|< Day Day Up >|| |
John Quarterman described the Internet as the product of a 'chaotic mÈnage ‡ trois of government, academia, and business.' [1 ]The starting point of this relationship was a very orderly research project, the ARPANET, funded by the Information Processing Techniques Office (IPTO) of the U.S. Defense Department's Advanced Projects Research Administration (ARPA).
The ARPANET was an experimental backbone of leased lines connecting research scientists in university, military, and industry sites. Its purpose was to facilitate time sharing on mainframe computers. A request for proposals circulated in 1968 called for the construction of a packet-switching device called an interface message processor, the development of software, and the design of a physical network to connect them. A Cambridge, Massachusetts, based research firm with longstanding personal and financial ties to ARPA, Bolt, Beranek, and Newman (BBN), won the ARPANET contract in 1969 (Hughes 1998, 269-270).
The ARPANET was not the Internet. The Transport Control Protocol/ Internet Protocols (TCP/IP) had not been invented yet, and the word Internet was not used to describe it. The ARPANET was difficult to use and connected at most about 200 people at 21 nodes. The project did, however, bring together the people who played a continuous role in the Internet's technical development and its governance for the next 30 years. ARPANET created the nucleus of an Internet technical community.
Robert Kahn was one of the leaders of BBN's interface message processor project. He later co-authored the basic TCP architecture and helped to form the Corporation for National Research Initiatives, which supported the IETF in its early day. The site of the first ARPANET node, installed in September 1969, was the ARPA-supported computer science research center at the University of California, Los Angeles, headed by Leonard Kleinrock, an inventor of queuing theory. At UCLA, Kleinrock's graduate students Steve Crocker, Vinton Cerf, and Jon Postel were given most of the responsibility for implementing the ARPANET protocols. [2 ]It was Crocker who formed the Network Working Group and initiated the Internet's uniquely open method of developing and documenting standards, the Request for Comments (RFC) series. [3 ]Jon Postel eventually took over the task of editing the RFCs and gravitated to the administration of unique number assignments for ports and protocols. Cerf went on to become one of the principal designers of TCP/IP, the most persistent and effective advocate of the TCP/IP standard for internetworking, the founder of the Internet Society, and board chair of ICANN.
It was also the ARPANET project that brought Keith Uncapher, a RAND Corporation computer engineer, into contact with the military research agency. In 1972, Uncapher formed the Information Sciences Institute (ISI) as an affiliate of the University of Southern California (USC) but located it in Marina del Rey, apart from the main campus. ISI was conceived as a kind of West Coast BBN-a university-based research center focused on applications of computer science. Unlike BBN, however, it was nonprofit, and deliberately set up to obtain funding from ARPA exclusively. 'I was totally captivated by the freedom that ARPA had, the excellence of the people, and their ability to commit to a good idea based on the back of an envelope drawing or a telephone conversation,' Uncapher later said. [4 ]ISI became one of the main centers of Internet research and administration, supporting the work of Jon Postel, Robert Braden, Steve Crocker, Danny Cohen, Daniel Lynch, Paul Mockapetris, and others. Throughout the 1970s, many of the ARPANET principals moved seamlessly between the ARPA and ISI.
[1 ]Matrix News, 9 (December 1999).
[2 ]Robert Braden, later a major figure in the Internet Activities Board, worked at the UCLA campus computing center at the time and was active in the Network Working Group. Another contemporary UCLA graduate computer science student, Kilnom Chon, returned to his native Korea in 1981 to establish the first TCP/IP network in Asia and later became one of the leaders of the Asia-Pacific toplevel domain administrators.
[3 ]A history of the Network Working Group and the origins of the RFC series is contained in RFC 1000.
[4 ]An interview with Keith Uncapher, OH174, conducted by Arthur Norberg on July 10, 1989, Los Angeles, California. Charles Babbage Institute, Center for the History of Information Processing, University of Minnesota, Minneapolis.
|< Day Day Up >|| |