space holder space holder March Contents | Gazette Home space holder

Feeling Around
Farber's first exposure to computers came in 1954, when as an undergraduate at Stevens Institute of Technology in New Jersey, he participated in a thesis that involved building an automatic chemical analyzer. He and his colleagues built a relay computer, using a huge punch card to program it. "It was about two feet long by a foot wide, and, you know, you put holes in it, BUT IT RAN!" exclaims Farber. "And that's what got me interested in computing, though I didn't know there were computers at that point."
The following summer Farber, then twenty, landed a job with the U.S. Navy in Washington, D.C., working for Wally Deitrich, the person who built the first transistor analog computer. Farber spent the summer programming it, though he says he never planned to go into computing. He was set on attending graduate school at MIT to become an electrical engineer. But before he had a chance to pursue that goal, a special opportunity popped up that he couldn't resist: an interview at Bell Labs.
"I talked to this strange guy in New York City who was building something that I thought was a computer-programmed electronic telephone system. And I got in a big fight with them, because I thought they were doing it all wrong. I had nothing to lose, I was going to MIT. I left the interview and called the personnel people. They said, 'What the hell did you do? They want you desperately!'"
At Bell Labs, Farber found himself knee deep in helping design the world's first computer-operated telephone switching system. Early on in his ten-year stint with Bell, from 1956 to 1966, he became obsessed with solving a complex series of equations that dealt with the behavior of the telephone switch. The two-year process of solving the problem -- in his spare time -- got Farber "hooked" on programming languages, he says. During that period he also helped develop the String Oriented Symbolic Programming Language. Its better -- known acronym, SNOBOL, came, as Farber tells it, "From a comment that we did not stand a snowball's chance in Hell of finding an acceptable name." Since the 1960s, SNOBOL has been a widely used programming language designed to manipulate symbols and is the forerunner of some of today's programming languages.
After a few years at Bell, Farber was tempted to return to school to earn his doctorate, thinking that academe was the best place to gain the kind of knowledge and experience he was looking for. He'd been accepted at MIT for the second time when Richard Hamming, a famous numerical analyst and a friend of Farber's, asked him why he was going back to school. When Farber replied that he wanted to learn more about computing, Hamming told him, "'You're at the center of it!'" Farber recalls. "Which [Bell Labs] was at that point. I got a masters from Stevens later on."
Soon after deciding not to pursue a doctorate, Farber became involved in creating one of the first high speed networks for Bell Labs, which ran at speeds equivalent to today's T-1 connections. To gauge how far ahead of his time Farber is, the average home computer that dials into the Internet today uses a 28,800 baud modem -- several magnitudes slower than the T-1 he helped design thirty-five years ago.
This was also the project that got the grandfather of the Internet interested in networking and digital communications, which, combined with his telephone experience, landed him at the Rand Corporation, headquartered in Santa Monica, Calif., where from 1966 to 1968 he worked on several communications projects for the military. From Rand, Farber went to Scientific Data Systems and then -- finally -- on to academe. "I'm not sure what attracts me to academe, in hindsight, because I had a very good job, with a very good stock option. I had taught in the evenings, and I liked teaching. I had gotten into the commercial side of the house, and I enjoyed it, but it didn't feel right, so I decided to take a couple of years off and try teaching. During that time, I came upon this notion of a distributed computing system with one of the very early local networks and applied to the [National Science Foundation] and got what was at that time an enormous amount of money to go build it." Continued ...
March Contents | Gazette Home
Copyright 1997 The Pennsylvania Gazette Last modified 6/23/97