Inventing All The Time
John Backus spoke of his days wrestling the Super Calculator with a sense of nostalgia for the frontier. Oh, the machine was so complex. It was. And there was no textbook back then. The constraints were such a challenge... There was so much opportunity for ingenuity. You were inventing all the time.
Cuthbert Hurd - Thumbs Up
In the 1953 letter to his boss proposing his programming project, Backus emphasized the economic dimensions of the problem. At most installations, the cost of programmers' salaries in computer centers, typically 30 programmers per installation, was at least equal to the cost of the computer (the monthly rental for the 701 was $15,000, the equivalent of nearly $100,000 these days). In addition, Backus noted, one-quarter to one-half of computing time was spent in debugging. Accordingly, programming and debugging represented as much as three-quarters of the cost of operating a computer. With hardware improving rapidly and becoming cheaper, the proportionate cost of programming seemed destined to rise even further. It was a big problem, and getting more so. Cuthbert Hurd read the letter and immediately approved the request to begin the programming research project. He really understood, John Backus recalled.
Irving Ziller Joins The Fortran Team
In January 1954, John Backus got his first conscript, Irving Ziller. A graduate of Brooklyn College, Ziller joined IBM in 1952 and had been put to work programming plug boards on electronic calculators. The calculators were made from a series of these plug boards, roughly 8" 1/2 by 11" inches, filled with holes into which wires were connected by hand. It was another form of hard-wired programming. When complete, a plug-board would look like a miniature jungle of wires rising up from the board.Irving Ziller quickly proved to be both bright and extremely adept as a plug board programmer. In his apartment in the Riverdale section of New York, Ziller described his plug-board programming days in animated detail. This, as you can imagine, was a fairly tedious job, he said. Anyone doing plug boards understood the emerging need to simplify the programming process. So, when asked, Ziller was an eager recruit to John Backus's project.
Harlan Herrick Joins
Soon after, the team got its third member, Harlan Herrick. He was a math major at Iowa State University, and an outstanding chess player, who had won regional tournaments in the Midwest. He was awarded a scholarship to Yale University for graduate studies, but he was unhappy there. After reading an article about IBM's SSEC machine, he applied for a programming job and was hired.When he joined the FORTRAN team, Harlan Herrick had five years of experience programming IBM's SSEC and 701 machines. That made him a wizened veteran among programmers at the time. Within IBM, Herrick was known as a naturally gifted programmer, and his work was instrumental to the success of FORTRAN. At the start, though, he was the most skeptical because he was the most steeped in the programming practices of the time. Herrick was a member of the priesthood. When Backus first told him about the project, Herrick was incredulous. I said, John, we can't possibly simulate a human programmer with a language, this language, that would produce machine code that would even approach the efficiency of a human programmer like me, for example, Herrick recalled in 1982. I'm a great programmer, don't you know?
Next page > Part Two of FORTRAN The Turning Point
