September 22, 2023



Chapter 1, (Where Wizards Stay Up Late - Hafner and Lyon) introduces Bob Taylor and ARPA. His decision to construct a network. Origins of the Advanced Research Projects Agency (ARPA):
JCR Licklider Two-year tenure as Director of the Information Processing Techniques Office. Succeeded by Bob Taylor.

Interactive computing and time sharing

Licklider's views about computing - shared by Taylor. Licklider came to computing when most people thought computers were "vastly-expensive and cumbersome calculating machines". Individuals had no access to computers. Computers operated only in batch mode. Licklider (ex MIT) was one of the first to see that computers were machines with which humans could interact.

The origins of packet switching

Three elements to this in Wizards:

  1. ARPA, 1957 - 1967. Developments at ARPA as the Agency moved to thinking about the design and construction of the network.
  2. Paul Baran and his work on resilient networks at the RAND Corporation between 1960 and 1965. He invented the basic technology needed for the ARPANET
  3. Donald Watts Davies, his work on the independent invention of packet-switching at the NPL (National Physical Laboratory) in Britain in 1965.


Important technical ideas are introduced e.g. the distinction between analogue and digital signals, circuit-switching and packet-switching. The book highlights the resistance of AT&T to digital communications and discusses circuit-switching versus packet-switching.


Despite numerous technical developments, the public telephone network - Public Switched Telephone Network (PSTN), although reliable, was inefficient and expensive.

A phone network connects the wires of two telephones (or fax, modems). Sounds from one end are transmitted to the other. This is "circuit-switched", or "switched", network architecture.

When a connection is made at the beginning of a call, part of the network is reserved exclusively for that conversation, or data exchange, whether or not there is traffic.

T171 states that:

"If one party puts down the phone or is silent, or neither computer is sending or receiving data for a period of time (as is the case when using the Internet), that circuit as well as the 'ports' on the phone switches between the two devices are still unavailable for other activity even though they are not being used at the moment. Estimated that up to 50% of a typical voice conversation is silence. (Company must build double the network it really needs for a given number of simultaneous calls at double the cost.) "


T171 states that:

"Instead of keeping a connection open for the entire length of the call, packet networks break the digital stream of ones and zeros into chunks of the same length. These chunks, or 'packets', are then put in the computer equivalent of an envelope, with some information such as the origin and destination, or 'addresses', of the packet, and a serial number that indicates the sequence number of the packet - its 'place in line'. In the place of switches which merely connect and disconnect circuits, packet networks use routers - computers that read the address of a packet and pass it to another router closer to the destination. At the destination, a few thousandths of a second later, the packets are received, reassembled in the correct order, and converted back into the original message. The routers in a packet-switched network are permanently connected via high-speed lines. This may seem expensive at first sight, but it makes sense economically (and technically) if the network is heavily used, i.e. effectively flooded with packets."

AT&T was extremely reluctanct to accept Baran's ideas about message switching. T171 argues that a number of factors will determine whether or not a technological breakthrough is to be accepted. Apart from manufacturing and technology issues, economic, social and legal factors of the time should be considered. The cultural and intellectual "mindset" will also bear upon decisons taken by companies. Thomas Kuhn describes the requirement to accept technological progress as a paradigm shift.

T171 states that:

"Baran's ideas required not one but two paradigm shifts. A move from a paradigm based on analog communications to one based on digital principles. A move from the circuit-switching technology which dominated analog telephony to the packet-switching technology envisaged by Baran and Davies."

Wizards Chapter 3, introduces a number of key elements in the story of the construction of ARPANET.

  • Bolt Beranek and Newman (BBN), the company that was awarded the contract to design and build ARPANET.
  • Key figures within BBN who worked on the project, notably: Frank Heart, Robert Kahn, Dave Walden, Severo Ornstein and Will Crowther.
  • The choice of the Honeywell 516 minicomputer as the basis for the Interface Message Processors (IMPs).
  • The writing of the computer programs to implement packet-switching on the IMPs.
  • The writing of BBN's bid for the ARPANET project and the award of the contract.


Next page » ARPANET and Systems thinking

Previous page « What is the Internet?



















Up to top of page