I think we're getting rather far from the base issue described in issue 15. For
Role in particular, many in this group thought the element MUST be handled by a
receiving MSH but the schema described an element with ordinality of 0 or 1.
This conflicts with the current text using the 2119 OPTIONAL sense. As Dale has
well described, we still have an issue here.
Issue 15 also calls out a general problem in the specification where elements are
described as OPTIONAL but the associated service is really what implementations
may or may not choose to include. Correcting this terminology problem is
probably of lower priority than determining what we really want to say about the
Role element. I propose we really want to say everybody MUST support this
element but it may appear in an instance of an ebXML message.
thanx,
doug
David Fischer wrote:
> This is a perfect example where the RFC2119 definitions are extended within a
> single spec.
>
> Regards,
>
> David.
>
>