Dave Pitts is making progress running Fortran II

I have been negligent in reporting impressive progress made by Dave Pitts emulating IBM 7090 software. As Leif Harcke posted to alt.folklore.computers and bit.listserv.ibm-main on 2 February 2005:

Dave has developed his “asm7090” cross-assembler to the point where it can assemble the core of IBSYS from MAP source. The resultant IBSYS image will run the Fortran II(?) compiler on a modified version of Paul Pierce’s “s709” 700/7000 series emulator. The object code produced by the Fortran compiler does not run under IBSYS, however.

The cross-assembler and emulator are written in C, and build under Linux. Presumably they are easy to port to other POSIX-compliant systems. Details on the project are available here:

http://www.cozx.com/~dpitts/ibm7090.html

If anyone is interested in helping out, Dave could use a hand debugging the emulation and getting IBSYS and other related system tools working.

Today, Dave told me:

With the current version of the emulator, 2.0.4, I’ve been able to run the following:

1. FORTRAN IV – Compile and run both Primes and Laplace programs.
2. COBOL – Compile and run the hello world program.
3. FORTRN II – Compile the Primes program. The exec doesn’t work, I get a checksum error loading the runtime (bad tape??).
4. MAP assembler – I’ve only run the assembler, didn’t try to run output.
5. FAP assembler – I’ve only run the assembler, didn’t try to run output.

My current IBSYS tape has the nucleus assembled with my ASM7090 cross assembler. Also, I re-assembled the COBOL compiler with ASM7090 and put on the tape.

I know that Leif Harcke has been hacking on the FORTRAN II. I think he’s stuck at the same point as I am. I’ve been trying to get the FORTRAN II parts to assemble with ASM7090 to replace the tape image with a NOPed checksum test.

Systems Manual for 704 Fortran and 709 Fortran

In January, Peter Capek told me that while cleaning up a file cabinet in his home he’d come across:

… a detailed description of the FORTRAN compiler, dated in 1960, and explicitly distinguishing between the 704 and 709 versions, but covering both. It looks like it was typed and what I have is probably not an original, but likely one of very few copies. It’s a couple of hundred pages, and describes each section of the compiler, including table structure, in considerable detail.

Peter very kindly made me a photocopy of this 264-page document, which is a wonderful complement to the actual source code. I hope soon to be able to provide web access to this and the other Fortran documents I’ve come across.

Peter is a frequent contributor to alt.folklore.computers and supplied information for Frank da Cruz’s “The Columbia University Computer Center in 1965” article.