From nowicki@Sun.COM  Tue Nov 24 13:02:30 1987
Posted-Date: Tue, 24 Nov 87 11:59:09 PST
Received-Date: Tue, 24 Nov 87 13:02:30 PST
Received: from SUN.COM by venera.isi.edu (5.54/5.51)
	id AA19856; Tue, 24 Nov 87 13:02:30 PST
Received: from snail.sun.com by Sun.COM (4.0/SMI-3.2)
	id AA00303; Tue, 24 Nov 87 12:30:19 PST
Received: from speed.sun.com (speed-ebb) by snail.sun.com (4.0/SMI-3.2)
	id AA11810; Tue, 24 Nov 87 11:55:28 PST
Received: from rose.sun.com by speed.sun.com (4.0/SMI-4.0Beta)
	id AA08138; Tue, 24 Nov 87 11:59:09 PST
Date: Tue, 24 Nov 87 11:59:09 PST
From: nowicki@Sun.COM (Bill Nowicki)
Message-Id: <8711241959.AA08138@speed.sun.com>
To: end2end-interest@venera.isi.edu
Subject: Network Computing Forum meeting
Content-Length: 2468
X-Lines: 57
Status: RO

Here is a short summary of the recent Network Computing Forum meeting:

Attendance: about 250 people this time, about half of which seemed to
be marketing types (grey suits).  DEC sent one "observer", and nobody
from IBM showed up.  Not many Internet people, except for Dan Lynch (of
course) and Lorenzo Aguilar.  In fact, many speakers started their
talks with: "I don't know anything about protocols, but..."

Wide Area Networking Working Group:  evidently this group has been
disbanded. Bill Melohn from Sun was the only person to show up!

User Interface Architecture:  never-ending debate on which X toolkit to
use.

Network Administration:  Dominated by a big Apollo user who is coming
up with a model with entities and attributes to feed into a database
system.

Practical Aspects:actually bogged-down in discussing "human"
(i.e. political) aspects.

Non-Unix Networks: wants to change name to "Mainframe Networks" to
reflect the fact that they are only interested in IBM 370 systems.
Langdon of Amdahl complained that they should call a spade a spade and
rename it to "MVS and VM Networking" since they are not concerned about
IBM 370s that run Unix.

Core Network Services: Greg Chesson is the chair.  So far they are
bogged down in the ASN.1 debate.  Since the Apollo and X people are
claiming that their techniques for saving byte-swapping are necessary,
this group has agreed on a scheme in which everyone can send in ASN.1,
but you can also send your private format, and the receiver will tell
you if it does not want that format.  I came up with some one-line
descriptions of some data-representation standards:

Apollo NDR	Receiver makes it right
NCF proposal	Receiver says it's wrong
X (double ports)It's right half the time
Sun XDR		It's always good enough
ASN.1		It's never quite right, but standard!

Application Partitioning: hopeless problems of trying to use RPC
to distribute already-written programs.

User Requirements: things like standard file system formats that
become uncovered when you have heterogeneous networks.

Software Licensing: vendors say that network licensing will make
cost structures "more flexible", while users think it is a smoke
screen to make software more expensive.

There were also a dozen "papers" that varied quite a bit in quality.
The only one I remember was by Apollo who did a name server very similar
to our Yellow Pages.  Marshall Rose (who has now moved to Wollogong!)
gave his usual ISODE speech.

	-- WIN