InfoMagic Standards 1994 January/InfoMagic Standards - January 1994.iso/inet/ietf/sip/92nov.min

Text File | 1993-02-17 | 7KB | 150 lines

Editor's Note: Minutes received 11/20/92

CURRENT_MEETING_REPORT_


Reported by Christian Huitema/INRIA

Minutes of the Simple Internet Protocol BOF (SIP)

The Simple Internet Protocol BOF attracted a wide audience.  The first
part of the meeting was a quick review of the proposed SIP Charter,
which was approved by the Group modulo alignment of the milestone dates
with the proposed IESG decision schedule.  The participants were
reminded of the name of the mailing list:  <[e-mail]>
and that preliminary versions of the documents can be obtained by
anonymous ftp from ``parcftp.xerox.com'' in the directories ``pub/sip''
or ``pub/net-research''.  Related documents on IPAE can be obtained from
the same server in the directory ``pub/ip-encaps''.

The discussion turned next to the SIP specifications, addressing a set
of characteristic design points, and in particular some issues that were
marked as provisional in the current specification:


   o Steve Deering presented a problem posed by the difference between
     the TCP pseudo header ``conceptual layout'' and the actual layout
     of the payload length and type fields in the packets, and asked
     whether conceptual and physical layout should be aligned.  It was
     observed that the pseudo header remains constant (modulo the packet
     length) for the duration of the connection, while changing the
     layout would makes the hop count handling in each packet somewhat
     slower.  Moreover, the relation between packet layout and pseudo
     header will have to remain ``conceptual'' when options like source
     routing are used.  It was decided not to change the packet layout,
     but to explain more clearly the pseudo checksum computation rules
     in the documentation.

   o Some Group members questioned the absence of a checksum in the
     network header.  This item had already been debated in the mailing
     list.  The arguments for omitting the checksum will have to be
     presented in detail in a SIP overview document.

   o Some Group members questioned the small size of the payload type
     field, and the need to provide an extension mechanism, e.g., for
     student projects.  Various solutions were proposed, e.g., to
     reserve the value ``255'' for an extension mechanism.  The need for
     a payload type indicating ``intermediate options'' (to be processed
     by all routers) was mentioned in the same discussion.  An example
     of a request for such options may be the need of performing ``trace
     route'' on a multipoint address.  This mechanism will have to be
     documented in the specification.

   o The discussion on ``flow-ids'' showed that there was no consensus
     on this point that many members feel as deserving further research,
     and that the corresponding bits should remain reserved in the

                                   1





     initial specification.  However, the first implementors reported
     that the presence of a TOS field similar to that of IPv4 would help
     the transition process.  This field will have to be added in the
     revised specification.


One of the results of the discussions of the specifications was to
outline the need for an ``overview'' document.  The discussion turned
then to addressing.  Ross Callon objected that the 64 bits SIP addresses
were smaller than the 160 bits NSAPs, so could not so easily be used to
incorporate link layer addressing, e.g., telephone numbers.  The
discussion showed that the Working Group did not believe that the NSAP
size was justified or needed, and that there is virtue in keeping the
addresses compact.  Steve Deering presented then the ``metropolitan''
addressing plan.  One of the result of the discussion was to outline
again the need of more explanations.  The overview or the addressing
documents should explain how mobility, renumbering and policy routing
are supported, based on concrete examples.

Attendees

Cynthia Bagwell          [e-mail]
David Bolen              [e-mail]
Ross Callon              [e-mail]
Ken Carlberg             [e-mail]
Stephen Casner           [e-mail]
Rob Coltun               [e-mail]
Michael Conn             [e-mail]
Chuck Cranor             [e-mail]
David Crocker            [e-mail]
Michael Davis            [e-mail]
Steve Deering            [e-mail]
Barbara Denny            [e-mail]
Kurt Dobbins             [e-mail]
Jon Dreyer               [e-mail]
Ralph Droms              [e-mail]
Donald Eastlake          [e-mail]
Robert Enger             [e-mail]
William Fink             [e-mail]
Karen Frisa              [e-mail]
Shoji Fukutomi           [e-mail]
Robert Gilligan          [e-mail]
Joseph Godsil            [e-mail]
Masayoshi Gohara         [e-mail]
Heather Gray             [e-mail]
William Haggerty         [e-mail]
Joel Halpern             [e-mail]
Robert Hinden            [e-mail]
Don Hoffman              [e-mail]
Christian Huitema        [e-mail]
John Ioannidis           [e-mail]
Ronald Jacoby            [e-mail]
Charley Kline            [e-mail]

                                   2





Tracy Mallory            [e-mail]
Greg Minshall            [e-mail]
Dave Monachello          [e-mail]
Andy Nicholson           [e-mail]
Erik Nordmark            [e-mail]
Joseph Ramus             [e-mail]
Benny Rodrig             [e-mail]
Henry Sanders            [e-mail]
Henning Schulzrinne      [e-mail]
William Simpson          [e-mail]
Frank Solensky           [e-mail]
Tang Tang                [e-mail]
Richard Thomas           [e-mail]
Jim Thompson             [e-mail]
Stuart Vance             [e-mail]
Gregory Vaudreuil        [e-mail]
A. Lee Wade              [e-mail]
Chuck Warlick            [e-mail]
Luanne Waul              [e-mail]
Douglas Williams         [e-mail]
Kirk Williams            [e-mail]
Daniel Wilson            [e-mail]
Robert Woodburn          [e-mail]



                                   3