Back to top

Chapter 6 - Networking: Arpanet 1969-1972

6.0 Overview

In 1969, the effort to bring into being the Arpanet swung into full gear. BBN had won the right to build the IMPs that would be interconnected using leased lines from AT&T to create the communications network, or subnet. Now the many design issues left open had to be decided. Most importantly would be how packets would be routed through the subnet.

Personnel from the Host sites had been meeting since mid-1968 to craft the host-to-host protocols that would function over the subnet yet, organizationally, they remained unsure of their role or how much authority they had to make decisions. No wonder, most were graduate students inexperienced with managing such a complex, unknown process. In February, BBN convened a meeting to discuss how this host software would communicate with the IMPs. Everyone left realizing just how much there was to do, and that so many questions remained unanswered. The meetings of the representatives soon assumed the formal name of the Networking Working Group.

BBN delivered the first IMP to UCLA on schedule at the beginning of September. That same month, Robert Taylor left IPTO. Larry Roberts became the new Director, making Arpanet but one of his many projects, rather than his principal responsibility. Importantly, his expanded budgetary clout would help him persuade Host site personnel to make the network a priority - a clout very much needed for a network was not everyone’s vision of the future.

Roberts had other concerns. To make sense of how to expand the initial four-nodes into a cross-county network, he sought help from both new experts such as Howard Frank, and trusted friends such as Leonard Kleinrock. After months of testing, expansion began and pleasant surprises confirmed the value of having computers interconnected into a network.

Arpanet was not the only computer network experiment underway during the late 1960’s. Two others that would prove important were the NPL network under the leadership of Donald Davies in London and the ALOHAnet led by Norm Abramson in Honolulu. Both would have lasting impact on the history of computer communications.

In mid-1971, Robert Kahn counseled a frustrated Roberts that a public demonstration needed to be staged to force Host site conversion to Arpanet. In October 1972, at the International Conference on Computer Communications (ICCC), a hugely successful coming out party for Arpanet was held. The who’s who of computer communication attended and interacted, with many soon moving to new organizations and assuming new roles to lead the continued evolution of computer communications. It was a magical and exciting time, when a powerful new idea - packet switching - proved real and a torrent of activity and thinking was unleashed to reshape the way computers would communicate.