Re: T2 PLEAE READ - Suggested solution to RM Issues

[Dan Weinreb <dlw@exceloncorp.com>]

Chris, I'd like to understand better your idea of how decision is made
to use an IM, and which IM, and how the "routing" is done.

When the From application invokes the From MSH (when the software
layer above ebXML calls into the ebXML stuff), clearly it has to state
who the ultimate destination of the message is, i.e. the To MSH.  And
normally there would be a CPA, reflecting an agreement between the
From and To MSH's, to tell the From MSH useful things about how the
communication should go.  So the From application tells the From MSH
what CPA to use (or some other data structure or object that acts as a
surrogate CPA, whatever).

Now, how would the From MSH know to actually communicate with some IM
instead of with the To MSH directly?  (1) Because the CPA specifies a
URL or email-address that isn't actually the To MSH's physical
address, though the From side has no idea that this is going on.  (2)
Because the CPA says, explicitly, that the From MSH ought to use a
particular intermediary.  (3) Because the From application, when it
asked the From MSH to send the message, explicitly said "and
furthermore, send it directly to this IM rather than to the To MSH,
and depend on the IM to get the message through".  (4) Other?

In a chain with two IM's, how does the first IM know that that rather
than delivering the message to the To MSH. it should send the message
to the second IM?

-- Dan