To Top Page

Computer Networking

The Internet, like all early software development, was created in open-source mode, with no patents taken and detailed documentation being freely shared among participants, which enabled it to progress rapidly through seven main inventions so far. Only one person in the world has contributed to more than three of these inventions: I accidentally helped six of them.

     Unfortunately, we messed up in several ways, so more work is needed. I propose developing several more inventions to make things work better, as discussed further on.


Seven Inventions

Computer Networking began with two new kinds of services, that were initiated at MIT in the 1950s as part of the SAGE air defense system funded by the U.S. Air Defense Command. It had 23 air defense zones spread across North America, including one in Canada, and ran for 25 years even though it was a fraud from the beginning. Its two main inventions were:

1. Interactive Computing, which enabled around a hundred users at each site to promptly see the results of their actions, unlike the then-used batch processing, and

2. Packetized Data Communications between computers, enabling prompt exchange of information.

      However, those two terms did not come into general use until much later. SAGE used large-scale vacuum-tube computers that each occupied the area of a football field and used two additional floors of the same size for power supplies, cooling, and communications equipment, plus another floor for operators’ display stations.

     It was initiated by MIT faculty but turned out to be an operational fraud, so the MIT administration backed out, turning it over to two profitable nonprofit corporations (SDC and MITRE) to assist the Defense Department and Congress in stealing trillions of dollars from taxpayers and giving it to crooked corporations such as IBM, AT&T, GE, Boeing, Lockheed, Convair, etc., making it the biggest fraud of the 20th century. However, that information was classified “SECRET” so that if anyone involved talked about it they would go to jail, a practice that is still widely used for modern Military, Industrial, Congressional Conspiracies (MICC). See SAGE like Forrest Gump.


MIT also played a central role in creating

     3. General-Purpose Timesharing, which was inspired by SAGE. A number of us at MIT wrote proposals for how to do it. The person who got there first was a new Assistant Professor John McCarthy, who devised a scheme using program interrupts. He called it timesharing and inspired a number of MIT groups to build prototypes, including Project Mac, which was the most popular of the eight or so timesharing systems developed at MIT and went on to largely replace batch processing, initially using large computers supporting many users, which caused them to slow down under heavy loads. Advancing technology enabled the use of personal computers, which appealed to many people but provided no new capabilities.


4. General-Purpose Networking first appeared in Arpanet, the specifications for which came out of a committee that was dominated by former MIT graduate students, including me, and was funded by ARPA (Defense Department’s Advanced Research Projects Agency) under the leadership of J.C.R. “Lick” Licklider from MIT, who was an old friend of mine, then carried on by Ivan Sutherland, another friend, who had recently earned his PhD at MIT. It was developed under contract by BBN, an MIT spin-off corporation.

     Ivan then tried to recruit me to his ARPA group but having worked for a decade as a Defense Department pirate (see “Piracy Tax,” below), I said that my goal was to get as far from the Pentagon as possible. Ivan then kindly invited me to go to Stanford University to set up a new computer research lab, which I named SAIL (Stanford Artificial Intelligence Laboratory) because John McCarthy would be the Principal Investigator and had introduced the AI term, which I came to hate because it too was fraudulent, but that is another story.

     I thus got to create and manage SAIL, a very enjoyable and productive but challenging experience, while Ivan recruited Bob Taylor to take over his office as he became a Harvard Professor.


5. Internet Protocols were developed by a Stanford research group headed by Vint Cerf and enabled computer networking to utilize diverse networking elements, including satellite links. I helped Vint round up funding using our SAIL contract with ARPA.


6. World-Wide-Web standardized graphical interfaces and was initiated at CERN by Tim Berners-Lee, who now maintains it at MIT. His formatting language, HTML, used many of the same elements as the PUB documentation language that had been developed years earlier by Larry Tesler and me.

     Thus, MIT has played a significant part in developing all networking technologies and I have had at least a finger in on all of it except for core memory, which has since been superseded. These developments were all done in open-source mode, enabling rapid progress. I expect more such developments will soon improve security and enable a more versatile and useful Internet.