Interpreting an incoming message's conversationId

[Dan Weinreb <dlw@exceloncorp.com>]

Suppose I am an MSH, and an incoming ebXML message arrives.  It would
be nice if I could tell, just from looking at the message, whether
this message is part of a conversation that was initiated by myself,
as opposed to part of a conversation that was initiated by my partner.

Why would I care?  Well, that way, I would know whether the
conversationId was made up by me, or by my partner.  If I know that
the conversationId was made up by me, I can parse it into fields and
interpret it in various ways, whereas if it came from the partner it
has to be treated as an opaque string and looked up.  If I could tell
that it was made up by me, I could avoid the cost of the lookup.  In
practice, I think I can arrange for the lookup to be cheap in common
cases, using caching and so forth, but in some cases it's going to
have to be a lookup in a persistent mapping table, which costs
something not entirely trivial.

Is there any way to tell, when a message arrives, whether the
conversationId was generated by "me" as opposed to the partner?  (I
don't think I see any way to do it, but perhaps I'm missing
something.)  Thanks!

-- Dan