INTERNET: THE EARLY DAYS

By Dave Wallace

RIVER BENDER - June 1998 

Little did I know at the time that Paul Baran, a researcher at the Rand Corporation, would later become known as the founder of the Internet. The year was 1964 and Baran had completed a study titled "On Distributed Communications Networks" which I, an AT&T engineer in Colorado Springs reviewed to determine if any of his ideas might fit into our proposals for air defense communications. Basically, Baran proposed a nuclear survivable network by breaking data into packets and sending them over many paths to be restructured at the destination. It was a novel and fascinating idea but way ahead of the times.

Several years passed and I moved on to Washington where, working on Government communications, I met Dr. Larry Roberts. Roberts had proposed a packet-switched network to the Advanced Research Projects Agency (ARPA) called ARPANET which sounded strangely similar to Baran's earlier proposal. Indeed, Roberts admitted getting many ideas from Baran's earlier paper.

ARPANET, a packet-switched data network, was approved as an experimental network and by the end of 1969, 4 nodes were connected. By 1971, 15 nodes, mostly Universities, with Government research contracts, were connected. It worked. But the real coming out party was in 1972 at the annual International Computers & Communications Convention, where a demo was set up for the public to try out ARPANET. I was there and very skeptical, because as a GE time-shared computer user I knew that an interactive system had to be fast. How could ARPANET be fast enough after sending data in packets all over the place? Sitting at a model 33 Teletype machine, I interacted with a program on the West Coast and was surprised how responsive ARPANET was. It was not as fast as GE, but certainly acceptable. The backbone at that time was 50 kb/s, far slower than today.

ARPANET grew and grew and eventually became worldwide. I recall Dr. Roberts trying to sell ARPANET to AT&T and other carriers since the experiment was completed but there were no takers. Other packet networks started appearing, but the name "Internet" did not come into play until Vinton Cerf and Bob Kahn developed a protocol called TCP/IP that permitted different networks to be interconnected. The Internet was thus born in 1983 when TCP/IP was started. Some say that Cerf and Kahn were the real fathers' of the Internet.

While the Internet grew rapidly, the general public was not permitted access until the National Science Foundation got into the act by funding the high-speed backbone, regional networks and other organizations responsible for maintaining a degree of order. By 1989, 100,000 host computers were on the network and the following year ARPANET ceased to exist.

What exactly is the Internet? The very name suggests an interconnection of networks and indeed there are well over 150,000 networks interconnected worldwide. By the end of '97 there were 30,000 host computers and an estimated 300 million people using the Internet. And still nobody and no government owns the Internet.

For an interesting timeline of Internet events I suggest the following URL:

http://info.isoc.org/guest/zakon/Internet/History/HIT.html

Next month, we'll talk a bit about "Spamming" on the Internet.