Bright Boys: The Making of Information Technology Book

In the late 1940s and early 1950s the United States Air Force was spending too much money on preparing for the first nuclear strike on the Soviet Union, and too little on defending the United States from such a strike. Wargames and false alarms (from civilian airliners that have strayed off course, for example) have proven that, should Stalin's falcons strike the United States in their Tu-4s (reverse-engineered B-29s) carrying atomic bombs, the United States military could not impede them in a significant way, and the country could kiss Seattle and San Francisco goodbye. A decade earlier, the British won the battle over their skies by circling their island with radars, and having them transmit warnings of German bombers to the headquarters, which would dispatch fighters from airfields to engage them. The same had to be replicated on a much larger scale: a network of early warning radars in the Arctic and the Atlantic sending data about Soviet bombers over phone lines to command centers, which would in turn send orders to jet fighters and surface-to-air missiles. The amount of information flowing through the system would overwhelm anything but the new invention, the electronic digital computer. However, all the existing computers were built for offline batch processing: things like computational hydrodynamics of thermonuclear explosions. So the real-time computer needed to be invented, and so did the computer network connecting several such computers, radars and airfields.

A bunch of 20something MIT engineers started doing that. Their design was improved upon a few times, and after all was done, there were 23 installations with 2 vacuum tube computers apiece hardened against a nuclear explosion, one in a bunker under 600 feet of Canadian granite. These computers could fly F-106 fighters on autopilot and launch BOMARC missiles. In the age of ICBMs they were already obsolete, but the technical achievement was huge, and the spinoffs very influential. A transistorized version of the real-time computer evolved into the DEC PDP-1 minicomputer. When American Airlines asked IBM for an airline reservation system, IBM built one based on the air defense system, with travel agents taking the place of radars and airfields; the software is still in operation.

This is a great story, but it deserves a better written book than this one, one which would tell it without straying off into too many irrelevant details.