[ih] Ethernet, was Why TCP?

[Brian E Carpenter]

On 01/09/2016 14:09, John Levine wrote:
>> So if one wanted a cheaper mechanism, tossing out the extreme discipline 
>> made sense to me.  We always hear about computing/storage tradeoffs. 
>> This was a cost/channel-efficiency tradeoff.
> 
> 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.
> 
> As we all know, Ethernets worked just fine.  A lot of people didn't
> believe it until they saw it, and sometimes not even then.

There was a paper around 1984 from either UCB or LBL that showed (by
simulations, I think) that the performance curves were the same shape for
Ethernet or (IBM) Token Ring. For Ethernet the drop-off was caused by
collisions on the wire, for Token Ring by I/O buffer overflows. At that
time memory wasn't dirt cheap, so buffer sizes tended to be small.

At CERN we were bombarded by DEC with reasons why we should choose Ethernet
and by IBM with reasons why we should use Token Ring. Being CERN, we chose
both (Ethernet for general purpose use, and Token Ring for the control
system of LEP, the machine that preceded the LHC).

When I spent a couple of years in 2001-3 at the IBM Zurich Lab, where
the IBM Token Ring was conceived, the wounds were still raw. Many IBM
sites were still on Token Ring then. I had to travel with PCMCIA cards
for both Ethernet and TR, plus an adaptor for sites that ran Ethernet
over the IBM Cabling System instead of RJ45/UTP5.

I think Token Ring really failed because the IBM Cabling System was
too good, and therefore very expensive. The Ethernet discourse (one
simple coax cable goes everywhere) was very persuasive, especially
when Cheapernet (thin coax) came along.

Regards
    Brian