minutes of 5-Apr-2001 TR&P con-call

[christopher ferris <chris.ferris@east.sun.com>]

Attending:

Colleen Evans
David Burdett
Bob Miller
Gary Morin
Ralph Berwanger
Chris Ferris
Doug Potter
David Fischer
Dale Moberg
Ingrid
Ian Jones
John Ibbotson

Minutes:

Discussion of adding a new element DeliveryReceipt as distinct
from Acknowledgment.

Bob Miller raised issue that things like Acknowledgment really belong
in the SOAP:Header. Based on discussion surrounding synchronous
messaging and the combination of Acknowledgment and DeliveryReceipt
it may be the case that Acknowledgment does indeed belong in the 
SOAP:Header with an actor attribute of "next" so that the chain
of intermediaries can push and pop their respective acknowledgments
along a chained synchronous request/response.

The BP spec which went public yesterday also has a ReceiptAcknowledgment
and AcceptanceAcknowledgment defined as business signals. Chris has
raised an issue with the BP team that we (TR&P and BP) need to 
come to resolution so that there is a consistent approach that
works for both.

Bob also raised the issue of having the Signature element as a 
direct descendent of SOAP:Header removes it from the "ebXML"
processing. Much discussion ensued. In the end, the real issue
was the figure on page 13 which is intended to be non-normative, yet comes
across as otherwise. Ralph took an AI to amend the diagram to remove
the interrelationship between the various components.

Ralph, as editing team lead suggested that we close all open
issues by Wednesday of next week, so that the editing team can
finish the document for submission to QR by Friday.

Prasad raised a couple of issues related to RNIF mapping
onto ebXML Message Service specification. The firsst issue
was the lack of a place to put the digest of a received (signed)
messsage in the Acknowledgment element. This issue is somewhat
related to the issue above w/r/t DeliveryReceipt and ReceiptAcknowledgment
especially as the ReceiptAcknowledgment element is straight from
RNIF.

The second issue raised was the issue w/r/t header elements/attributes
that derive from CPA which might be in conflict with CPA. It was noted
that the resolution is not to define whether or not the CPA can or
cannot be overridden, but that the behavior of a receipient of a message
that is in conflict with the CPA is free to do with that message
as it sees fit. If it chooses to reject the message as inconsistent,
then we have defined how the error will be produced. We have already
concluded that we can not come to consensus agreement one way or the
other.

Prasad will follow up with other issues via the list.

We agreed that we'll need two meetings next week to resolve open
issues by Wednesday. This is compounded by the fact that many of
us will be attending XML DevCon next week. We'll decide on a time
for the calls. AIAG offered to host Monday's call.