When Dreams Become Real
For Henry David Thoreau, Walden Pond was never the remote wilderness that many imagine it to be. Along its west shore ran the tracks of the Fitchburg Railroad, on whose right of way Thoreau often walked in the evenings to nearby Concord to chat with his friends. Other than that, he had little use for the railroad. “We do not ride on the railroad; it rides upon us,” he wrote in Walden.1 He compared the crossties, on which the rails were laid, to the Irish laborers who built the line and over whose backs the trains ran. Railroad fever was about to grip the nation, but Thoreau would have none of that, in spite of his almost daily—and free—use of the Fitchburg line’s right of way.
A few miles west of Walden Pond, the railroad built a spur to the falls of the Assabet River in the town of Maynard, where a woolen mill was established to take advantage of the falling water. By the time of the Civil War, the mill was one of the world’s largest and did a healthy business supplying blankets and uniforms to the Union Army. By 1957, New England’s textile industries had largely relocated to the South (eventually to settle in China), and the building sat mostly empty. That year, two engineers, Ken Olsen and Harlan Anderson (ages thirty-one and twenty-seven, respectively) founded a company, the Digital Equipment Corporation (DEC), and moved into a corner of the building.2 Olsen was a graduate of MIT; Anderson had also worked at MIT’s Digital Computer Laboratory. Both had worked on advanced computer engineering at MIT’s Lincoln Laboratory, a nearby Air Force–sponsored lab working on problems of air defense. DEC’s first products were “logic modules”: building blocks that others would insert into designs for a variety of machines. Olsen and Anderson were reluctant to call DEC a “computer company,” since a number of well-established companies had entered the computer field and lost a lot of money trying to compete with IBM, which dominated it. Unlike those companies, DEC prospered by pioneering a new class of “minicomputers”: machines that were relatively small and cheap and thus suitable for laboratories or engineering groups that otherwise could not afford the large and expensive mainframes offered by IBM and its competitors.
The Digital Equipment Corporation’s PDP-6 and its engineers, summer of 1964. Sitting, from left to right: Lydia McKalip, Bill Coburn, Ken Senior, Ken Fitzger- ald, Norman Hurst, and Harris Hyman. Standing: Peter Samson, Leo Gossell, Gordon Bell, Alan Kotok, Russ Doane, Bill Kellicker, Bob Reed, and George Vogelsang. (Source: Jamie Parker Pearson, ed., Digital at Work: Snapshots from the First Thirty-Five Years [Bedford, Mass., 1992], p. 55. Photo © Hewlett-Packard Company, reproduced courtesy of the Computer History Museum, Mountain View, California.)
In 1964, DEC took a bold step and introduced another computer of radical design, the PDP-6. It was not a minicomputer but a large mainframe, and thus it appeared to challenge IBM’s product line. During the summer of 1964, a group of proud engineers at the mill gathered around a PDP-6 for a photograph. A student of computer culture will find much of interest in this photo: crew cuts, white shirts, narrow ties, breast pockets stuffed with pens, corporate badges pinned to shirts. At least one engineer is wearing white socks, another has thick, black-rimmed glasses. The one woman among the group likewise dresses appropriately: white blouse and dark skirt. Only one member of the group—Gordon Bell, the chief engineer of the project—is wearing a jacket.
But what about the PDP-6? Was it foolish of this small company, located in a corner of a drafty textile mill, to go up against IBM? Historian David Alan Grier once wrote about “the Great Machine Theory of [computer] History,” comparing it favorably to Thomas Carlyle’s notion of history driven forward by “Great Men.”3 Grier did not include the PDP-6 among his Great Machines, but he should have. It was a large computer designed to be employed by numerous users interactively—not as other mainframes were used. It would do this by means of a technique known as “timesharing”: letting the computer serve multiple users simultaneously, taking advantage of the fact that the PDP-6’s processing speeds were much faster than the eyes, fingers, and brains of the humans who would use it. The closest analogy would be to participate in a game of simultaneous chess with a grandmaster—an exercise the author strongly recommends to anyone who fancies him- or herself a good chess player.4 For the PDP-6, Bell was inspired by the work of Professor John McCarthy of MIT, who was a champion of timesharing: “John McCarthy’s definition of timesharing, to which we subscribed, included providing each user with the illusion of having his own large computer.”5
“The illusion” indeed. The MIT-trained engineers, with their white shirts and narrow ties, were selling a “dream machine.” This was the machine that opened up people’s eyes to see that computers, when coupled with human intellect, would change the world in ways far beyond what they were already doing in the fields of finance, business, and science.
After selling a few dozen PDP-6s, the engineers at the mill quickly upgraded it to an even more capable version, the PDP-10. Here was a Great Machine that Carlyle would have admired. Bill Gates learned how to program on a PDP-10, and he and Paul Allen used another PDP-10 to develop their first product for Microsoft, the company they founded. McCarthy left MIT, and he and his colleagues at the Stanford Artificial Intelligence Laboratory used PDP-10s to develop the new science of artificial intelligence, using the programming language LISP that he had invented. In 1972, when Whole Earth Catalog founder Stewart Brand visited the Stanford laboratory and saw its denizens playing the proto-computer game Spacewar!, he was looking at a console connected to a PDP-10.6
The illusion was powerful, and it did not stop with individual time-shared computers. What if one created not just a symbiosis between humans and one timeshared computer, but worldwide networks of time-shared machines? Program managers at the Defense Department’s Advanced Research Projects Agency conceived of such a network; as it took form, it drew heavily on the products of DEC. A logical map of the ARPANET, published in 1979, shows a rapidly growing network of about 175 computers and terminals; of them, over 17 percent were members of the PDP-6 family, and overall, two-thirds were DEC products.7 A decade later, the ARPANET had evolved into the internet, which, as its name implied, was intended to interconnect heterogeneous networks to one another. That was the plan, but in 1988, the internet was very much a monoculture of DEC machines. In that year, a mischievous twenty-three-year-old student at Cornell was able to cripple the internet with a “worm,” taking advantage of a minor flaw in the operating system of DEC computers, which comprised the majority of computers on the network.8
In the years that followed, DEC’s fortunes rose rapidly. The company leased all of the remaining empty space in the mill, and it acquired more modern facilities in the Boston suburbs. Many of these plants, as well as those of DEC’s competitors, were located close to Route 128, Boston’s circumferential freeway. The term “Route 128” became synonymous with “high technology.” Massachusetts governor Michael Dukakis, who had a reputation as a technocrat, erected signs along the road proclaiming it “America’s Technology Highway.”
But just as the signs were going up, the fortunes of DEC and those of the region were beginning to turn downward. The epicenter of dream machinery was moving to California, especially to Silicon Valley, the site of Brand’s epiphany at the PDP-10 console. Perhaps the New England winters were too harsh for illusions to flourish: McCarthy had long ago moved from Cambridge to Palo Alto, and Dukakis’s loss to George H. W. Bush in the 1988 presidential election was symbolic of the fall in the fortunes of the Bay State, the end of the “Massachusetts miracle.” In 1992, as the company faced heavy losses, Olsen resigned as DEC’s president. His successor, Robert Palmer, could not revive its fortunes, and in 1998, DEC was sold to Compaq. What little remains of DEC is now a part of Hewlett-Packard, based in Palo Alto.
And this photograph remains: a record of the day that the illusions— or are they delusions?—of the information age became real. Here are the men and women who built the underlying hardware of cyberspace, thereby letting the rest of the world dream. What would Thoreau have made of this? He had little regard for the instant communications afforded by the telegraph: “We are in great haste to construct a magnetic telegraph from Maine to Texas; but Maine and Texas, it may be, have nothing important to communicate.” But perhaps he would have softened his criticism. He also had this to say: “If you have built castles in the air, your work need not be lost; that is where they should be. Now put the foundations under them.” A century later, his fellow New Englanders took him up on that.
1. This and other quotations from Walden are taken from the online resource, Project Gutenberg, at http://www.gutenberg.org. It is unpaginated.
2. Throughout its history, the Digital Equipment Corporation was also known as simply “Digital” or by its acronym “DEC,” pronounced as a word.
3. David Alan Grier, “The Great Machine Theory of History,” IEEE Annals of the History of Computing 25, no. 3 (2003): 95–96.
4. As a teenager, the author once played—and lost—such a game, along with about a dozen others, at a chess club in Times Square.
5. C. Gordon Bell, Alan Kotok, Thomas N. Hastings, and Richard Hill, “The Evolution of the DECsystem-10,” in Computer Engineering: A DEC View of Hardware Systems Design, ed. C. Gordon Bell, J. Craig Mudge, and John E. McNamara (Bedford, Mass., 1978), 489–518, quote on 492.
6. Stewart Brand, “Spacewar: Fanatic Life and Symbolic Death among the Computer Bums,” Rolling Stone, 7 December 1972, 50–57.
7. Daniel P. Siewiorek, C. Gordon Bell, and Allen Newell, Computer Structures: Principles and Examples (New York, 1982), 397. The central switching device of the ARPANET was a modified computer made by Honeywell, whose plant was also located in Boston’s western suburbs.
8. Charles Schmidt and Tom Darby, “The What, Why, and How of the 1988 Internet Worm,” available at http://snowplow.org/tom/worm/worm.html (accessed November 2009).
©2010 by the Society for the History of Technology.
Filed under: Vol. 51 No. 3 (July 2010)
RSS