[ih] Peter Salus / Baran's work

[Noel Chiappa]

    > From: Bill Ricker <bill.n1vux at gmail.com>

    > (D)ARPAnet initially had fixed routing, not useful in damage-prone
    > environment.

I don't think that's correct. The BBN Proposal (IMP P69-IST-5, 6 Sept 1968)
says (III-47) "We plan to provide an [alternative routing] algorithm which is
adaptive, free from routing loops". The first ARPANET paper (the 1970 SJCC
paper by Heart, Kahn, et al, "The Interface Message Processor the ARPA
Computer Network") says (pg. 555): "The routing table in consistently and
dynamically updated to adjust for hanging conditions in the network. The
system is adaptive to the ups and downs of lines, IMPs ..".

Everyone: please check before making these kinds of statements. This is how
these errors get started - and then it can take a lot of work to straigthen
them out.

    > It was TCP/IP that introduced adaptive routing around damage.

Ironically, some parts of the early Internet used static routing - I
certainly recall it being used in some of the MIT routers.

(Although I _think_ all the BBN routers used dynamic from the earliest -
various versions of GGP, to start with. Note that BBN at one point was using
an un-documented version of GGP, and when I sent a BBN router a GGP packet
formatted according to the spec, it promptly crashed... :-)

	  Noel