Chapter 4 - Networking: Vision and Packet Switching 1959 - 1968
4.2 The Seminal Experiment: 1965
By early 1965, Sutherland had consulted with key ARPA computer scientists and had decided it was time to begin exploring the issues faced in interconnecting computers. In February 1965, Sutherland turned to his friend Roberts, who always had the time and energy to tackle something new, and authorized Roberts to proceed with an experiment to study how to interconnect computers. In July, Roberts contracted with Thomas Marill, founder of the time-sharing company, Computer Corporation of America (CCA), to devise a network experiment. Marill, a former student of Licklider’s, took the project on largely because he too was motivated by Licklider’s Intergalactic Network vision. In October 1965, Roberts and Marill connected a TX-2 computer at MIT Lincoln Lab with a Q-32 computer at System Development Corporation in Santa Monica, CA using a lease-line from Western Union.5 The simple yet path-breaking experiment proved that the computers could run programs and retrieve data on the remote machines, but that the circuit-switched telephone system presented problems. Roberts remembers:
And what we found is that you can connect the computers fine internally and the time-sharing systems both could call on each other, but the communications was slow and unreliable and difficult – and made the whole process so slow it wasn’t very attractive.
That the telephone network proved unattractive as a communication system posed a very significant problem, because other than the totally unsuitable telegraph system, the telephone system was the only nation-wide communication system in existence at the time.
Roberts and Marill documented their work in a paper published in the Proceedings of the AFIPS 1966 Spring Joint Computer Conference. (Lawrence G. Roberts, The ARPANET and Computer Networks, November 1966)