Computers

​

Throughout my time at Boving I was involved with computers.  This was not through an interest in computers per se, but an interest in what they could do for you.  At Boving hydraulics department when I arrived, one of the big problems was calculating the transient conditions when the turbine lost the connection to the grid.  The turbine, with no power line to hold it back, would start to speed up towards “runaway speed”, about 80% above normal speed.  How quickly it speeded up depended on the inertia built in to the generator rotor.  The generator had to be designed not to fly apart (10s, or 100s of tons) and destroy itself at the maximum speed reached.  The turbine control vanes would start to close, to shut it down, and thus reduce the speed rise.  Closing the vanes would mean stopping the flow of water in the pipeline, which would cause the pressure in the pipeline to increase, by say 30%, and the pipeline had to be designed to withstand this pressure.  And factor in that the turbine characteristics would change, depending on the vane position and the rotating speed.  A Frenchman called Bergeron had devised a method of calculating all this by drawing lines on paper.  You would have an A3 sheet of graph-paper with lines all over it to work towards the answer.  One calculation, for one particular set of circumstances such as the rate of closure of the vanes, might take two days to complete.  And you had to do it several times over for each project.

So I think a primary criterion in my selection was how successful I might be in doing this task.  It was fun for a bit, and I became quite good at it, but I soon thought “this is no good for a game of soldiers”. 

​

Computers more or less existed (1964), and there was still doubt about whether to use analogue computers (where you devised electric circuitry to mimic the real world conditions) or digital computers (where you wrote equations to represent the behaviour of the real world conditions).  Of course, digital computers won in the end.  I wrote the equations for the behaviour of the turbine vanes, its speed and flow, the speed and mass of the generator, the dimensions of the pipeline and the characteristics of the water in it, and put them together in the form of a program which I could take to Marconi in Greenford to run on the computer there at a large cost for 5 minutes.  I would type out the program on a sort of typewriter which produced a paper tape, 5 holes across for each character, about 10 meters long.   When it was your turn on the computer you would run your tape through it,  It had to be perfect – one comma or character wrong and it would be rejected (it does not compile) and you would have to produce a corrected tape and take your turn later.  So in those days you really thought about what you were doing – now the process is so blindingly fast that you just tend to try various things until it seems to be right.  Anyway, you compiled your program and ran it for various conditions – turbine vane closure rate, generator inertia, etc., and you got your answer.  So if it all worked smoothly you could get a much more complete answer in an afternoon that you could by 2 weeks with the graphical method.

​

This was all quite cutting edge stuff.  There were a couple of US academics, Wylie and Streeter, who were doing the same sort of thing, but I reckoned that my method was better.  I published some of my work, but Wylie and Streeter, being academics, had the publishing of their work as their main interest, and eventually the rest of the world followed their inferior method.  Yah boo sucks!

Anyway, I think Boving realised the importance of my work, and my reputation as a whizz kid was established, even if I was a bit odd.

​

Gradually things moved along.  After a while we could have a typewriter in our office connecting to the computer via the telephone.  The rate was 15 bits per second, compared with 50,000,000 bits per second today (2014).  Eventually I wanted us to have our own computer, at a cost of perhaps £250,000.  To help this along I wrote software to do the company accounts, and I wrote a word processor as well (this was before Xerox introduced their first word processor.  We got our computer, a DEC PDP40.  Eventually that was superseded by a DEC VAX machine.  I bought a hard disc machine for it, a Fujitsu Eagle, for 20,000 1970 pounds, which was the size of a fridge and stored 20Mb!  I was more or less in charge of computing (as well as Chief Hydraulic Engineer), and I would do things such as run wires in the ceiling space through to various offices. 

​

Duncan McDougall was our main supplying agent, and he and I worked closely together.  He took my word processing software to market to other customers.  I was vaguely thinking along the lines of going into business myself too.  We needed a metal box with connectors to lead 24 users in to the computer.  The cost would have been enormous, so I proposed (and was accepted) to build it myself at a cost of about £1,000.   It was this money which I proposed to have available for any business venture which might arise.

​

It was at this time that I started sending 3 kids to expensive public schools, without adequate resources.  Also at this time the Amstrad CPC home computer came out, which had an 80 character screen (as opposed to 64 or 40 for the competition) on which one could write serious software, so I bought one, and that was the birth of Money Manager, my personal accounting software.   George, who was about 18, was involved in my planning.

​

I had been thinking of starting a business in about 1980, unspecified, but with the name Connect Systems, had opened a bank account, and had deposited in it the £1,000 from the connector box I had made for Boving.  Connect came from Only Connect, which I thought was to do with usefully connecting thoughts.