David,
Agreed, but even if the "dispatching" was performed from
the Manifest (by some new processor), then there still needs
to be MIME parsing (although possibly not dispatching) to
resolve the attachments.
What this seems to be suggesting is a whole lot more work
and I'm not convinced that it is really necessary or even
that it solves the problem.
I wasn't suggesting that there were any changes to the existing
MIME headers, what I was commenting on was the need (for digest
calculation) to specify canonicalization of MIME headers which
is clearly (IMO) outside our scope, responsibility and ownership.
Cheers,
Chris
David Fischer wrote:
> Chris,
>
> I didn't take this to mean NOT have MIME headers on the payloads but just to
> copy them (Suresh used the word 'Repeat'). This presents two possibilities,
> either the ebXML signature verification must compare the signed MIME headers in
> the Manifest to the ones on the payloads to make sure they have not changed, OR,
> I suppose the ebXML processor could dispatch off of the headers in the Manifest
> (except that some headers might not be there - like CTE headers).
>
> I didn't understand this as any change to the existing payload MIME headers.
>
> Regards,
>
> David Fischer
> Drummond Group.
>
>