ARPANET: How a Crashed Login in 1969 Gave Birth to the Internet

Got something to say to this podcast? Send a voicemail, and you might hear yourself on the next episode!

On October 29th, 1969, a student programmer at UCLA tried to send the first message across a new network to the Stanford Research Institute, more than 300 miles away. He typed L, then O, and the system crashed before the G of login arrived. This episode traces the ARPANET from that stuttering start back to its real origins: the waste of circuit switching, the packet switching ideas of Paul Baran and Donald Davies, J.C.R. Licklider's memos on an intergalactic computer network, and Bob Taylor's annoyance at needing three separate terminals in his Pentagon office.

The story then moves to the hardware and the people. Universities refused to spend mainframe cycles on networking, so Wesley Clark proposed refrigerator-sized Interface Message Processors, built by Bolt, Beranek and Newman after AT&T dismissed the whole idea. Email soon swallowed the traffic, rule breaking followed, and Bob Kahn and Vint Cerf wrote TCP/IP to join separate networks together. The episode also settles the nuclear survival debate and follows the network to its formal shutdown in 1990.

• The first four nodes went live in late 1969 at UCLA, the Stanford Research Institute, UC Santa Barbara, and the University of Utah.
• Ray Tomlinson invented network email in 1971, and within a year it made up three quarters of all ARPANET traffic.
• In 1978 Gary Thuerk of Digital Equipment Corporation pitched computers to about 400 users at once, the first spam, and claimed $13 million in sales.
• On Flag Day at the start of 1983, the entire network switched from NCP to TCP/IP, turning it into the backbone of a network of networks.
• ARPA director Charles Herzfeld denied the nuclear origin story, yet Baran's packet switching research was explicitly designed to survive a nuclear war.