Bob Bright (getfiddle@home.com) asks: 
"Is Al Gore the Father of the Internet?
Is he taking credit for a little or a lot too much of the Internet?"

Al Gore has been one of my heroes for the last decade. I became 
aware of him around 1990 when he started being quoted a lot by the
engineering types working on internetworking issues: He was the 
first legislator who actually appreciated what the Internet was all
about, and he helped guide the 'net through a very tricky transition.

When the 'net got started in the 1970's, every computer scientist who
heard about it was jazzed, but only a very select clique could get
to touch it: The hardware for the internet was these special
computers called IMPs (I think that was short for Intelligent
Message Processors) built by Honeywell, and outfitted with software
and some minor hardware modifications by Bolt Beranek and Newman,
and engineering company in Cambridge, Massachussetts. In order to
get one of those, you had to be a research institution with contract
funded research for the Advanced Research Projects Agency of the
US Department of Defense. I think the rental for an IMP was something
like $100,000 per year, which had to be paid out of the overhead on
the research contracts, so small colleges need not apply!

Around 1980-82, the ARPAnet had grown to include major military
posts, defense contracting companies and most universities that
had any defense research contracts at all. It was now carrying
several different classes of traffic:
- administrative traffic for the military
- administrative traffic between the military and its contractors
- and acting as a testbed for research experiments in protocol
development.
During this period, TCP was developed, and the network switched from
the original NCP protocol to TCP/IP. Shortly after that, the network
had grown so large that it had run out of numbers for the IMPs
(the hardware allowed 8 bits for the IMP number) and it was split
into two separate networks connected by some routers called
"mail bridges":
- network number 10 - ARPAnet
- network number 26 - MILnet

This split also helped calm the fears of some military people who
were worried about sharing a network with potentially subversive
students. This fear is why the connection between the networks
was called "mail bridges" implying that only the relatively safe
e-mail could get across. Despite the name, however, those were
really full-fledged routers, providing a completely seamless 
connection.

With IP installed, and the newly invented ethernet allowing for
affordable campus networks, the major universities started attaching
campus networks to the ARPAnet backbone, using VAX-11/780 mini-
computers with the network-aware version of UNIX that ARPA had
paid University of California at Berkeley to develop.

Many of the smaller universities wanted to participate, but did
not have any military reaserch contracts to qualify them, so they
banded together to build a compatible network running TCP/IP over
X.25 (Telenet, Tymnet). This was known as CS-NET (for Computer
Science network).

By 1989, the university-to-university traffic had dwarfed the
military traffic, and the DoD wanted to divest itself of the
overheads of running the network, so they asked the National
Science Foundation to take over. Around this time, the NSF had
started a program to build - I think it was 9 - national 
supercomputer centers, and needed to link them with the potential
users at universities. They rented a bunch of 56 kbps lines -
of the same kind that ARPAnet ran on - and installed a bunch of
routers built out of inexpensive PDP-11/23 minicomputers, using
a software package called FUZZBALL, developed by professor Dave
Mills of University of Delaware. This created a second backbone,
parallel to the DoD-sponsored ARPA backbone. Since NSFnet had
no military funding, there was no longer a requirement for
military contracts to be connected, but since it was paid for
by tax dolllars earmarked for reasearch in the national interest,
it was not available to businesses, except in support of govern-
ment paid research.

It was at this point that Senator Gore stepped in, and basically
brokered a deal where NSF stopped paying for the network, and
instead gave the universities money to buy network services.
This made it possible to start network companies to compete with
NSFnet and its regional affiliates. Several of the NSF-funded
affiliates re-invented tehmselves overnight into for-profit
ventures. NYSERnet became PSI, for example.

Without this visionary plan, there would not have been a commercial
Internet. Because I had seen how elegantly Senator Gore pulled off
this very good thing, I was happy to see him run for president,
and even happier to see him join forces with Bill Clinton. I still
think Al Gore is the better man.