Aucbvax.6444
fa.works
utzoo!decvax!ucbvax!works
Tue Mar  9 16:41:11 1982
Re: Ethernet Doomed?
>From guyton@Rand-Unix Mon Mar  8 23:26:11 1982
From: guyton at Rand-Unix
Via:  Mit-Ai; 9 Mar 82 18:56-EDT
Via:  Brl-Bmd; 9 Mar 82 19:01-EDT

Quotes from an article in Mini-Micro World, Feb 82, pg 17-18.
(Also, the same article has some information on a couple high
level protocols recently announced by Xerox.)

" . . .  Davide Liddle, vice president and general manager of the
OPD, defended the single-channel, baseband approach used by
Ethernet against multi-channel, broadband LAN methods.  He also
took issue with the critical Strategic, Inc., report, entitled
"Xerox - The Key Issues."

"Out of about 100 pages," Liddle said, "only about two and one-
half pages deal with Ethernet.  Most of the remainder consists of
historical data, untrue information and gossip.  The Ethernet
material that does exist is incorrect."

[ . . .]

Liddle claims assertions of broadband cost advantages are
"rubbish."  Although many broadband components are the same type
of units used for CATV transmissions, and are, therefore,
produced in quantity at low costs, the broadband transmitters are
relatively expensive and difficult to maintain on a one-per-work-
station basis, Liddle says.  Broadband systems also require a
separate modem and controller for each terminal at each node,
while Ethernet requires only one controller per node.

Broadband also suffers in comparison to baseband during the
planning stage, Liddle says. "You need a galactic plan to
implement a broadband network because you must ensure systgem
balance, since a strong signal can overpower a weak one." Drop
cables to terminals must also be of the same length.