Computer History

The hardware we used has been and gone. Programming languages go too but they last longer. COBOL [ COmmon Business Oriented Language ], the first of them is still out there in legacy systems. C lives on, just like C++. We used Assembly language, which talks direct to the machine. One line of code produces one machine instruction. This is more reliable than a language with powerful instructions that do complicated things, which may be more or less what you want most of the time.

It all started around 1980 with getting access to a real 8 bit computer, the Research Machines 380Z [ Research Machines Ltd ] offering which had a Z80 made by Zilog in its heart running at 4 MHz. It had a maximum of 64 kilobytes of RAM [ Random-access memory ] enough to do useful things in the real world. A major virtue was the front panel, a display that showed what was stored in various registers and a block of RAM. It meant that we could go through a programme one step at a time; great for testing.

Upgrading to a single sided 5.25" floppy drive was a definite improvement with 72 kilobytes of storage. Having twin floppies was better yet. Porting programmes from RML's system to the new CP/M [ Control Programme for Microprocessors ] was part of it. Changing to the new fangled MS-DOS [ Microsoft Disk Operating System ] from Bill Gates was another step, a fairly easy one in the event.

It may sound primitive to modern geeks who expect gigabytes of RAM. Bloatware was not an option back in the 1980s.

The RML was able to run an implementation of DES, the Data Encryption Standard which was used by banks to make their money transfers private. It kept criminals out of the system for several years. My version ran in 4 kilobytes. It had to because that was all we had. It was a good way to learn programming, by making something work.

Having got the possibility the next step was the idea that was going to make it useful. This was to be using computers that could talk to each other using a standard radio voice link. It was achievable because CML [ CML Microcircuits ], a firm in Essex produced leading edge chips which converted digital signals to voice frequency sine waves. One of their customers was the aerial warfare research department of the Irish Republican Army. Surface to air missiles can be useful against helicopters.

They were working with the FX101, a voice frequency operated switch; a strange choice. The perpetrators were given a fair trial but not hanged, sad to say. The IRA also used Mercury tilt switches bought from RS Components [ Radio Spares ] to detonate their bombs but this is about our system which was legal.

John was the man who had the idea of using radio links to connect parts of the public water system to each other. Measuring water levels, turning pumps on and off, monitoring sewage systems are all important. Connecting them to the operations room meant using land lines which never all worked at the same time. It was a maintenance nightmare. That is why radio made sense and still does.

The system was designed from scratch. John did the circuitry then the artwork for the motherboard. It was double sided with plated through holes to join wires on opposite sides. The tracks were laid down on a transparent sheet using blue tape for one side while the other was red. The maker of circuit boards used photographic techniques to produce the prototype. John was responsible for the hardware. I did the easy bit, the programming. We had a whole 4 kilobytes of ROM [ Read Only Memory ] to store the programme in with another kilobyte of RAM [ random access memory ] to store real time variables. Multimegabyte memory chips were not even on the horizon.

Then it was a matter of putting the software into action on the motherboard. This was the real challenge, one that nearly cost a divorce as well as many sleepless nights......

The programming language came later. It was called METAL, the Mike Emery Telecommunications Algorithm which made it easier to put sub routines together in different versions.

Having the design team was not enough. We needed a salesman and a production manager. They turned out to be Tony and Chris. Both played the same roles in the real world full time. This was a shoe string operation that was going to go big time - we hoped. Getting sales was one make or break issue; the one that went wrong. Then there was finance. Installing a major system was going to be another.

We had one major customer in our sights but nothing came to pass. This was when John and Tony, hardware man and salesman decided to branch out on their own. Chris and I kept our full time jobs while a dream faded. Steve Jobs made it from a garage to a world changing force. Lots more do not. Business is not easy.

What should we have done to break through is the question that lingers. We knew nothing about marketing which is about how to sell. Deciding how to approach the right customer in the right way is the aim. Mass markets may call for television advertising. Cultivating people with major budgets is the way in other areas. Selling our design to a big company with big overheads would have made sense. It has the credibility, the financial resources that we just did not. Management consultants might have been the way to go but they would probably wanted cash up front or a stake in our future. America has a Venture capital industry. England does not. Banks are unhelpful when it comes to start ups. They are also irresponsible when it comes to housing loans.

Would we do it all over again? I think so but the second time would be different.

Mike Emery
October 2011