Protocol Confusion: Networking, 1972–1979

mous amount of work to fulfill the vision of the ARPANET’s creators. A functional network presented a conundrum: DARPA did not have charter authority to operate a network. With high demand for networking, and the growth of the network­ ing equipment companies documented in Chapter 4, there was a clear need for the ARPANET, or something like it, to support commercial needs. Efforts to com­ mercialize ARPANET technologies were not immediately successful, and networks operated by private companies met some but not all market needs. Elsewhere, research into packet-switching continued along new paths, pro­ pelled by committed teams of researchers in Europe and the United States. In France, Louis Pouzin led a team of researchers in a project called CYCLADES, with the goal of learning from and improving upon the experience of the ARPANET. And in the United States, Robert Kahn worked with other DARPA-funded researchers to devise methods for passing packets from the land-based ARPANET to the radiobased ALOHAnet. They soon realized that they would need a new protocol for transmitting packets, which led Kahn and his collaborator Vint Cerf to develop the Transmission Control Program (TCP) in 1973. By the mid-1970s, a proliferation of projects and protocols came from Ameri­ can universities, private companies like Xerox, and European research institutes— all seeking to provide the foundations for network interconnection and support robust commercial and scientific applications. Scientists and engineers working in institutions devoted to collaboration—including the International Organiza­ tion for Standardization and the US National Bureau of Standards—sought to use Protocol Confusion: Networking, 1972–1979