Part Three: FORTRAN The Early “Turning Point”

FORTRAN was presented to the computing world in February 1957 at the Western Joint Computer Conference in Los Angeles. The gathering was mainly attended by the few dozen members of SHARE, an aptly named group of companies that used IBM computers. SHARE was made up of managers and scientists mostly from aircraft makers, big manufacturers, and government labs, and the members freely shared information, complaints, and even programs. This was before software was a separate business, so programming expertise was regarded more as useful knowledge to be shared than as intellectual property to be protected.

IBM's First Public Demonstration of FORTRAN

At the Los Angeles conference, IBM had arranged for what today is known in the industry as a “demo” – a public demonstration – of FORTRAN. Shortly before the conference, IBM had asked a few of its customers to come up with real-world computing chores, like calculating air flows for the design of a jet wing. The problems would be coded by assembly programmers, but also written in FORTRAN. When the assembly coders were done, the two programs for each problem were put on a 704 – one hand-coded and the other FORTRAN compiled. The FORTRAN programs ran nearly as efficiently, sometimes as efficiently, as the code of the assembly programmers – that is, the FORTRAN program consumed no more machine time than the hand-coded program to solve a standard problem. FORTRAN had proved its skeptics wrong. “It was a revelation to people,” Ziller said. “At that point, we knew we had something special.”

Daniel McCracken

Daniel McCracken first encountered FORTRAN in 1958 at New York University, where he had a graduate fellowship. By then, he had been a programmer for seven years, working mostly for General Electric. He thinks, though he is not absolutely sure, that his first FORTRAN program calculated heat flows in liquids. But McCracken does recall the sense of excitement he felt. With FORTRAN, he could suddenly program in a way that closely mirrored the mathematical problem he wanted to solve, instead of having to focus so much on coding for the machine.

FORTRAN, he figured, enabled a veteran programmer like himself to program at least five times faster than if he was working in an assembly language. “Maybe it’s hindsight, but I think I also realized even then that this would open up computing to a lot more people,” said McCracken, who is now a professor of computer science at the City University of New York. His instinct proved accurate, and he benefited personally from the growth in the programming population. For 20 years, McCracken was the Stephen King of how-to programming books. A Guide to FORTRAN Programming, published in 1961, was his first big winner, selling 300,000 copies.

A High Level Programming Language

The achievement of FORTRAN, perhaps most of all, was that it demonstrated that higher-level languages were possible and practical. With FORTRAN, a huge barrier fell away, opening up the software field for a steady succession of innovations over the years that would make it easier for people to program computers. Most young programmers today regard FORTRAN, if at all, as a historical curiosity. But those in computing during the 1950s still appreciate the impact of FORTRAN. John McCarthy is the creator of the influential Lisp programming language, he helped start time-shared comput-ing in the 1960s and the field of artificial intelligence. The professor emeritus at Stanford University is not impressed by the current state of computer science research, which he finds too focused on small steps and incremental improvements instead of thinking imaginatively and aiming for breakthroughs. But when asked what really impressed him in the early days, McCarthy replied without hesitation, “FORTRAN, that impressed me.”

Previous

• 1
• 2
• 3

Next