Pretty nifty programs people have posted. You can thank Elmer Clausen (from Buffalo NY) for all of the math functions in FORTRAN 99/9640, he basically developed them all and gave them to me to include in the package, as my original implementation of the same was so poor. I was more focused on the compiler itself rather than the math libraries. Elmer was actually driving a plotter with the FORTRAN and he had a real interest and need for rock solid math libraries. FORTRAN still seems pretty unique in that is supports complex math easily. Interesting as well in that the TI-99 and Geneve both had interesting BCD based floating point arithmetic which was a lot more precise than what we ended up with IEEE floating point becoming the standard. I seem to remember that had something to do with TI being used for accounting type functions before it was migrated to the home market, so we ended up with those slow BCD floating point representations that actually were quite precise and didn't have the binary to decimal round off issues.
The original FORTRAN compiler was based on FORTRAN 66, but I added a lot of FORTRAN 77 stuff (like if/then/else/endif, do while, etc.). The compiler is pretty interesting in that it used an interpreted machine to parse the FORTRAN source code, that interpreted machine was loaded into video ram. We were so tight on memory and resources back then, it took some real tricks to do useful things.
I managed to pull together most of the FORTRAN 99/9640 and TIC (Full-C) source code. It is pretty massive, most of it was written in assembly (the TIC compiler is written in C). One thing I am missing is the tasm assembler, I have a couple of Amiga floppies labelled a99 (original name for tasm) but I don't have a machine to read them on any longer. The tasm/a99 assembler came from a posting on BIX (Byte Information Exchange, I used to be a moderator there) from a professor from Arizona from memory. I greatly enhanced it, one nice feature it had was macro capability. The TIC compiler was based on some public domain compiler that was pretty buggy at the time. One of the fun things I did for both FORTRAN and TIC was to write peephole optimizers for the code generation.
The original FORTRAN was coded entirely with Editor Assembler / floppy disks, etc. Hard to believe I got all that working with those tools at the time. Eventually I got a Geneve which I really liked but it was SO buggy with the operating system MDOS. I ended up getting an Amiga eventually, that became my favorite computer for a long time, I actually ended up owning three of them. I ported the tic + tasm tools over to that environment and did a lot of development there.
Professionally I still work for a company called Schneider Electric as a project manager, and work on computer systems primarily to manage the electric power grid in the USA. I also work on a lot of new stuff like Microgrids, and even did a lot of projects in New York City for the transit system. I will retire in the next few years and hopefully will move from New England to South Carolina (I have a second home there). I have had a great career traveling the world to a lot of places that most people would never get to see.
I can post this source code for the TI-99 and 9640 compilers/linkers/demo programs/tic compiler, etc. but not sure where to post it. Any ideas?
LGMA Products (Little Green Men Associates)