[ih] Ethernet, was Why TCP?

[Paul Vixie]

John Levine wrote:
> It is my recollection that at the time a lot of people thought that
> Ethernet sounded too good to be true.  If it were heavily loaded, all
> those collisions would surely cause a storm of interference and
> performance collapse, unlike a token ring that shared the capacity in
> a predictable way.

david boggs had, posted on his office window at DECWRL, a photocopy of 
an internal D|I|G|I|T|A|L memorandum from the head of the DEC Office 
Connect project, who was looking for a way to connect office computers 
that wasn't twisted-pair. boggs used a yellow "highlighter" to 
illuminate one key passage which i shall never forget:

"Ethernet is a laboratory toy. Its random exponential backoff mechanism 
is clear evidence that <blah blah blah>."

somehow DEC ended up adopting ethernet, and hiring boggs, in spite of 
this memo. later when DECWRL visited the main "NAC" (Networks and 
Communications) office building to give a series of technical talks, 
someone came into a conference room where boggs was sitting and told him 
they had the room reserved. boggs should not have had to leave. it was 
the Metcalfe and Boggs conference room, after all.

later on, DEC poured billions (that's like millions except with a "b") 
of dollars into DECnet Phase V, proving after all that they, and not the 
community, knew what the future of networking was going to look like. 
(they should have saved the money so that Compaq could have it.)

anyway the list of things that can't be done, and the naysayers who have 
proved those negatives, continues to get longer every few minutes.

-- 
P Vixie