Sinclair User Profile

MIKE MITCHELL works for RICHARD DANA Corp as a mechanical engineer. The company designs and builds automatic assembly equipment which need custom cams.

Mind you that I never intended to marry it but like many of my projects, it began with the thought that I could do something better than someone else and has progressed to becoming a part-time obsession.

When I began my present job, my employer was renting a time share computer with an old 110 baud teletype machine as a terminal. That computer had special software to generate cam profile numerical data that was then used by an numerically controlled milling machine to cut the cam profile.

To design a cam, the engineer made a long distance call to Michigan, logged on the computer, input some parameters, waited and finally received numerical control codes punched onto paper tape.

A machinist spooled the paper tape onto a portable reader and downloaded the control codes into his automatic milling machine.

The time-share computer service cost $300.00 a month plus about $80.00 per cam design. In addition, the 110 baud date transfer rate and the ‘user-unfriendly’ software cost an engineer 40 minutes per design.

On my own time, I began writing SuperBASIC software to replace the time-share system. I convinced my boss that the math was correct but was unsuccessful downloading the numerical control code from my Sinclair QL to the milling machine.

It turned out that the RS-232 on the particular machine was synchronous. not asynchronous, a distinction lost on many who are computer literate and harder yet for those unfamiliar with computers to understand.

However, my boss was interested in the idea. He believed that it was only a matter of getting a ‘better’ computer since mine was ‘just a toy’.

We got a CAD/CAM syste comprised of a Compac computer with Computervision software. I planned to translate my program over to the Compac.

I had often heard from MS-DOS ‘experts’ that writing programs in BASIC is impractical and unprofessional, but I assumed that there would be a BASIC language for IBM compatibles similar to SuperBASIC. Turbo Basic by Borland looked promising and I set out to translate my 50K program.

It didn’t take long to find out that Turbo Basic does not hold a candle to SuperBASIC. The MS-DOS ‘experts’ who denigrate BASIC as a programming language are talking about BASIC, not SuperBASIC.

As good as Turbo Basic is compared to other BASICS for IBM’s, it is woefully lacking in features we take tor granted in SuperBASIC. It does not let you label your loops. String slicing is tortuous and STRS$ behaves differently depending on whether the humber is positive or negative.

Emulating QDOS’s coercion defeated me and WINDOWing and SCALEing in SuperBASIC had no equivalent in Turbo Basic.

Every procedure and function required a Turbo Basic SHARED statement if a variable were global, chewing up bytes at an alarming rate. My QL knew that if I didn’t declare it as LOCAL, it must be global.

Bugs. errors in documentation and inferior code finally forced me to admit to my boss that I had made a judgement error in recommending an MS-DOS equivalence as a solution.

Well, I talked my boss into doing what we should have done in the first place. He got a full blown QL with dual floppy drives, Trump Card and color monitor. My program is now TURBO-CHARGEd in modules and works faithfully everyday.

The code for the automatic milling machines is transferred to the COMPAC because it has a 40 MEG hard drive to store the stuff and is wired for the necessary data communication — the QL does the thinking.

MS-DOS ‘experts’ simply do not know about SuperBASIC and many who cut their teeth or whatever on an expensive MS-DOS machine are too embarrassed to even acknowledge the Quantum Leap.

The only problem that I still have is the constant temptation to make improvements to the program which, by the way, paid for the QL system in less than three months of not needing to time-share on another computer.