A PERSONAL ODYSSEY
My interest in electronics goes back to before the second world war, but I suppose my first involvement with anything that might be termed a "computer" was just after the end of the war, when I was in the RAF as a ground radar mechanic. This particular equipment was associated with the "Chain Home" radar stations. Housed in a Nissen hut, it comprised banks of uniselectors - in effect being based on an automatic telephone exchange of that period. It was known as the "Calc. Room", and converted ranges, bearings and elevation angles of target aircraft into grid positions and heights - a "polar/cartesian converter".
A few years later, I was writing a technical manual for an advanced transmitter/receiver for airline use. This employed a large number of relays for automatic tuning and aerial matching, driving servomotors for moving the various tuning elements. It was more complicated than it sounds, and really amounted to a special-purpose computer designed to solve some specific problems.
The next family of computers which came my way was quite different. Analogue computers gained quite a foothold in the late fifties, at a time when the digital machines were few, very costly and bulky, and used electronic valves (or "tubes").
An analogue computer solves numerical problems by using a physically measurable quantity (in this case, voltage) to represent numbers. It also used valves (although far fewer) but with a basic element - if you will forgive the technicality - of the "drift-corrected DC amplifier".
This equipment was very good at solving complex differential equations (such as one might use - for example - in the design of automobile suspensions) and also for real-time simulation of systems with a number of interactive variables.
Some large-scale radar and military systems simulators were produced using this technology, and some "hybrid" computers, combining analogue and valve-digital elements were developed. At about this time I was attending evening classes on computer design and programming (the programming was in machine language, and recorded on five-hole punched-tape!).
The transistor was now available to replace the clumsy valve in digital computers, and I soon found myself working as a systems engineer on a major government contract for an Action Speed Tactical Simulator. The details of this are not really relevant, but the way the technology was evolving at the time does give some insights into the pace and nature of technical developments. The central special-purpose digital computer was housed in some forty eight-foot high rack cabinets. The circuitry was carried on printed-circuit boards which slotted into "nests", and the back-wiring was incredibly complex, with many thousands of soldered connections.
The design started with point-contact germanium transistors, but within a year the superior silicon co-planar epitaxial transistors became available - they started the year at about £100 each, and soon dropped to a few pence - so the design was changed. As the equipment neared completion the new integrated circuit devices arose - but too late to be used.
We would wryly say that if it was built it was obsolete, and if it was on the drawing board it was obsolescent!
Not long afterwards, general-purpose scientific computers appeared, and these were eminently suitable for use in such applications, obviating the need to design large special-purpose computers. "Memories" were small, and much use was made of "hard-wired" and "patched" programming.
In the sixties I moved out of electronics proper, and into the advanced machine-tool industry, where numerical control and automated systems were beginning to be adopted.
A few years on, and I changed to a quite different environment - in Research. Minicomputers (the smaller "mainframes") were used for modelling, and for applications like automatic analysis, and with equipment such as the gas-chromatograph/mass-spectrograph.
The very first "microcomputers" - little more than toys - had come on the scene and had useful, though limited, applications.
At that time I was concerned with the worldwide distribution of specialised research papers and other literature, newsletters, announcements of special interest, international conferences, and so on. The mailing system in use was primitive, labour-intensive and error-prone. The availability of the new microprocessors and dot-matrix printers seemed to me an ideal opportunity for a new approach, so with the aid of a computer bureau I put this in hand. It was a success. People found it incredible that, once all the data had been captured (from questionnaires and other sources), the system could locate, say, a dozen scientific specialists out of a list of several thousand.
My final move was to the Computer Applications and Technology Division of a large multinational organisation. Here, to my amazement, I found yet another primitive distribution system (the cobbler's children oft go barefoot!) and immediately set about creating an effective computerised matrix, since there was now a greater range of microcomputers and relational database packages available.
Again, this transformed a clumsy manual operation into an efficient worldwide selective mailing and distribution system. The ongoing hard work lay in data-capture, updating and amending of records.
These last two examples are indeed trivial compared with full-scale computer projects, but they gave me great satisfaction and the sort of insights that can only be obtained by the detailed solving of a real problem - learning by doing!
Any of the decades from 1940 could be looked at as one in which computing technology underwent significant development. The forties might be seen as the decade of gestation, the fifties as that of infancy, the sixties as that of childhood, the seventies that of adolescence. It was in the eighties, however, that the technology really came of age and moved from specialised and scientific applications into the broad fields of business activity generally.
I was fortunate to experience this last decade from the point of view of (amongst other things) the Editor of a Computing Journal for a major multinational organisation. In this I had the dual role of participator and onlooker. Some four or five hundred articles dealing with computing applications and developments passed through my hands. Much of my editorial comment dealt with specific articles and the technicalities involved, but wherever possible I tried to use a "broad brush" and bring out what seemed to be the underlying principles and philosophy.
It is this more general aspect that is represented in the next two examples.
