Early history of Fortran

A very rich site devoted to tracking down the source code for the original Fortran compiler:

My name is Paul McJones. I hope to use this weblog to discuss software history among other topics. For several months I’ve been studying the early history of Fortran, and trying to track down the source code for the original Fortran compiler. Although I just set up this weblog recently (June-July 2004), I’ve created back-dated entries to document my quest in chronological order

It seems most items recently are about programming language history... This site describes an interesting quest, which makes me wonder if the evolution of more recent languages will be easier to document, given the Internet and so forth. It would be rather amusing if LtU will once be used as an historical resource ;-)

The idea of preserving classic software is a good one. I think programming languages (and programming technology in general) are very good indecators of the state of the art and the major issues of the day (e.g., Java and the Net), so building a timelime by considering PLs sounds like a good idea.

We should also keep in mind that John Backus of FP fame was famous even before that for his work on compilers, and was involved with the Fortran team at IBM.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

More copies flying around these days

So someone's bound to have the source (even if it happens to be on an 8 inch floppy). As evidence, you can download Perl version 1.0.

I suppose the FORTRAN I compiler was developed before the hard lesson of backing things up took hold. Or was it because it was done before they actually invented backup devices?

And for bonus points, today's trivia quiz is to name these suspicious looking characters of Algol '60 fame?

8-inch floppy?

There were no 8-inch floppies back when the first FORTRAN compiler was developed (1954 on). The IBM 704 for which the first FORTRAN compiler was developed had up to 8 Kwords of magnetic core memory, up to 16 Kwords of magnetic drum storage, plus magnetic tape storage (1/2 inch tape, 2400 feet). The machine was described as an "electronic calculator".

The most likely form you'd find the source to the compiler in would either be punchcard, or magnetic tape.

I suppose the FORTRAN I compiler was developed before the hard lesson of backing things up took hold. Or was it because it was done before they actually invented backup devices?

They took backup quite seriously on those machines, but also had the benefit that most things had to be stored on tape, since there was no other mass storage medium. The "hard lesson of backing things up" was something that the PC world had to learn, not the mainframe/minicomputer world. But, I don't think the historical implications of those early compilers were foremost in their minds, and of course keeping data in a usable format from one generation of hardware to the next can be non-trivial.

Old days

That reminds me of a link that
Darius Bacon sent me recently.

    ...Indeed this was perhaps the first large program I had seen where the
    output of the assembler would actually run correctly. Debugged
    programms had generally been patched versions of assembler output. The
    programmer would patch by punching a few binary commands on a card and
    adding that to the end of the binary deck from the assembler. He would
    hopefully annotate the assembler listing so that the same change would
    be made to the source. The debugging cycle did not afford an assemble
    step in general. The cost of reading 2000 cards thru a card reader at
    150 cards/min and then printing at 75 lines/min made that
    infeasible. In one (just one) extreme case the programmer threw out
    the source assuming that he would never be able to afford to assemble
    it again.

Suspicious looking characters

Front row, left-to-right: John Backus, Edsger Dijkstra, Alan Perlis;
Back row, left-to-right: John McCarthy, F.L. Bauer, ???

Better late than never

Four out of six ain't bad.

Front Row: John Backus, Peter Nauer, Alan Perlis
Back Row: John McCarthy, Fritz Bauer, Joe Wegstein

Wegstein was the leader of CODASYL (COBOL) just a year before.

Source code trivia...

This reminds me of something that happened to me in
January 1991, when I was touring German universities
to talk about my thesis work on compiling Prolog.
At ECRC in Munich a scientist from East Berlin came to
me after the talk and told me that they had typed in
the source code of my earlier WAM compiler from the
appendix of a 1984 research report. I.e., they did not
have network connectivity to the West. I asked whether
they got it running and he said they did!

John Backus of FP fame was fa

John Backus of FP fame was famous even before that for his work on compilers, and was involved with the Fortran team at IBM

An amusing reversal.
John Backus, Turing Award

That was my point

I thought giving the link to the Turing award lecture wpuld be redundant, seeing as we disccued it in the past at length.

Maybe not such a reversal?

I know that John McCarthy was interested in developing a referentially transparent language when he invented Lisp, but pragmatism, hardware limitations and the like led to a more imperative approach in the end. Backus, as leader of the Fortran project, wanted the language to be able to express mathematical formulae, and I imagine he was forced to make similar compromises. Both Backus and McCarthy seem to have been functional/mathematical programmers at heart.

It's a pity that some language designers since then don't seem to have been as smart. I was pleased to note Alan Kay's preference for a functional approach in Felleisen's Functional Objects paper.

Now we just have to work on people like Hejlsberg and Gosling...