John Backus proposed the FORTRAN project 70 years ago 🎂

1

70 years sounds crazy in the computer science world, where it seems like geologic times :t_rex:
 The exact date being unknown, let’s celebrate it today as it is also the birthday of John Backus (December 3, 1924 – March 17, 2007), who would be 99.

A few quotes telling that story

Backus, J. « The history of Fortran I, II, and III ». IEEE Annals of the History of Computing 20, no 4 (December 1998): 68‑78. The history of Fortran I, II, and III | IEEE Journals & Magazine | IEEE Xplore :

This economic factor was one of the prime motivations that led me to propose the Fortran project in a letter to my boss, Cuthbert Hurd, in late 1953 (the exact date is not known, but other facts suggest December 1953 as a likely date). I believe that the economic need for a system like Fortran was one reason why IBM and my successive bosses, Hurd, Charles DeCarlo, and John McPherson, provided for our constantly expanding needs over the next five years without ever asking us to project or justify those needs in a formal budget.

Lorenzo, Mark Jones, Abstracting Away the Machine: The History of the FORTRAN Programming Language (FORmula TRANslation), SE Books, 2019, ISBN 978-1-0823-9594-9:

Backus had larger ambitions than simply writing a better assembler; rather, he was interested in automating and routinizing algorithms at a higher level of abstraction than mere systems programming. He wanted to, in the words of Richard Hamming, “buffer
the user from the machine itself.”
So, in December 1953, Backus penned a memorandum to Cuthbert Hurd [
] suggesting the development of an “automatic programming” or “automatic coding” system.

Backus, J. W., H. Stern, I. Ziller, R. A. Hughes, R. Nutt, R. J. Beeber, S. Best, et al. « The FORTRAN Automatic Coding System ». In Papers Presented at the February 26-28, 1957, Western Joint Computer Conference: Techniques for Reliability on - IRE-AIEE-ACM ’57 (Western), 188‑98. Los Angeles, California: ACM Press, 1957. https://doi.org/10.1145/1455567.1455599 :

THE FORTRAN project was begun in the summer of 1954. Its purpose was to reduce by a large factor the task of preparing scientific problems for IBM’s next large computer, the 704. If it were possible for the 704 to code problems for itself and produce as good programs as human coders (but without the errors), it was clear that large benefits could be achieved. For it was known that about two-thirds of the cost of solving most scientific and engineering problems on large computers was that of problem preparation.

The name FORTRAN will be find after the project was started, as stated in the already cited book Abstracting Away the Machine:

Third, a name was settled on for the language. “We would continuously invent very trite names for the system,” Backus said in a documentary film released by IBM in 1982. “And, I would come in with today’s name, and try it out on my friends. And they would all say, ‘Ugh, God, Backus—no, not that!’ ”
Herrick: “And one day he [Backus] came in and he said, ‘I’ve got it! FORMula TRANslation—FORTRAN,’ And I went, ‘Ewwwch?’ ”
Nelson: “But it was the only thing we had, so FORTRAN it became.”
Herrick: “FORTRAN. It sounds like something spelled backwards.”

Transcript of a presentation by John Backus in 1978:

So, in closing, let me remark that there seems to be general agreement among the original FORTRAN group that it was one of the most enjoyable projects any of us ever worked on. We were graced by this wonderful independent status that Cuthbert Hurd and my subsequent bosses conferred on us, so that we were completely on our own. We knew we were going to provide a valuable tool, even if few others did, and we worked quite hard. But perhaps the best part was the uncertainty and excitement of waiting to see what kinds of object code all that work was finally going to produce. I only wish that programming today could be half as exciting and enjoyable as the FORTRAN project was.
Thank you.

Other posts and documents about J. Backus

12 Likes
2

Note that I believe that Backus would have been interested by our discussions about generating programs with AI
 I don’t know what opinion he would have developed, but sure he would have keep an eye on the subject. His vision of functional programming was that you should define the problem and its constraints and the computer should solve the problem, not the programmer (by imperative programming)


1 Like
3

They already had Speedcoding, so something like FORTRAN was the next logical step. And what a step indeed


I’m guessing when the idea of a high-level programming language and its compiler (“translator”) was first mentioned to their bosses the initial reaction wasn’t very positive. Probably something like “Are you smoking stuff?” or “What kind of witchcraft is this?”.

4

I guess most people here are aware of this but for those who aren’t, 50 years ago someone filmed a short documentary for the 25th anniversary of FORTRAN. You will see and hear the people involved in the project talking about it, including Backus of course. Some quotes @vmagnin posted above are also there.
Backus himself seems to be the most modest person of the team by the way - despite the fact pretty much everybody mentions him as the leader.

Two quotes from the documentary that summarize what those people were doing:

We were in completely uncharted waters, we were trying to do something that basically nobody had ever tried to do before.

It was all invented at the time, it wasn’t a case of choosing between this method and that method, this theory and that theory. They were no theories.

2 Likes
5

His bosses seemed OK with the project. Inside IBM, the main obstacle was von Neumann, as stated in “JOHN BACKUS: a restless inventor”:

But first, Backus had to overcome the persuasive arguments of von Neumann who was then a consultant to IBM.
He didn’t see programming as a big problem. I think one of his major objections was you wouldn’t know what you were getting with floating point calculations. You at least knew where trouble was with fixed point if there was trouble. But he wasn’t sensitive to the issue of the cost of programming. He really felt that FORTRAN was a wasted effort.
Hurd approved the project anyway and von Neumann didn’t fight it any further. Backus hired an eclectic team of experienced programmers and young mathematicians straight out of school.

It is also related in the book Abstracting Away the Machine, with more details:

von Neumann exclaimed, “Why would you want more than machine language?”, although this quote may be apocryphal.

But yes, I think von Neumann’s brain was too powerful to need a high level programming language


They had also to convince programmers that a machine (a compiler) could generate a code that is as fast as a code generated by a human being. In fact few people were believing it would be possible. But it was also partly because previous attempts did not succeed to do that. So Backus’ team work hard to create from nothing an optimizing compiler, knowing that if it could not compete with humans in every applications the project would not succeed.

3 Likes
6

I read somewhere they initially expect the project to be done in six months. Finally, for the above reason, it took nearly three years (the delivery of the compiler on the 704s of IBM’s clients began in April 1957).

7

Of course people did ask how long it will take, and the answer was “come back in six months”
 This was happening for three years. :laughing:
Backus himself says that in the documentary, around 6:28.

Not that three years was much for such an achievement


1 Like