Editor's Note: Minutes Received 11/4/92 INTERIM_MEETING_REPORT_ Reported by Mark Knopper/Merit Minutes of the TCP/UDP over CLNP-addressed Networks Working Group (TUBA) This session took place on October 28, 1992 in conjunction with INTEROP Agenda o Introductions o Transition Plan and End States o Routing and Addressing Plan o Protocol Changes: - TCP/UDP pseudoheader and checksum - SMTP/Telnet/FTP - DNS o International Standard Profile for TUBA/CLNP o Writing Assignments The Agenda was loosely agreed to and followed. Bill Manning started off by asking ``What constitutes a CLNP address for TUBA?'' Possible answers included use oft he ISOC-AFI (as proposed by Juha Heinanen), or an end-system id that is globally unique and is routable by the IS-IS protocol. The consensus was that the latter is preferable, and that a 6 octet system id plus selector for the ``IP protocol field'' is a TUBA address. It was agreed that any arbitrary AFI should be usable as long as global uniqueness of the system id is preserved. The basic outline of the transition plan is that IP will become vestigial as an increasing number of hosts start to use TUBA. A discussion of the routing and addressing plan for CLNP in the internet then ensued. Dick desJardins pointed out that RFC 1237 on OSI NSAP allocation by Richard, Eva and Ross should be followed. It was agreed that current practice in the CLNP world should be followed with respect to routing. It was agreed that Sue Hares and Cathy Wittbrodt would provide documentation on the CLNP routing schema as a contribution from the NOOP Working Group to the TUBA Group. It was agreed that it would be useful for the transition plan to have a standardized value for the first 2 octets of the system id to indicate that the next 4 octets are an IP address. It was suggested that we request such an encoding from the IEEE. Peter Ford was charged with this task. Denise Heagerty raised the question of whether TUBA is going to support CLNP-only hosts, i.e., TCP/UDP over CLNP with no transition plan for addressing. These hosts would not have been using IP addresses and 1 therefore would not conform to the 2-octet system id prefix above. There was no immediate answer to this but it was agreed this was important to work out, since the unique address needs to be provided for higher level protocols in TUBA. Notable quote from Dave Katz: ``That's somewhat orthogonal but obviously related''. Dave Piscitello has an Internet Draft describing the CLNP profile for TUBA. This could be submitted to ISO as an International Standard Profile for CLNP, but the Group was not sure this would be useful. The consensus was that Dave should decide, and Mark Knopper would follow up with him. There are 8 documents completed or in progress for TUBA: 1. DNS extensions, revision of RFC 1348 by Bill Manning and Richard Colella (in progress). 2. Routing Plan from NOOP - Cathy Wittbrodt/Sue Hares (to be supplied). 3. TUBA, a simple proposal for internet addressing and routing by Ross Callon, RFC 1347. 4. Addressing and end-point identification, for use with TUBA, Internet Draft by Ross Callon. 5. IEEE encoding for TUBA NSAPs, by Peter Ford (to be supplied). 6. CLNP Profile for TUBA, Internet Draft by Dave Piscitello. 7. IESG deliberations packet, to be generated by Peter Ford and Mark Knopper. (Brian Carpenter has provided a significant contribution on the mailing list.) 8. TUNE, Internet Draft by John Curran. Attendees George Chang gkc@ctt.bellcore.com Richard Colella colella@osi.ncsl.nist.gov John Curran jcurran@bbn.com Richard desJardins desjardi@boa.gsfc.nasa.gov Dino Farinacci dino@cisco.com Peter Ford peter@lanl.gov Susan Hares skh@merit.edu Kenneth Hays hays@scri.fsu.edu Denise Heagerty denise@dxcoms.cern.ch Dave Katz dkatz@cisco.com 2 Mark Knopper mak@merit.edu Randolph Langley langley@scri.fsu.edu Bill Manning bmanning@sesqui.net Vesa Parkkari vesa@relevantum.fi Yakov Rekhter yakov@watson.ibm.com Keith Sklower sklower@cs.berkeley.edu Cathy Wittbrodt cjw@nersc.gov 3