RE: Reliable messaging vs business retries - the sender is not done once a message is reliably delivered.

["Jean-Jacques Dubray" <jeanjadu@Attachmate.com>]

Steve:

Your scenario is perfectly valid, but why deal with it at the protocol
level rather than at the collaboration level? Is it still the same BTA
instance if the sender is changing the content of the message and
"retries"? (This is a philosophical question so I am not expecting a
clear yes/no answer). You could do exactly the same thing at the
collaboration level by allowing a transition "to self" on that
particular BTA. That would accomplish exactly the same thing but give
you more expressivity about when retries are allowed (timeout, receipt
nack, acceptance nack, any failure, technical failure, ...)