[ih] Separation of TCP and IP

[Stephen Casner]

I've just had a chance to read through the TCP-IP split discussion.
Here are a few snippets responding to points relevant to me.

1. Vint said his search for The Oceanview Tales as an ISI/RR report
was unsuccessful and asked if anyone had a copy.  The (or one) reason
the search failed is that it was not published as an ISI/RR.  I have a
pristine copy that includes yellow cover and end pages like an ISI/RR,
but it is spiral-bound.  "Printed by ISI as a courtesy to one of its
research efforts."  Copyright 1979.

Noel also has a copy and listed the chapters therein.  Some are in
Prof. Finnegan's book, but not all.  Noel offered to scan but said his
scanner is broken; mine is working, so I could prepare a scan if
someone wants it.

2. Toerless said: "Starting maybe with Steve Casners
unwillingness?/inability?  (don't remember which one it was ;-) to
hack SunOS kernel sources when implementing RTP and renewed ever
since, not QUIC to the latest."

I honestly don't remember this debate.  But that was around the time
when we were still developing RTP itself, and I imagine working in
userland was much easier.

3. Craig Partridge mentioned seeing a video showing Danny running
around Marina del Rey, using packet voice.  If that video was the
Digital Voice Conferencing movie we made at ISI in 1978, it did show
Danny running in Marina del Rey, but he was running to a phone booth
so he could join the conference from the payphone through the switched
telephone network interface we had built and considered a big deal
because it greatly expanded the access to the packet voice system.
There was nothing mobile in those days (packet radio was still in
progress?).

Craig also mentioned the multimedia conferencing over the Wideband
Network that Claudio Topolcic and I managed.  It use the ST-2 protocol
that was widely denigrated by most of you because it was connection-
oriented, but that was needed for 1-hop transit on the WB net.

4. Jack Haverty wrote about access to uncontrolled packets on the
ARPANET (type 0, subtype 3) being very restricted.  We wanted that
service for packet voice experiments.  Initially the access was very
carefully controlled and monitored, but later in the program BBN was
willing to enable it for our communication sessions without much
concern.  We had demonstrated that it did not kill the network.

                                                        -- Steve