1241.
And yes, it's more general than WSDL; any of the transports we've
discussed (and I've talked to people doing EBXML, CIP, and others)
have the same structure.
The patterns for message exchange are OneWay, so there's no expected
messaging system response unless that's built in - we're building
for highly scalable systems without a lot of message chatter, hence
the application level acknowledgements that combine a typical
messaging system ACK with a real message, saving considerable
interaction.
You have a similar effect with responses on oneway messages when
you're doing REST.
Hope this helps.
Yes, the spec could be more general there.
Thanks!
bill
--
On 10/16/11 6:44 PM, Ed Koch wrote:
Are the message headers described in section 6.6 simply being proposed at the moment or are they actually defined somewhere. They seem to address some of the issues that are being discussed in the Alliance and may have value beyond just the WSDL based interactions that the section implies they are for.