Scor­pio News


May 1989 – Volume 3. Final Issue.

Page 23 of 43

Developing Programs On Microcomputers
Over The Years

by Richard Beal

Paul asked me to write an epitaph, as this is the end of an era of 8-bit computing. I thought it would be interesting to look back at the changes in ways of developing software on microcomputers, concentrating on the 80-BUS machines. I am sure some of you will have feelings of nostalgia when you read this!

The first program I wrote for my Nascom 1 (in 1978 I think) was written out in pencil on pieces of paper, with assembler mnemonics on the right which I had learned from the Z80 CPU manual. I then hand assembled the program, looking up each instruction in turn. Calculating the jumps was a painful process, particularly the backward relative ones. When the program was typed in (in hex) and executed, it crashed. I had no clues, and the next stage was to single step the program until the problem was found. Naturally the bug involved a missing instruction which meant that every jump had to be recalculated and typed in again. Still, the program worked in the end.

Once ZEAP was available, which provided a text editor and an assembler, things were much better. It was not possible for me to assemble a whole 2K of object code, and this time it was NAS-SYS that I was writing. This code was particularly hard and slow to debug, because I could only test it by putting it into EPROM, turning off the machine, plugging in the chips, and turning it on again. If it didn’t work the only way to debug was by a minute examination of all the changes since the previous version.

The next great leap forward was floppy disks and CP/M. The MicroSoft M80 assembler was an excellent product, and WordStar was a fast text editor. In fact only recently has WordStar been made to run as fast on a PC as it used to under CP/M. I wrote another ROM based operating system, RP/M, using these tools.

I made several attempts to get away from Z80 assembler and wasted some effort on a language called PL/M. Then good C compilers were developed and I wrote a communications program called DIAL in a mixture of C and assembler.

Then the years passed and I had very little free time. I used IBM PCs and had a Compaq Plus followed by a Toshiba 3100 at home, but have written very few programs. Recently I tried out the latest MicroSoft products consisting of MicroSoft C 5.1, Codeview, Quick C, and MASM 5.1. These are a very professional set of tools

Page 23 of 43