Abstract: A lively, chatty interview full of personal anecdotes relating to his life and work experiences. Concentrated on Maths at School and awarded scholarship to Cambridge. But opted to do National Service in Royal Signals, for two years, rising to rank of Lance Corporal. Opted for a degree in Physics at University of Birmingham, graduating with II/1. Saw LEO Computers advert for engineers, taken on at Minerva Road and after training school became maintenance engineer on a range of LEO/IIs. Spent 3 years with LEO, then joined Honeywell as manager of software development. Subsequent career with Rank Xerox in marketing, as consultant with CACI, then working with James Martin and after a spell in Australia (1997), set up as independent consultant before retirement.

Transcript :

LEO COMPUTERS LIMITED  -  Oral History Project
Interview with Michael Mills by David Phillips

So, it's the 7th of February 2017.  I am David Phillips and I am interviewing Michael Mills.

[MM]:   I was born in Wimbledon in 1935.  My parents had moved to Wimbledon and opened a furniture shop in Broadway, Wimbledon.   I was born in Stanton Road, just off Worple Road.   My father, unfortunately, had chosen the wrong place for the shop and they went bankrupt or closed the shop down.  He ended up with very little money but luckily they had enough left over so that, with my uncle, they opened a new shop in Gravesend which was successful and it served them for the rest of their lives.

We moved from London to Gravesend and then came the war and after being evacuated briefly to Norfolk, we moved to Luton, which was not a good idea because Vauxhall make cars there which, you can imagine, during the war was a prime target for bombs and so.

My father stayed on in Gravesend and worked for the fire service, was the water officer responsible for providing water throughout for officers putting out fires.

 I started school in Luton. I went to primary school there.  And then at the age of nine, I think, I went to St Albans School, a private school.  I used to travel every day on the bus for half an hour.   I was at St Albans for a year when we moved to Southend. We had a family connection with Southend.  And we stayed in Southend.  I went to secondary school, Westcliff High School, where I had a very good education. I was lucky.  And then when I reached the age of eighteen [1953] I was due for military service or I could go on to university.  

[DP]:  Were you showing any particular strengths in subjects?

[MM]:  Yes,  I had a very unbalanced sixth form education.  I had maths, maths, maths and maths.   And morning prayers, I think, which, I suppose nowadays would be ridiculous. But I was very lucky. I was in a year with half a dozen really bright people and I think that helped me to be with people because we stimulated each other.  A nice story was doing maths and I lived very close to the school as well.  I think every day of my school career I was late and did detention regularly.  But a good friend of mine and I used to do our homework together and we used to write up alternate questions and the masters used to be very upset about this, which was absolutely ridiculous.
  We worked hard. We enjoyed doing the work and they just  found it quite difficult that we worked together. But I always thought it was actually a real plus point in motivating people. 

 I had done physics and  I might have been doing a tiny bit, but it was primarily maths.

[DP]:  What was the aim?

[MM]:   I think we enjoyed maths and in fact I applied to go to university to study physics, which I enjoyed, and I got a place at Downing College, Cambridge.  But I didn't want to go straight away because I felt I was too young. So I decided to go in the army and do National Service early. This did me an awful lot of good because I was very soft and gentle, if you can imagine. I'd been bullied at school earlier on when I was eleven, twelve, thirteen.

 Well people just do, you know, because I wasn't tough enough I suppose.  And going in the army  you mix with an awful lot of people. Well you did in those days. You may remember that youths who had committed minor offences were offered jail or the army.  And I had to learn to cope with all these people and that did me a lot of good. I think that was two years.  

 I rose to Lance Corporal. Well  it's actually the lowest job in the army above private.  Privates can escape and they can always find excuses but the lance corporal is really at the bottom of the pile.  

[DP]:  Was your maths being used in any way at this point?

[MM]:  No, I was in Royal Signals and I was repairing radios. I had an easy life, you know, for quite a long period.  I was excused guard duty, which was always a good thing, because we were on call all the time.     I was in England for six months at Catterick Camp with people committing suicide regularly, which is rather nice to know.

Then I went out to Germany, only to discover that the colonel in charge of the camp where I'd been, where people committed suicide, was now my new colonel. 

We were in a detachment at a place called Paderborn, which  had been a German army camp.  Actually it was much better than that. Near there was a small town called Klipspringer and the army had taken over the Spa Hotel. So you have to believe this, in the army for a while we had our own bath house and marble baths. Our own man used to adjust the water because he had a big brass key to turn the water on and off.  So we, squaddies, used to luxuriate in these marble baths and, ‘Over here! Over here! More hot water, please boy!’.  And this guy would come with his big brass key.  Beautiful grounds, being a Spa Hotel, you can imagine it was beautiful.

With the trees and the lake. As far as I know they replaced it. Because the army had taken over the hotel there was a new Spa built subsequently.  I remember doing guard duty there.  And then we moved to Paddo. Then we moved to Senni Lago. Paderborn was a big town. Klipspringer was a small town, quite a small town.  Senni Lago - Lago’s a camp I think.  And Senni Lago was the army camp where I was for about a year I think.  

Well they did exams and I don't think you had a choice.  I think you did exams and they put people where they wanted them to.  And as I had a reasonable education the signals was fairly high up in this hierarchy. So I got into signals.   In Catterick my barrack’s corporal, who looked after us, was a very nice guy who was essentially in the army. 

Going in through the interview system which they take you through, they said ‘Did I play any sport?’ And I said ‘rugby’. And I discovered, at that time anyway, with this big intake of people every month, if you played rugby you went to a certain training regiment at the Royal Signals.  So they were all rugby players there.  Now the barrack corporal would go off and play games. He primarily played for England. If he didn't play for England he played for the north of England. If he didn't play for the north of England he played for Lancashire and if he was really pushed he didn't play for the army.  So we hardly ever saw him. An amazing life - I didn't realise this.  That was his army career, it was playing football, playing rugby.  So we didn't play.  But we just got chatting to various places and their signals and I knew something about electronics. I'd built radios before and so on - at home.

[DP]:  Would that be valve radios?

[MM]:  Yes and transistors and also crystals. I had some old equipment that would have been worth a lot of money nowadays - old loud speakers, big horn speakers and crystals with earphones.  So  I was a natural for signals - I fitted in.

[DP]:  What did that involve?

[MM]:  Repairing radios. I’ll tell you one lovely story. We were signals attached to the seventh armoured division and in fact we were a fairly small attachment. And the radio repair group was about five people, led by a staff sergeant. So we were all very pally. It was very relaxed.   This wasn't formal army stuff.  In fact I was amazed because you used to get visits from the guards regiments. And there it's unbelievably strict. I mean absolutely.  If a corporal goes by they stand to attention and all that.  I was amazed because this wasn't the life I was leading at all. One day I was there and the phone rang and the staff sergeant said ‘You've got a problem.  I can't come but  I’ll send you an expert along’.  And he put the phone down. He said ‘you're going to repair a tank’.  I said ‘I can't repair a tank. The nearest I've been to a tank is in the cinema’.  Because even though it was an armoured division regiment we were a separate detachment. We were in the armoured division.  We had a badge on our shoulder but never saw these things.  He said, ‘Off you go’.  So I went to the tank.  This tank had a small problem. The internal communications between the commander of the tank and the driver was in a separate compartment. It didn't work.  And this tank had always been like this and it went around with somebody sitting on the outside. You had to believe this. It’s the army so it's legal.  And the commander used to shout down to this guy and he used to then tap the head of the driver and go...

[DP]:  ‘Left’.

Anyway, my first challenge was how to get into the tank.  And, remember I'm the divisional expert because, I'd been sent down.  And as I marched across this vast square I remember, as a young officer, I was trying to think  ‘what can I do?’ and suddenly one of those moments of clarity came. As we arrived at the tank I said ‘after you sir’, and he climbed in and I just followed him in and that's how I learnt how to get into a tank.  So I learnt  how to be sharp with  my brain.  Anyway we did repair the tank. It had a very simple fault which took me two days to repair because I hadn't a clue what I was doing.  I had to just work it out by trial and error.  Eventually the fault was extremely simple.  Do you remember, the turret revolves?  Well there was a communication system for lots of brilliant purposes. But for communications between the two the designers decided that the communicators might get damaged. So they had two sets of communicators; A and B.  And the driver was on A and the commander was on set B. So  that was all that was wrong and it had been built that way. Well once I found out there were two lots of things I just switched over the wiring and it worked perfectly well.  It was just a case that they'd been wired incorrectly.

Well I started with these funny little junction boxes which I recognised. We had the same kind of things in jeeps and so on.  I unplugged the junction box and nothing was working. I thought ‘Oh, I need a new junction box’.  But in fact there was a control circuit that went through all the junction boxes to make sure that everything was plugged in.  I didn't know this, so every time I unplugged something, nothing would work.

My other fun time was when Princess Margaret came out to inspect the division and we went out on a vast grass plane.  I think you'll see why.  And I think there was a landing strip, which was the main one. They put a few viewing seats out along there and the main procession was  all out in the field - the division.

And the Princess would go past all the people in the field and then we’d move off. Then she’d go into the sand and the few would move off around through the tracks till it got to the runway. Then all would drive past her and they would all salute.  And at the end of the runway, well that was the day.  But when it was first tried I had this privileged position because I was the radio mechanic for the event.  I was actually in the brigadier’s jeep, repairing his radio at the time.  So he said, ‘Off’, and there was a tremendous noise and all the vehicles moved off.     And when the dust had settled half the vehicles were stuck in the mud.  I didn't think this was going to be a very good display. The army couldn't even move round. Absolutely extraordinary. Well, they sorted out the vehicles and on the day it worked alright.

Actually it was during one of the exercises when we were trying it out (because we did it several times).  I was in the brigadier’s jeep, repairing his radio.  And I knew the brigadier and we got on quite well together.  And his radio was always breaking down.  He said, ‘Why does my radio keep breaking down’, and I said, ‘I wouldn't know sir’.  But he was doing the tour of inspection of the troops, and I was waving at him. He didn't realise at the back it was me.  The guys are going ‘What in the hell is he doing?’

[DP]:  You were waving to the crowd?

[MM]:  The crowd as a royal visitor.  We had excitement from time to time. So I left the army. I did two years in the army. Then I went to Birmingham University.

[DP]:  Right.  What happened to the Cambridge offer?

[MM]:  Well I'd done Latin at school for a year and given up.  And Cambridge at that time required you to have Latin.  They said 'As soon as you've done Latin and you've got your GCE Latin well you can come to Cambridge’.  And I thought, ‘I don't want to waste another year of my life studying’.   

But then another part of my story, ‘Why did I go to Birmingham?’. Well back in 1947’ish, my father borrowed my uncle’s Ford Prefect car and me, my sister and my parents had driven down to the south of France, which was quite something in 1947. There weren't that many people on the road.  Petrol was drawn from the French machines. They used to have a cylinder and a handle you turned. You turned the handle and you pumped fuel into the cylinder and you could actually watch it till it rose to ten litres or something.  And having filled the cylinder with that, you then took the pipe and put it into your petrol tank and you knew you were getting ten litres because you were watching it. I'd never seen any machines like this.  

We were in the south of France one evening and we’d gone to some outdoor restaurant and my father asked for the menu. He’d got talking to a Swiss couple and the guy took one look at the menu, and said ‘You don't want that’.  And he called the waiter over and demanded, in Italian, an Italian menu. The menu prices were half what they were on the English menu.

And this couple had three children and they said, ‘Why don't we exchange holidays for the kids?’  So I'd gone out to Switzerland. I went several times by myself, at the age of twelve or  fourteen by train. Great fun.   While in Switzerland I'd gone with Rolf, the boy in the family, to the mountains.  Anyway, while I was there I met another family called Muntz and he said he was a dentist and had twin daughters, about my age.  So I met them, and would you believe   we used to write to each other.Pen pals.  

So when I was in the army I used to write to them and they lived in Birmingham.  So when I had to choose somewhere to go - I had no idea at all - I wanted to go away from home.  I'd got to an age where, my parents were lovely, I got on very well and I respected them, but I always felt constricted. I wanted to be on my own. I would be about twenty.  So I had to choose somewhere so I chose Birmingham because of the Muntz twins.  All this chain of events is quite extraordinary . It so easily could have been somewhere else.

[DP]:  So you applied to do what at Birmingham?

[MM]:  Physics. I don't know why I didn't do maths but I liked using my hands or something. I think they accepted me by post while I was in the army.  I was very cross because I was hoping to get some leave, but they wouldn't - they just accepted me.

[DP]:  Where did you live when you went to Birmingham?

[MM]:     I lived in digs in a little house.  Perhaps irrelevant, it was Jewish.  And this was a house, I don't know how I came to know it but anyway, there was a Jewish couple and a sister and they'd come out of Russia, I believe. He was a tailor and he still sat on the table, which I'd never seen, cross legged.  And he was married and the other woman must have been the sister.  They were quite elderly at the time I stayed there and they looked after me very well. They made very nice food and I did very well and I stayed in a lot.

[DP]:  What did you pay?  

[MM]:  A few pounds, very little, and my parents supported me.  I remember I met my father’s accountant because they were somewhat friends.  And I explained that my grant had gone up, because I had a living grant.  If university was paid for you got a living grant. Five pounds my grant was. He said, ‘Is it five pounds a week?’  I said, ‘No, per year’.  And he was amazed because even in those days five pounds a year was not a lot of money. It was spreading it very thin across a year.

[DP]:  Was this because there was some sort of look at your parents’ income?

[MM]:  Yes,  my parents’ income was too high. I've got a feeling that I lived off something like ten pounds a week. It might have been ten pounds a month. So, I went to university from 1955 to 1958.

[DP]:  So you're there, you're settled in, and you started ...

[MM]:  In the physics department.  And I did geology for the first year.  I never did a subsidiary.  And I failed.  And I asked and we got our papers back and I noticed that they had made a mistake in the marking. So surprisingly enough I managed to squeeze through by accident.  I was alright, I got a 2A degree. Not first class because I wasn't willing to work hard enough.  I had a ball at university. I was quite frightened when I first went. I'd been out of education for two years - remember I've been in the army.  So I was wondering what was going to happen. All these bright people coming straight from school and it did take me a few weeks to catch up. But by Christmas I was up to speed and all was well.

There was one other person that had been in the air force and may I say we were outstanding - we stood out - from a maturity point of view. Because all of the others were ‘Yes sir’, and ‘no sir’.  And the lecturers and the professors hated that. They didn't want to be treated like school masters. They wanted to be treated as intellectuals, you know, sharing ideas and challenging them. And the two of us did because we were two years older and we’d been through the army.

[DP]:  Confident?

[MM]:  Yeah.  And we knew how to bullshit people and to some extent deal with people.  So I felt very pleased with the choice I made.  And I did lots of things at university.  There was a university newspaper - I was on that.  But there was also a war newspaper and I became editor of that newspaper, which was a big pin board where the news went out. It was well respected so that was rather fun.  

[DP]:  What about the course itself. Can you remember the physics course?

[MM]:  Well I remember our professor.  He was an old man and he described physics in the year 1900 and we covered the entire course.  The subject didn't matter, which was an entire degree in 1900.  We covered it in the first term of university because things had moved on.  Birmingham was interesting. Few people realised but on top of the cupboards, there were some boxes. I looked in them and they were the early magnetrons. Because Birmingham was where the radar magnetron was designed and there they were just stuck on top of the cupboards - real bits of history which nobody bothered about.

 I remember the course and my tutor was Professor Frenlin.  I remember I had a rather young, naive young girl, not a girlfriend, who we  went to sessions with our tutor and he used to say to her, ‘Now you, relax, it's holiday time, take your time’, and to me ‘and you can bloody well work’. It was quite interesting. He was quite bright. He knew what he was dealing with.  And I got on fairly well with him.  

I got called in to see the professor - not Frenlin, but the head of the department, who was a really  old man.  He opened the discussion. Poor man, he was a little bit nervous and I wasn't. ‘I hate to say anything but looking at your results for practical work al the students have done about twenty experiments and you've done eleven. Can you explain?’  I said, ‘Is this a discussion on quantity or quality?’  because I was really cross. I thought it was an invalid question because I found I was interested in the things I did. When I did experiments I often found that the experiments were often very straightforward. There was a little booklet telling you what to do.  And then they show the figures for this formula which I didn't think was  university level stuff. But anyway that's how it was.  And I often found that when I did my experiment my results didn't fit exactly the formula - they kind of fitted.  And so I was quite interested in why, in what I had done wrong?  Why were there these variations in the theoretical value?  And that's why I was doing so few experiments - because I'd spend another week trying to understand where it went wrong.  

The only university level experiment we had, I remember, was  when I opened this little booklet and it said ‘make an amplifier with ten million amplification'.   I turned over the next page and it was the back.  So I went up to see the lecturer and I said ‘I think we’re missing something’.  He said ‘No, that's the experiment’.  There were no instructions. It was just because all the other experiments were, you know, ‘Take this piece of wire and screw it into there and measure this and show if it's ...’  And that, to me,  was the only experiment I felt was  real, because that was university level.

[DP]:  Did you discover why your figures didn't match?

[MM]:  Sometimes.  For example, we had work with a telescope (You shine a light through a prism). You Look through the prism there and you can look at the lines, and I realised that my eyes had a problem focusing.  They could focus on the cross hairs, or they could focus on the light, but that was a different place.  So when I made the measurements I never knew whether I was measuring to where the light was or where the cross hairs were.  And so I got a second spectrometer and the cross hairs were lit by a little hole at the side. The main arrow went this way, and there was a little window here that light came in.  So I got a second spectrometer and got it to shine the same colour light on to the cross hairs as I was looking at.  Now my eyes could focus on one colour and I could get an accurate measurement.   But that took me another week to research that. But I felt I was learning something.  

Well the professor, after his one embarrassing question, said ‘Okay, that's alright’, and that was it.  I don't think I impressed as being a particularly brilliant student but  I graduated.

[DP]:  Yes, 2A. 

[MM]:  Oh I was very happy with myself.   A friend of mine had gone, not in the army, but had gone straight to university.  And where had he gone but to Manchester University and he had written to me. He said, ‘There's a funny machine up here called a computer’.   Of course, Manchester University was with Ferranti anyway, so that interested me.  So when I was looking for a job I looked for a job with computers.  I mean I had no idea what I wanted to do.   

I don't think we did anything about computing. Well, Birmingham, where I was, was much more interested in nuclear physics.  We had synchrotrons and I could tell you stories about the EC’s atoms and machines with accelerators.  A synchrotron used to discharge electricity for one second to a massive coil and you could produce a terrific field and control the spinning of charge particles round the synchrotron.  And then it used to take a further nine seconds to recharge the electricity and then it would do it again.  And this produced a magnetic pollution across the entire campus, because people had quite delicate instruments and everywhere, every ten seconds it went 'boing', and  just as the instruments settled down the ten seconds was up.   Because the synchrotron was under the enormous forces the coil became loose and they had to close it down, and there was a kind of cheer all round the campus because everybody else could get on with their experiments.

[DP]:  So, tell me, when you looked at the possibility of studying computers.

[MM]:  My friend just wrote enthusiastically and it got me interested in the subject. So I applied. I can't remember who I applied to apart from LEO.   But LEO accepted me.  And they were in Minerva Road. Well Ferranti would have been the company to go to.  Elliott Automation, they made paper tape readers, which we used at LEO.  So I went to LEO.

I went to their school for six months.  And what was amazing the school was attached to the factory, so there was a door.  And we studied the computers completely theoretically. There are the computers out there but they never took us round and showed us except on one occasion. I realised they were busy in building computers.  So we went through this. I remember particularly Doctor [David] Pinkerton, the brains behind the whole thing, explaining to us how they did temperature compensation of the memory store.   

Well the memory store on the LEO II  was mercury tubes, some sixty four of them.   They were in a big box about as high as me, so that would be a metre and a half high and a couple of metres long. In the side were these tubes which we were told  ‘Do not move, they're delicate', but occasionally we had to.   Of course, with the temperature change, the tubes would get longer or shorter. So this would affect the way they worked and they had temperature compensation.  So the cabinet was cooled or if it wasn't cooled it was certainly insulated.  Dr Pinkerton went to great lengths to explain how they were doing this,  that and the other. The temperature compensation was absolutely unbelievable and we were all very impressed - it was a load of crap.  

Our machine used to give up at three o'clock in the afternoon, before we installed any air con, 
because it got too hot.  And we had on top of the machine a control, I never did undo the box, but there was a knob that you could turn which changed the frequency of the thing, and we actually controlled things. As the temperature went up we turned this knob on there.  It would blow this automatic system, and nevertheless, it would give up. It would get too hot and you would just close the machine down and wait.  And when it got a bit cooler things would go back and run.
 [The box was know as the vault. It was insulated and had a fan in to circulate the air. The tubes would still vary in length and Pinkerton’s clever solution was to vary the basic machine oscillator using a reference tube one pulse shorter. It was the first thing we commissioned on each machine. The vault also contained tubes for the central computer:  Accumulator,,Order, Sequence and Input/Output buffers.]
 
[DP]:  Were you applying for a specific advertised job or position?

 [MM]:  I was clearly applying to be a computer engineer. I must have done an aptitude test.   I did a training course for six months and my one memory of that course,  or just after the course, is going to the card reader, and I never, to this day, know what happened.  I was manually moving machinery for a card to go into the card reader.  I can remember turning the crank, which is normally motorised, to do it slowly.  And nothing came out. And to this day I don't know what happened to that card. I put it in and it just disappeared - a punch card.

 LEO I, I only saw briefly. I'm essentially a LEO II man. I'd been to Cadby Hall and I must have worked there for a brief period.  I was a bit surprised that the flip flops on LEO I were hand tailored.

 [DP]:  Tell us what a flip flop is.

[MM]:  A flip flop is an electronic circuit. It will be in one of two states. It could be flip or flop.  And these two states are equivalent to zero or one in binary.  And I think they were on all three machines but I'm not very good on the LEO III. Any time you actually wanted to look at   some information, it would be recorded on a set of flip flops.  So if you had a thirty nine digit word, as we did in LEO II,  you had thirty nine flip flops.  And each flip flop could be set to zero or one and they consisted of six amniotic valves.  And the circuit, which was not unique, had been invented by somebody in America.  And when you sent a pulse to it it flipped.  And when you sent another pulse in the same input,  it would flop and go to the other way round. So you could move it backwards and forwards. We had lots of them, not just the current order, (programming order executed, they were important),  but lots of places.  And he would also use flip flops when he wanted to hold information.  

There were some, presumably because of the shortage of money, when they designed LEO II, in the few places it took. It only came to me in my dreams many years later, but long after I'd left LEO.  They save money. To start off with they couldn't always carry out a full logical instruction in one go. It required too much technology or required too much information and you needed two addresses.  So there were a few commands which required two instructions.  You did the first instruction and the results of the half cooking, you saw and then followed it with the second instruction and that gave you the answer. This was unfortunate because you needed to store the results for the first one before using the second. Now the designers said, ‘You always do these two instructions one after the other.  You always do instruction number one and then number two’.  But, of course, the programmer is much too clever for that and they'd do instruction number one, and then go off and do something because the answer was stored, and then come back and do instruction number two, for whatever reason.  

Unfortunately, because flip flops are moderately expensive, these status flip flops were sometimes used for several purposes.  So the programmer thought he was doing the first half of instruction A, and then later the second half of instruction A.  But unfortunately in between the two he did another instruction which used the same status split, so the results were  wrong.  And this frequently happened, and it only came to me one day, for the price of one flip flop they could easily have put that second flip flop somewhere else and let somebody use this one.

 Because if you had been following these theoretical rules then it would have been safe.  But if you separated the two halves then of course you could  play games in the middle.  So that was one of the places where lack of flip flops let us down.  They just saved money. However the flip flops were fairly good circuits. They used to break down occasionally but we had lots of them.  On LEO I, the engineering tolerances weren't very good, and people didn't understand them exactly.  Individual flip flops had been modified electronically so they'd have different components in them.  So if you took out this flip flop from this place there, you damn well put it back in its slot because it wouldn't go anywhere else. Even though, theoretically, they were all the same, they were actually tailored because things were very edgy.

They had rules about how many logic positions could be placed, one after the other, before you   put an amplifier in. Amplifiers cost money of course.  And I remember rightly the basic pulse height on LEO II was twenty volts.  And you were supposed to keep it above something like twelve volts.  But we found places in the machine where there were a series of logic data. Every time you went through a logic data you lost a bit of voltage.  And we found the places where these rules might have been forgotten so that there were really very tiny pulses being processed. Of course, the smaller the pulse the more likely you could get interference.  So we used to, occasionally, wire additional amplifiers and hang them from the rigging or find a space you could put one in to correct these bugs.

 We left school and I was assigned to LEO II/5.

[DP]:  Which was where?

[MM]:  In the workshop being built. And that was very interesting because that was the first machine with magnetic tape on it.  LEO had tried on LEO I to use a wire recorder.

Well if you imagine what a tape recorder looks like.  Instead of having the tape you've got steel wire.  And wire can change its orientation along its length - twist backwards and forwards. So the kind of precision they wanted they couldn't get so it was not very successful.  But LEO II/5 was the first one they were going to put magnetic tape on and I helped to get it working.

[DP]:  So no more paper tape or, or cards?

[MM]:  Oh, we had all those.  I’ll explain what the peripherals were. So, I left the school and went in and saw these wonderful machines shining.  I think I'd seen them one day only during my training.  I was assigned to this thing and I stood back and watched and it was easy. I did bugger all and stood back and watched while the machine was warmed up in the morning. You had to turn the power on and then run tests on it.  I had no idea (wet behind the ears, I'd just left school).  But the guy was very nice. I don't know who he was. He told me not to tread on the equipment and he got on with the job and that was perfectly alright.  

So the first day passed and I must have some things, who knows what. After the engineer started the programmer got access to these machines because they're generally in use at that time.  Day two I came in and whoever it was, wasn't there. So I thought he’d be in soon.  And people started umming and ahhing because they wanted to get on the machine,  ‘When are you going to run tests?’ 
 I had this absolutely frightening thought. 'Here's me and this million pound machine'.   I was a young guy with nothing, from nowhere and, I was absolutely petrified trying to get this machine to run on the second day.

  Well, we got through the tests somehow.  I probably bypassed most of them and we handed it over. So on the second day it turned out the guy had gone out and got drunk.   He just didn't come in because he was pissed out of his mind.  He came in the following day. So that was my introduction.  But I got used to the machine and in particular the magnetic tape, because they knew how to build the machine.  

I should explain to you about the peripherals. There was the computer and there was a magnetic drum from Ferranti.  There was magnetic tape, paper tape and punch cards.  Programmes were written, in fact all new data of any type (it could be commercial data or programmes) were punched into paper tape - five hole paper tape.  The programmes were assembled on the computer by a very simple piece of software and converted into binary instructions which were put on to punch cards.  So punch cards were a method of storing data but the initial data always came in on paper tape.  

You couldn't put data in on punch cards. My very first program failed (it was a very simple program - probably had one plus one equals two, or something like that) because it extended over two cards.  The problem I had was I'd forgotten to give the instruction on the first card to read the second card.  So my little program didn't work, it was my first programming failure.

[DP]:  Were you required to take a programming course?

[MM]:  No. We had test programmes and as far as we were concerned test programmes were a bunch of cards. You put these things in and they ran and you looked at diagnostics.  But not what was in the programmes. We knew vaguely about programming    because we were engineers and we knew what the instruction did indeed.  It  had been known to wire up the next instruction, a modification of a standard instruction.  And on one occasion when the programmer was coming to discuss something, I said, ‘I can do that. We set that at seven'. We could do a sort of change. Changes we made to wiring were done in white wires so you could tell where we’d made changes to the machine wiring.  

So there we are in the factory, testing this every day and fairly some came to getting the magnetic tape to work.

 And we had a team, three of us, working on it.  Because I'm safety conscious, I got the guys to say, if it was stripped,  was open,  no covers on it,  we would never have the power on unless all three of us were in front of it. Never had it on because people are quite careless and we actually nearly killed someone.

Magnetic tape. Well there were two drives,  one above the other.  It stood about two metres high, because I had to stretch up and about two metres long.  And the tape reels were 2,400 feet.  And there was a reservoir, a fracking reservoir. The tape came out of that and the little reservoir was required because you didn't want to break the tape when you started or stopped  the reels or the tape going through the head. This reservoir was kind of a spring. What was amazing was occasionally the server motors which kept the reservoirs full, wouldn't work. So you could get 2,400 feet of  tape into one reservoir. You won't believe it but you really can.  If something would go wrong you'd look up and although it didn't often happen there were these two empty reels. Then you'd look and there was this brown, because it had a glass cover, column and that was the tape wrapped.  
[ The Decca tape decks had glass-fronted vacuum reservoirs to buffer the high stop/start speeds given by the precision, beautifully engineered vacuum capstans. There could big snarl ups if a faults developed in the vacuum system: Motor failure. Broken belt. etc. ]

[DP]:  What happened when that happened?

[MM]:  Well you've lost your tape. I mean it was bad news.  Well the tape was peculiar - the technique for reading the tape. They didn't want to damage the magnetic coating on the tape so the plastic back of the tape ran over a wall and the reading head was above the coating on the other side.  But above the coating so the coating was never touched. Now unfortunately, magnetic field falls off by the inverse square law of the distance.  And the gap between the head and this wall,  had to be very, very precise.  So precise that we had gauges. (They were used in cars for the ignition, I can't think what you call them).  Anyway, these gauges usually came in different sizes.  But this one, because we were on a very precise gap we needed a tapered gauge, not a flat bit of metal. And these are quite thin bits of metal.   
 [Feeler gauges. ]
 
[MM]:  Springy metal.  And then there was one remaining person in England they found who could make these tapered gauges.  And we had to set them so that half way along the gauge was the right distance.  So you can imagine quite an accurate measure.  But of course if you pushed it too far it got jammed in the head and it got crumpled, because this was a very thin metal.  So we used to get through gauges at some rate.  And eventually this guy died. Well this was a disaster, because what were we going to do? These things need to be adjusted from time to time.  So they changed  the mechanism and went to contact heads.  And the contact heads were good because you didn't have this variation of distance. 

We did have another problem with the tape.  The magnetic tape probably had been designed for contact heads in the first place.  But this thing was running over a little wall over the back, and so bending the tape with the coating outwards, and this stressed the coating and so the coating would just fly off as flakes.  So another disadvantage of using this out of contact head was that we had real problems with supplies of tape.
 
I remember me or a colleague talking to a representative of the manufacturers and taking the tape over the edge of a cupboard so he could watch.  And we ran the tape like this and the tape came off, hence the floor was covered in coating.    As it went over the metal edge of the cupboard door, you can imagine, it just stripped off.

[DP]:  What was the response to that?

[MM]:  I think the tapes had just been made for contact heads where this wasn't a problem.  And so, they just weren't made for the kind of stress that we were putting on them. Eventually  the machines were modified by  Ferranti or Ampex or EMI, I think.  

[DP]:  And you mentioned the security aspect and the safety aspect because the covers were open for long times.

[MM]:  Well they were open.  Very high volts, you know. We were just worried about people touching things or forgetting where people were. 

To give you an example of the factory, an exhibition was coming up so they took a small part of electronics, a frame, perhaps two or three metres long.  And they got it ready for the exhibition, so it was all being beautifully polished and an engineer touched something while it was running and got thrown across the floor of the factory. Luckily he was alright.  And as young engineers,  we just said  ‘Oh he's still breathing. That's alright’, which was very unfair.  But I remember this poor guy.  Two people had been polishing this thing and getting it really beautiful and straightening it. This thing he touched caused the capacitor to explode because they're sort of a layer of wax paper and metal in a big roll.  You can imagine this splattered all over the equipment.  So not only had the guy had the shock but he was faced with this  completely filthy equipment which had been getting ready. So that was a day in the life of our factory.
[This sounds like the incident where George Manley got a big electric shock. The main contactors had large lever handles which could trip the machine if somebody knocked one walking by. The handles were therefore removed and to switch the machine on and off you operated the lever mechanism with your fingers. That’s how George got his shock. Thereafter the levers had to be in place. ]

 Meanwhile I was busy helping to try and get this magnetic tape to work.  We were at some disadvantage. The units were coming [Decca].  The tape units with these double drives, one above the other.  And there were circuits which we had no diagrams for. We had very little information.  Plus we had our own circuitry, so there was a combination of the basic drive plus the circuitry that LEO designed.  

And the LEO stuff had faults in it so my memory is of identifying a fault and getting the components out. We must have had four drives,  four sets, so two drives here and then another two there.  And if we found a component wrong we had to take them out of all of these things. If it was a member of the heads, eight channels, so eight channels for each head, so you got all the amplifiers or whatever it was, cycled to the factory to be repaired, because we changed the components.  And when they came back again, by that time we’d found another fault, so there was continuous cycling of equipment through the factory.  And I remember it didn't seem to be a very efficient way of doing it.

[DP]:  How many of you were working on the construction of LEO II/5?

[MM]:  Well I remember three of us working on the magnetic tape drives.  So there were three of us there; and then there must have been others - probably about five.  

Another day, we had a very alarming thing.  We had our cabinets and the power for the LEO II was coming off a motor generator set.   I have no idea why they had a motor generator, it's lost in my memory.  Anyway, it went through this and it was three phase electricity coming in.  One day there was a terrible noise coming from the motor generator system in the factory which was stuck over there somewhere.  And the current and voltage meters on the cabinets were doing windscreen wipers as this thing was going berserk - backwards and forwards. I think it was all out of phase and we’d lost a phase of electricity. I thought something was going to explode so we corrected that.  We turned the  power off.

And we had a problem because these cabinets had a handle sticking out because it's the four rows of cabinets going backwards.  The power cabinet is on the left and they had handles  coming out - about thirty centimetres - for the switch.  There was a slot in the front of them and handles through and we kept knocking into these. This would just infuriate everybody else because the power would go off and, ‘Oh bloody hell. What happened?'  So we used to unscrew these handles - that wasn't good.  Another time when we had a power failure we couldn't turn the power off.  And it turned out that the clever designer needed electricity to switch off the cabinet, which is not very clever!  Normally you need power to switch on but not to  drop the thing out.
 
[DP]:  Since then, subsequently modified?

[MM]:  Yes.  All these sort of things were corrected. And we had the power cabinet, I remember sometime when I was running II/5, because I became maintenance engineer for it and Whiteleys.  It was a Bureau machine.

    We had a power failure and I finally tracked it down and resolved it.  And it turned out we had four main voltages in the machine.  And the designer, quite sensibly, got four cabinets. so you'd put 250 volts in there, and 100 volts in there minus 10 volts of air, so that each cabinet had its own voltage. This was alright except only one of those four voltages was heavily used.  The other three were potentials that were used for the valves but they didn't have any current going through them. They were just a set of potentials.  So one cabinet was crammed full of equipment and got very hot because the air couldn't get past it. 

 The other three cabinets were this vast cabinet and two little boxes at the bottom because all they had to do was produce a reference voltage.   Needless to say, one day a transformer, or choke,  just burnt out. It just didn't like it.  And the problem there was the designer was just being tidy - not thinking.

I came in as an apprentice.  I worked in the factory and then I went to Hartree House, Whiteleys. From the first day, I was there  helping to install it.

[DP]:  So how long were you actually working on the magnetic tapes on II/5 in the factory?

[MM]:  I guess six months. And then it was complete - we got it working.  

[DP]:  So then you had to take it to pieces?

[MM]:  In fact there was a team of people who did that. And then it really  was taken to pieces and re-assembled on site. Well  there was a team of people  who were specialists in moving these things.  Everything had to be numbered.   We had a very good guy who organised that - an ex-army guy. [Arthur Clements and his Model Shop team.] 

And then we had to test it and get it ready.  So we got the machine running and T.R. Thompson, the boss, came round. Because it was a Bureau machine they placed it on the floor and in the lino on the floor in front of the control desk was some pattern of LEO.  And the engineers were sweating blood to get this thing down. T.R. Thompson came in and noticed that the design didn't line up with the machine, and being a typical director he said, ‘Could we move the machine over?’ and there was just silence in the room.  I mean, all they've got to do is lift the lino up and move it across. It's a trivial job. 'No, don't move the lino'.  But someone quietly took him on one side and explained to him it was perhaps a better idea to get some floor lights, which really upset us.

 My other story in the factory: I didn't really know about LEO II/5. We had a problem one day and I sent one of my colleagues, probably Mike Malloy, to the end of the aisle. He went to the end cabinet and opened the door, looked inside and said ‘It's alright, the indicators are on’.  I thought ‘We don't have indicators in the engine’.  And there was a loose connection, a power connection, the knob was bright red hot.  

Dead animals, we used to get the odd rat electrocuted.   
 
When we moved to Hartree House, after another six months or so, I became in charge of it.  We were all young.  I had a team of something like  eleven people to keep this machine running 24 hours a day.  We had a lady valve tester. What we used to do with valves is: we routinely  took valves out of the machine and put them into the valve tester and tapped them. Somebody later pointed out that that was not a good practice because tapping valves does cause them to fail. So we got lots of failures, most of which were self-induced.  
[Valve policy changed over the years. Routine replacement and testing with valves that did not come up to specification or were shown to have intermittent problems with the ‘rubber’ hammer (a bunch of elastic bands around the butt end of a pencil!) being rejected. This was known as ‘egg-boxing’ which involved pre- preparing a sectioned wooden box with tested valves for the next unit to be tested. Lenaerts research showed that different types of valves had different average lives- three, six or nine thousand hours. But then replacement then was abandoned and valves were only replaced when they failed or were not weeded out by the regular marginal testing. More problems were induced by regular intervention than were prevented. ]
 
Another famous story was LEO II/4 at Ford had a catastrophic failure. The machine was stopped  and very innocently the engineer in charge, the Ford engineer, (LEO trained him but he was a Ford man), sent in a report saying ‘This valve’. They'd traced the failure to this valve.  I knew this guy very well and I thought ‘This letter looks too innocent to me’.  So we put the valve in the valve tester and there was nothing wrong.  And then we read through the full report and realised the valve tester didn't test all the components inside the valve. There was a mistake in the valve tester.  And he was actually complaining that the LEO equipment, as provided, was not safe.  And  I suggested they quite rightly turn it down. I suggested, knowing who had sent this ‘innocent’ letter, that we would write back and say, ‘I hope your valve testers are up to date following all the modifications we provided' Because only this one went to the valve tester, but the people didn't think that lying was a good idea so we didn't do that.  But  there was some negotiating. I don't know what actually came out of it but Fords were actually very cross that the equipment had failed.  

The other story I'd like to talk about LEO II/5, because I'm succeeding on this one and a couple of stories. We had a visit from Prince Phillip and that was fun.  And the boss had his office converted into Prince Phillip’s portable toilet, because he can't use ordinary toilets - he's much too elevated to do this.  And ran the demonstration programme.
[For the Prince’s visit to Minerva Road an Honourable Artillery man was installed in the Gents complete with clothes brushes.] 
 It was a computer and this was quite something.  And for example we had to have this flag flown from the mast  at Whiteleys. And we refused because we were just still young men,   ‘We’re not here to fly bloody flags’. So they had to have a special man come in to fly the flag for him.

 And there were two incidences,  one was Stevenson who was my boss. His office was converted to the toilet. He told me to collect the frames out of the toilet,  which was down an ‘L’ shaped corridor, round the glass. The computer room was behind a glass wall and you could watch. So he walked down the corridor into this private room and followed by an entourage, a  snake, and I think he went down there to pick up his hat. He then turned round and came out and walked back down the same corridor.  Now this was terrible, breaking protocol, because he was now physically close to people who should be at the back of the queue.

 They were all kind of leaning backwards, ‘We can't see you, carry on Sir’. It was very, very funny.  We weren't disruptive. We were just very full of ourselves. We were good at what we did and that was it. 

 There were three groups of people and they lived three different lives. There were the engineers, the operators and the programmers. There were lots of other people of course but they were the three groups of people at Hartree House.  They really lived separate lives, I can tell you stories about any of these groups. 

Anyway in the demo I'm sure he didn't understand what was going on.  I remember taking my parents to see the computer and it was kind of ‘Yes’.  Well, I mean what do they see?  Blue cabinets?  I can't imagine what it meant to them, and certainly I couldn't imagine that being in my pocket as it is now.

[DP]:  But there will have been a little calculation or something going on?

[MM]: All the indicators lit up and then we’d have printed out something.
Anyway we had one of our operators standing to attention. He wasn't actually standing to attention, he had his back to the magnetic tape control unit because we knew that, we kept having failures. And each time it failed with his hands behind his back he was pressing the reset button.  And so the demo went through beautifully but thanks to that, there was some great fondness.

[DP]:  But now, as the computers were getting faster, the processing was getting faster, what about the peripheral equipment, was that developing at the same rate?

[MM]: No. The peripheral equipment was awful. We had two printers. Well the card readers we didn't have much trouble with. These had been going for fifty years and they were solid things. They weren't very fast. Much faster ones came round.  The paper tape was a problem because it goes all over the place. It gets torn and our paper tape readers, the Elliott ones, were just readers. They had no control over the paper tapes. It came from a bin, went through and went into another bin.  

Subsequently I worked for a company called Honeywell and they had the equivalent of magnetic tapes. They had reels, and that was that.  Printers were terrible. They were made by Powers-Samas and they were mechanical and so everything about them was mechanical.

[DP]:  Why do you classify them as terrible?  Because they were slow?

[MM]: No, they broke down all the time. And they were very difficult to manoeuvre.  They needed target, they were kind of a dot matrix.  What they had was a bar with 120 positions. It was 120 wide with 120 wires, each driven by its own little magnetic coil.  So the wires came down and then there was an ink ribbon and the paper below it.  So when you wanted a dot, the computer fired off the wire which made the plonk and the wire moved backwards and forwards just going through one of them - the dimensions of this thing. And the paper moved forwards which gave you the other dimension. 

 So this was a fairly clever idea, that's the good news. The bad news was the paper movement involved. If I remember rightly, five bits of metal, jointed, which were only secured at each end and wobbled in the middle. It was the most amazing piece of engineering I've seen.  And this thing kind of wobbled around and wobbly equipment is not a good idea and the speed of movement was  quite important. 

 But far worse than this, we needed to generate tiny impulses and when we would fire off these wires, there was a big stack. Instead of doing it electronically, it was all mechanical because that's what Powers-Samas were.  So there was a big stack of discs, and they had little metal inserts in them for each character. So each disc in the pile represented a different character in the alphabet.  So they were made and this thing rotated at whatever speed, but they had a commutator which had a big fat brush about the thickness of my finger, which was on a plastic block.  This you could rotate around the cylinder and you could change the timing when it would fire off.  And this had to be timed in micro seconds, and all you had was an ordinary bolt holding this thing in place.  I mean if it got out and moved it was absolute hell to set back. There was no micro meter or anything. You're talking about big hefty mechanisms moving a minute amount. I thought it was dreadful.  

On top of which the pulses that actually fired the thing came from valves, of course, because that sort of thing, was designed to be held. They were power output valves, so it could handle quite a lot of them, and they were designed to be held vertically. But of course our designers had put them horizontally and so, under the heat stress, every now and again the components inside the valve would collapse and it would catch on fire.  

So  I remember one day when they caught fire,  the engineer went up and, because they're all separate things, pulled out a unit which was burning whereupon another one set on fire. And we were being very helpful (the crowd that were there). We were saying ‘Come on, hurry up Harry.   There's another one gone’.  And he was juggling with his feet because they were setting fire to the floor at the same time as pulling these.

  So printers were so unreliable I was offered a third printer. We had two and I had one mechanical engineer to look after them. He spent his entire life with his head inside a printer. He never looked at the card readers as there was no problem with them.  And I was offered a third and I said ‘No, I don't’. Two were no good. Three were just going to be worse. Printing was a real problem. 

We did one amazing thing with this printer: It was full of compromises that should never have been allowed in. From an engineering point of view it was pretty poor. One day the operators complained that the printing was upside down.  So, if you can imagine a mirror, it had mirror- printed this line.  So right in the middle of some printing  was this line which was upside down.  And they gave it to us.

 Finally it occurred to us that it was the paper movement.  The paper was on tractors with holes with pins to pull it through.  But they were only in the front.  Now when the paper leapt forward it was kept in place by a spring roller at the back - a roller that sat on the top which held it back and kept a tension on it.  And what had happened was the paper had jumped forwards too much and then started printing. The roller had gradually pulled the paper backwards through the printer, taking out the slack while it was printing.  It looked like an absolutely perfect normal line of print, but upside down.
 
But my great success on II/5: We used to have monthly meetings of the engineers in charge of all the machines.  We came together to discuss what was going on.  And Lenaerts used to attack me.
 
 [Ernest (Len) Lenaerts. Seconded to Cambridge and one of Pinkerton’s research team. ]

Lenaerts who was in charge was a very, very bright guy.  First of all my engineering logs never added up to the time that the machine was down.  And I explained that this was due to the number of faults we had on the machine.  I said ‘If we had less faults we’d have more time to write them up’.  And I felt that criticising our administration instead of resolving things, I had some particular problems with that.  

We, for example, had a relay in the magnetic tape drives - gold plated contacts but these were going wrong by the dozen.  I and others complained bitterly at their monthly meeting and said ‘Look, just solve this one problem. There's something wrong with these magnets - these relays’.  And of course there was something wrong. They were German, Siemans and they transferred their manufacturing from Germany to Wales where they had a new factory.  And they'd missed out one manufacturing step which is tensioning the springs inside the relay. So the thing, instead of being pressed on the contact just kind of floated. That's why they kept going wrong.  But it needed a great effort on our part to convince these people to ‘Stuff your bloody records. We've had hundreds of failures of this relay. Just solve this one problem’, which they eventually did.  

But I used to get beaten up as the machines went out: II/6, II/7, II/8.  My computer, II/5, was peculiar.  We couldn't read and write on our magnetic tapes simultaneously because we had installed a modification to the machine that said you could read or write but you couldn't do both simultaneously.  And that machine was running at half speed.  Nothing was wrong, logically. The machine just waited.  I used to get beaten up every month because nobody else had this problem. I only got time on the machine when it broke, because the machine was used most of the time. So tracing this thing down was absolute hell.  But by being detectives we eventually found out what it was.  The magnetic tape looked very much like this box here but a bit wider - tall cabinets.

 [DP]:  Like a wardrobe would you say?  

[MM]: A part of a wardrobe. A tall bureau. And it had lots of cards slotted in. And the designers liked to be tidy, just like the guy had been with his power cabinet.  And he carefully placed the read circuits on the left hand side. There were eight channels on the tape, so all the read circuits for the amplifiers to pick up the signals were on the left and the right hand side. He had the right circuits doing all the power drivers.  And very cleverly the first amplifiers for the read circuits were adjacent to the power drivers for the right.  So when it wrote, the magnetic signal just jumped across from one card to the adjacent one. It took us a long time to find this.  Eventually we realised this was the problem. We wrapped the read circuits in tin foil temporarily, and whoopee, we could read and write simultaneously.  So I phoned up my pal at II/6 and said, ‘Could you go along to slot 33 and tell me what you find? Just go and have a look at it’. He said, ‘There's a bunch of aluminium'. Designers in the factory had discovered the fault on II/6 and repaired it and they never issued us with a report.

[DP]:  Wouldn't that have come up in your group meetings?

[MM]:  No, the engineers wouldn’t know that it was there because  a plate there was just part of the machine.   It's not obvious. And you've never seen people move faster.  Peter Mann, who was the engineering manager, organised transport and in a whisk, I had a little metal plate which slid into there. Forever afterwards nobody ever mentioned my name again at the monthly meetings.  

[DP]:  But  as the LEO’s were coming off the production line they were being improved and modified?

[MM]: Yes. We were doing back modifications. We did back modifications, for example on the magnetic tape. Transistors were just coming in, not just diodes. And I hadn't been an engineer for a long time. We had lots of diodes. Transistors were coming in in a big way and LEO came up with an idea. We were all given cross talks between reading and writing between channels to insert the transistor in each of the circuits. There were loads of these.  

 We had three magnetic tapes, each doubled, that's six drives. Each one had eight channels.  So we had a lot of wires coming into this control cabinet.  And we inserted a transistor into each of these circuits so we could hold them down so that we wouldn't get any interference.
 
And to make this modification they sent an assembler, a technician from the factory, who was going to do this.  And he did the most beautiful soldering you've ever seen at the rate of about one joint an hour.  Beautiful works of art - absolutely magnificent.  And of course we’d had to close down the computer while he was working and we had the weekend to work to. 

 And it took me a long time to think, but at one an hour it was going to take us at least a week and I'd be out of a job. So I thanked him very much and said now I understood what needed to be done. I'd watched very carefully how he did the job. I thought we would be able to do it and he didn't need to work any longer on this one and he went off.  

And we set to work, just a fraction faster, and we made one mistake in the whole lot. I was really proud of the lads.  We worked over the entire weekend and something wasn't working. It was just one joint that we hadn't done correctly.  

[DP]:   So just tell me a bit about Bureau life because Bureau life means that you have a series of customers coming in, buying time.

[MM]: Yes. They wouldn't come to us. I went to see them. There weren't people lining up in the streets.

 We had a contract with British Rail.  British Rail charged for freight, according to the distance between stations and they didn't have a good set of tables for doing this.  So we had a program that ran in the background, which we used to calculate the distance between every station in the United Kingdom. There are quite a lot of them and the computer was so slow.  And this is a very nice program because it will work so long. Then when you run out of time, half an hour, two hours, you flick a switch and it will punch out, how far it has got.  And then the next time you have some spare time you pop it in.

We used to do the payroll for the air force (or it might have been the army). Things would be delivered and the operators would run this.  We were only engineers.

[DP]: You were just waiting for something to go wrong?

[MM]: Yes.  I can tell you one story about Durlachers, who are stock brokers. In those days [before Big Bang] stock broking was all done on the floor of the stock market. The brokers would ask for a price of stock and agree something.  They'd write it down on a slip, but they didn't exchange contracts on the floor.  The next morning there was a call over at ten o'clock when a representative from each of the brokers would turn up and say, ‘Five thousand of so and so’, and  ‘You sold it to me’, and they would exchange slips at this point.  And you can imagine, there were hundreds of these things going on.  

So they sent us all their slips overnight. We then produced a listing. This was much better.  It was a straightforward listing.  One day we’d had a problem with the computer. I'd not been able to repair it fast enough, and a representative came down from Durlachers - very smartly dressed. You can imagine how we were dressed.  And very gently explained to us the system of ‘call over’ and how yesterday they turned up and the guy didn't have any slips because we had them. He didn't have any paper because we failed to print them, because we hadn't done the processing. So he's standing there, completely empty.  The stock market is a big old boys’ club. Everybody got their junk stock out and said ‘You've bought five thousand of these from me’ but he could say nothing.  So they dumped all their crap, and he said,  ‘We lost a 150 million’.

[DP]:  Did he hold you responsible?

[MM]: Oh, yes.  But it was all very nice. ‘Would it be possible’, he said very, very gently, ‘if we could use our best efforts to get it there by ten o'clock in the morning because it really would be helpful’.   It was just charming and it was a lovely old school.  He wasn't admonishing us - he was just explaining. A little bit sad.  But we did take it to heart.

[DP]:  So, as the engineer on,  II/5, you would be essentially sitting there waiting for something to go wrong?

[MM]: Correct.  But we would be repairing circuits. Do you remember we took the circuits off and we had spares?  So that the whole unit, the flip flop was something like a ball I could hold in my hands.  The whole thing was in a vertical unit, which is about a metre high, which came out.  And so if you had faults in these you’d have to repair them at the workshop. We had a small workshop. Sometimes we did that in situ.
 
Let me give you one lovely story about the magnetic tape.  We got all the magnetic tape together but it wasn’t running very well. In Hartree House,  we still had this rewrite issue going on.   One day my boss said, ‘Come on Mike’. We got our diagonal clippers, and started to remove components from the magnetic tape circuitry and checking the tapes.  The designer   of the hardware was very worried about things happening so he'd do something. He'd have some logic which would say ‘Did that work?’  And then he'd do something else and the logic would say, ‘do you need time to carry out a logic test?’  So there was a conflict because he had all this extra circuitry but there was no time in place for it to operate.  My boss decided it was just ridiculous. Nowhere else in the technology do we have these checking circuits. If you sent a pulse from A to B, you sent a pulse from A to B. You didn't have a thing at the end that said, ‘Have you sent the pulse?’ So we just chopped out this equipment and it was actually remarkable.  We were chopping out equipment, resistors and circuits and there was a pile on the floor and suddenly the tape ran into action. It just started going beautifully.  All because the designers were over-cautious.

If there had been a step ‘do’, and a set ‘check’, and then a step, it would have worked, but he didn't have any steps. He had a step ‘do’, and then another step ‘do’, but there was a check in between them.  So that was quite extraordinary and, it was the only time we ever attacked the computer.

[DP]:  So was life getting easier for the LEO II engineers as the machines were coming off the production lines? Did the man, the next one up,  have an easier life than you?

[MM]: I think he must have done.  We were getting  corrections coming through from the design department because whenever a fault was found in any of the LEO II’s it would be passed down the line.  I guess we made our contribution, but I think many engineers did in different places.  LEO II continued with fairly similar technology all the way through its life. Its basic structure was the same because design interest had changed to LEO III, which was the big thing that was coming along. That had big cabinets with back to back boards, facing each other, to try and increase the density of circuitry.

 I did some work in the factory on LEO III.  I worked on the pulse amplifier.  

I’ll tell you how I ended at LEO II/5.  I got involved in marketing towards the end.

So, when I was there we had pulse amplifiers in LEO II. When the pulses got too small you put some in and increased the pulses. LEO III had a much cleverer design. I improved the design and got it to work but it wasn't my thought.  I made a very minor contribution.  So instead of amplifying the pulse they regenerated the brand new pulse. So when a pulse came in we had a pulse detector. If it detected the pulse it said, ‘What you want is a nice pulse’, and it would then create a nice, big  pulse, which would pass out.
 
  A much better design because you weren't depending on the shape of the incoming pulse, as long as you detected it 

[DP]:  Yes.  Now as I recall there would be an oscilloscope on the control panel?

[MM]: There were two.

[DP]:  And what was the purpose of them?

[MM]: Mainly to look at what was in the memory.  I think one of them enabled me to look at the memory tube, which must have been in binary. It must have been 16 rows of 40 pulses in line with the pips. But, apart from looking at the memory, you could look at the registers. Well the memory was in the tubes and this is where the main memory was.

The way they worked was that there was a loudspeaker at the beginning of the tube, which generated a pulse. This travelled as a sound wave through the mercury to the far end where there was a microphone which picked up the pulse. It instantaneously at electronic speed, passed it back to the beginning of the tube where the process was repeated.  Of course we weren't dealing with one pulse, there was a train of pulses. If I remember rightly it was 160 pulses, four times forty, intermingled as it all went down the tube.   And it was relatively low sound wave, travelling through the tube, so it could store all those pulses.  And then as they came out they were teamed up and went round again.  And that's where the memory was. We could extract one word at a time as it came out.  

We didn't process anything in the tube. The tube was just there to store things.  If you wanted to do something with it you made a copy of it in one of the registers. 

There were many registers,  twelve or sixteen registers.  I remember seven and eight, but I can't remember how many and where, the registers were.  And they held the bits, and you could process those, because those would be held in static time.

  So you could look at the registers and see what the values were. In fact looking at seven and eight was important for the operators because the only form of communication between the program and the operators was by stopping the computer. Shall we say register seven, with value twenty three, and then stop the computer. 

 So the operator could look in and see, ‘Oh, very soon I've got twenty three’, and he’d have a look, and  if I'm running this program, error number twenty three is, whatever.

  So the operators used these tubes to get communication from the computer.

[DP]:  To identify the problems?

[MM]: Well they were programmed to do this.  We used the tubes to be able to identify an engineering or technical problem.  And we were helped in this because on the left hand side we had a panel (when we were under the machine) where we could change the voltage slightly.   The theory behind this was that if the circuit was sensitive, because  the valve was dying , we shook the vultures around it really pushed it over the edge.  

So we could identify bits of the computer that weren't working well.  The curious thing was that the panel that did this. You put a wire in so the tremble of voltage came out of one place and you could stick it into any of these points.   Well we never had the proper cable to do this. For some reason the equipment was non-standard and we never got really good signals through it.  

So the power on the left would be for shaking the machine, electrically, and seeing if anything would fail.  In the middle we would look at the tubes to see what was happening and we would write tiny programs, three or four instructions, typically to see if something went wrong.

  One day we had a problem which we identified. The arithmetic was sometimes failing and it was very obvious that we couldn't use the machine because it was unreliable.  It would do something, but it would fail.  

When you put a program in to do let's say ‘Add one plus one. Is the answer two?’  ‘Yes or no?’  The wretched machine would fail for the first few cycles and then correct itself.  So the program would say, ‘fail, fail, fail, okay, okay, okay, okay, okay’. Very difficult to find a fault if before you can look for it it's corrected itself.  But clearly the machine was unusable, and I think I had the machine for three days with enormous pressure from management as we hunted down this fault.  And in fact there were two faults. A second fault occurred while the first one was happening. We did eventually cure it, because it was a real. Bloody bad news.

 But I do remember there was an arithmetic fault and the fault was such that the circuit learnt.   It changed its voltage slightly and so after repeated use it would just go away.  But I remember management being very worried and I was very firm with them saying, ‘if you've got a better engineer send him in and I am perfectly happy.  And if you haven't got a better engineer then I need to be left alone’.   But,  I can understand that people were edgy.   But I felt that the job of management was to sort out the customers.

  There was nothing I could do about it.  And we occasionally had very long faults. I had a magnetic tape fault which went on for months.  I had the  one I  just mentioned.  I had a fire, a smell of burning, and that was a real sod because the only way we knew of curing it was to locate the part of the machine that was on fire, or was getting hot. But the trouble is your nose quickly adapts to smoke or the scent, so we used to go to the windows onto Queensway,  and breathe in fresh air to clear our noses, come back into the room when you had a few seconds to track it down before your nose said,  'I can't smell anything’.  

And we eventually located it to a free standing cabinet. Actually, the cabinet contained the Ferranti drum.  And inside that was a box, and inside that box was another box, and inside,  there, was a component that was burning out.  And that was a sod to find and, again, management found that quite difficult.  

Another bad fault we had Johnny Wheeler.   There was Peter Mann the head of engineering, Johnny Wheeler head of maintenance  and then there was me in charge of the machine.  

And we had trouble with the drum. Now in the drum, you had  two motors - a very clever idea.  One motor drove it forwards, the other drove it backwards.   This is the Ferranti drum in II/5 and later.   This is a memory device.  The drum is a piece of magnetic tape wrapped round a cylinder.  So you've got a little reading head and it can read.  I can't tell you how much there was on each track, I don't know what it was. We had thirty two heads in the same row on this drum, and that was fast access memory. 

 And it was not as fast as the mercury store but much bigger capacity.  Unfortunately, the heads were sprung. The way they were secured was they were sprung to drive, to go inwards towards the drum, so the heads were out of contact. This was out of contact recording.  And when you adjusted them, somebody had been concerned about cleaning the heads. I think this is what caused the problem in the first place.  

So they thought they were unscrewing the head, but the screw was actually the thing which stopped the head going inwards because it was being sprung inwards.  So, as they unscrewed, the head went in and tore into the surface of the drum.  Well we could manage two or three of these mistakes but we couldn't manage more of these. We managed to destroy a drum so it had to be sent off.  When it came back from Ferranti they hadn't got the same coating on the drum as originally, so it's electric responses were different so it was absolute hell. This was a Bureau machine. We had to get it working.  And John asked me if I would work that night and I said, ‘It's my wife’s birthday’.  He said, ‘Mike we've really got the Bureau’. I said, ‘it's going to need a very large present for my wife’, which he agreed to.

  And I wasn't trying to blackmail anybody but I had my own priorities.  And so I went out and bought a very large bottle of perfume, took it home to my wife, and when she finished smiling I said, ‘I'm awfully sorry darling but I'm off’.

  So I spent the night.  We had very, very primitive equipment to put a new drum in and get it working.  We had to, first of all make a pulse on the drum, that was the marker.  Then we needed, on a second track, to write a timing track with let's say a thousand pulses.  But we had to end up with exactly one thousand, not nine hundred and ninety nine and a half, not a thousand and a half, but we had  nothing to help us do this.  So we had a machine that would write so we would write these things and say, ‘Bugger, it's not quite right’. So you had to wipe the track and then change the adjustment, unbelievable.  It's hit and miss. It's absolutely hit and miss and eventually after hours and hours and hours we got a track of exactly a thousand volts.  Then we discovered that our circuitry wasn't working right. These two tracks were special tracks, called index tracks. We then had to adjust the rest of the circuits so you'd deal with the coating.  We managed to do it but I was a junior on that one. I was just mystified at how primitive it seemed to be.

But you got used to it. Leaving aside these stories, I had a couple of emotional things.  One was dealing with the microsecond. The LEO II/5 voltage was a quarter of a microsecond, so quite fast. A microsecond is nothing. A human being can just about detect a mili-second. Ten mili-seconds can be detected quite easily actually.  A microsecond is a thousand times faster and I couldn't cope.  

And we were using oscilloscopes with some and gradually I was able to break my mind into two things. There was time in seconds, days, months, seasons, and then there was computer time which in microseconds. And once I'd done that I could emotionally deal with them. Once you get used to this it doesn't worry you. A quarter of a microsecond and the shape of the wave form was in hundreds of microseconds. We hadn't got to nanoseconds yet.  

[DP]:  Did you sort of switch on and switch off, yourself?

[MM]: Well emotionally I could. I built two different time frames and once I'd got used to the idea that there was computer time, that was perfectly alright.  I had a similar problem when the microprocessor came out. Though I never had a microprocessor in my hand at the time, I obviously kept up to date with what was going on.  Here's me with this vast room full of equipment, much bigger than the room we’re sitting in here, and all that can go on to a microprocessor. 

Well it's not quite true but in those days there wasn't any memory. That took me a year or two to actually, emotionally, come to terms with. I just couldn't imagine taking this stuff, even in a crusher. If I crushed all my equipment there was no way it was going to fit into that amount.

 I was with LEO for three years and  I ended up in marketing.

 I can't remember exactly, but I was interested in what was going on.  LEO III had just come out and I asked a silly question, ‘What are we doing about LEO IV?’  And this really upset them.    If you're building a line of machines you realise that the next lot need to be started on even though you are just launching this one because this thing has got a limited life.   I showed an interest in marketing and then I gave my notice.  What had happened was, an American company, was advertising for computer people. 
[LEO IV existed as a large breadboard computer at Minerva Road. It Had a magnetic rod micrprogram store and twisted pair wiring which overcame all the technical problems with the LEO IIIF, the faster version of the original LEO III. It got nowhere when English Electric decided to go down the IBM road and adopt the RCA Spectra 70/45 as System 4. ]

[DP]:  Who was that?

[MM]: Honeywell.   Peter Mann, the engineering manager,  was applying for a job. So I thought I’ll apply.  And if he doesn't get the job, whoopee, I've got a good chance.  

[DP]:  But if you did get the job would you have got a promotion with it?

[MM]: By that time I was drifting away from engineering. In fact my life was drifting   away from that and so we can briefly talk about it.  So he applied for a job, I was interviewed as well, and they offered me a job as a maintenance engineer on an electronic computer. But it meant going to The States for six months and leaving my wife behind.  I wasn't prepared to do that. A number of people were quite willing to make that sacrifice.

 I remained at LEO,  because I believe in not saying anything till I'm ready to move.   And six months later out of the blue I get a phone call, ‘Would I come in for an interview at Honeywell?’  I saw the same guy I went to see before in America. Honeywell didn't exist in Europe. He was setting up an office here.  And he went through the identical spiel that he'd been through when I applied as an engineer. But this time he said to me, ‘Would you like the job of software manager?’  Now the only programs I’d written were engineering programs. They were about four instructions long. But you know, I was a young man and I knew more about computers than he did because I'd been on computers.

[DP]:  Do you reckon you were about  twenty four?

[MM]: Twenty four, yes.  So, what the hell, computers were coming in and there were plenty of jobs.   The price of units were going up. So I took the job.  And I was very lucky. Honeywell made two computers, the 800 and the 400.  The 800 was a big machine and the 400 was its little brother - a newer thing attacking the market.  And they were really interested in the 800. We installed an 800 in London Wall, Number 1.  And, so I got the job because it was a kind of junior job. We might sell the odd 400. The marketing went exactly the other way - we sold very few 800’s because they were very expensive. We sold loads of 400’s. So I had a lovely job with no experience. You learn on the job, and my colleagues were very kind and helped me.  I eventually became a  working software manager.  Charlie was the man who hired me.

 It was like going from the dark ages to a new life. This was really advanced stuff.   The 800 would have been comparable with LEO III.  The 400 I suppose a LEO II.  But these machines were really another layer of technology. I mean it was just a real eye-opener to me. 

  The memory was core memory as on LEO III. It had an ingenious way of storing it's instructions in a thing called ‘Snake’ which was microprogramming.  If you imagine there are cables with little cores all the way down.   I was told never to touch it. I did touch it once and the machine broke and they had to repair it, and I never touched it again. 

 But it was a remarkable.   It had advanced software so it communicated to a typewriter. No more tubes anywhere.  The 800 was a remarkable machine.  It could do eight levels of multi programming.  So it could run eight programs in the hardware.   This was a wonderful machine and  this was doing my brain.  I really enjoyed that.

[DP]:     Just give me a little run down of the rest of your career development?

[MM]: In Honeywell I went in as software, did this and then became marketing. I was still interested in marketing and I ended up as European Marketing Manager. I'm not sure I did a very good job  but we set up Bureaus around Europe and I trained them all.  So I stayed there for five years and I was selling. I did a very big sale to Xerox  (Rank Xerox as it was in those days).    Rank Xerox operated throughout the world. Xerox did North America and the rest of the world was Rank Xerox.  And I sold Honeywell computers to Xerox. It was the biggest order that Honeywell had had outside the US government, so I was quite pleased about that.
  
And the guy I sold it to said would I come and implement it. So I was quite challenged with that.  I was rather foolish because there was a hole in what we promised which we managed to solve in the end.  And so I left Honeywell and worked for Xerox and set up their Bureaus throughout the world.  We decided that the computers weren't very good. The software  wasn't very good. My boss in Xerox, an American,  got over an engineer to work with me, a software engineer. He wrote a new operating system which replaced the Honeywell one.  And this guy was a genius. Roger Moore, top class programmer, world class. 

 And we were kept in here while he perfected the software.  He kept saying, ‘IBM are waiting’,   and they had a new computer and they were waiting for him to write the software for it. And here he was in London.     

It's very funny because I worked on the development a little bit, but this guy was really the genius.  And we got the new operating system out, put it on to tape, and distributed it around the world.  And then we realised to our horror that we’d forgotten something. We didn't have  a loader. The software was there but there was no way of loading it into the computer.  The reason for that was we designed the software in a very sophisticated way which said that any operating system could load any other operating system of the series. Your operating system ran down here and it loaded the new operating system up here, handed over control to that and then moved it through memory down to here, brilliant.  And I didn't do this. This is Roger Moore.  We've never done a cold load to the computer because we have always had an operating system in the computer. When we wanted to try a new version we just said, ‘Load the new one’.

  And, luckily my colleagues around the world clicked on to this very quickly. But I think I'm one of a few people who actually issued an operating system that you can't boot strap from.  And there were sixteen boot strap records. I just forgot to put sixteen boot strap records on the front of the tape.  It was just disaster.

   What was funny was thatwe got this operating system to work and the guys who worked, I'd never seen people work like this.  I had an unbelievable team. I went to my boss and said, ‘I'd like to give them some money’. This is in Xerox., ‘ … to compensate’. He said, ‘No, time off in lieu, perfectly alright’.   

  This is a new operating system the whole country will use.  So I sent them home for three months with a brand new operating system around the world and the only person who knew anything about it was me, left in the home office.  My boss went white.   I said, ‘But you said  I should give them time off in lieu but I've given them about half of what they spent’, because they were working a hundred hours a week.  You've no idea, these guys were motivated.  Well, he kind of swallowed hard.
 
I remember the main software. There were some bits around that had been written by Roger Moore. He's the kind of guy who’ll just write it himself.  And so it was in fact we only ever found one bug in it.  

And one of the things we’d done with this software was to enable us to be able to print in parallel with running the machine.  So it had very similar multi-programming as on the 400.  But this made an enormous difference to delivery.  Printing is relatively slow. When you're dealing with large companies they print for six or eight hours. So you were depending on the computer not breaking down or anything during that time.  We’d removed the constraints so that the main program ran in something like thirty minutes and then you had a tape with all your print information, which you could print off. But you could stop it and go backwards and forwards on a print.  So you got through your work much, much better.

[DP]:  Could you not have done that with the LEO machines?

[MM]:  We did rely on a new thing which LEO III had which LEO II didn't have which was interrupts.  Interrupts are a way in which peripherals can communicate with the main processor.  It's a request for the program to do something. It's either started or stopped or had an error.  So you get an interrupt signal and the main processor can then do something about it. This is what we did when we were printing.  We would send a line of print to the printer and then go back to the main processing.
   And when the printer had finished printing it would send a signal back which would interrupt it.  So we’d leave the main program, we’d go off to our print program and print another line.  And then it would come back and tell us when it was ready.  What was quite funny was we set it up for maximum speed processing,  but this meant the printing wasn't at full speed. It was a bit slow and people complained about that.  So Roger just changed the balance of how we were treating the two of them.

 We actually got less out of the computer, but emotionally people were happier because the printer was going at a higher speed.

  I stayed at Xerox for fourteen years and then I joined, CACI.  I took early retirement.  I got a job with CACI, which is a consultancy. I'd moved away from computers. I was ending up doing budgets at Xerox.  I was doing the worldwide budgets for IT and certain policies and reporting to the director of IT.  Anyway I went to CACI and there was a guy called Tony Carter,  the boss. He died just a week ago, and this was 1980, approximately, and I was a year there when Tony got fired.
 
And he set up his own company with a man called James Martin who in his time was the the guru in IT.  
He wrote books on everything about IT, in fact he got instructions and staff wrote books for him.  And he died a mysterious death off this island in the Caribbean.  So James Martin, Tony Carter and a man called Dickson Doll set up a new consultancy company and a lot of people left CACI together and went to work for Tony. We thought he was a dynamic guy, and I stayed there.  And that went on till Texas Instruments with software for automatically generating  business programs. TI took over our company and then realised that it really wasn't their job.
 It was important getting software done but it really wasn't their job so they sold off their half of the company. One half of us got fired in 1996.    During that period I'd been doing consulting all over the world.  So when got fired I set myself up as an independent consultant.  

And in 1997 I was sitting here in the evening, and a phone call came through from a colleague of mine who said, ‘Do you want a job in Australia? -  a temporary job'.  So we went out to Australia. We had a ball. We were out there for two and a half years.    We were living in Sydney, which is absolutely unbelievable.

  Absolutely wonderful time and had lots of work. My first contract ended and then I, contracted with other companies.  I worked for the supermarket chain Wilson.  

  And then came back in the year 2000,  having stayed deliberately  to see the Millennium. We had friends who overlooked the harbour, so we were on their balcony, drinking, sipping champagne watching. It was a wonderful experience.

  And it's like a football ground except it was a vast arena which is the harbour. You could hear the people round it and I had never been in a place where I could hear this, so we came back and I then was running down, by that time I'd reached sixty five,.  I then worked for Learning Tree who teach technology to adults and they do professional courses,  they're typically one week long.  And I worked for Learning Tree for five years until I thought I'd done enough. I reached the age of seventy and I thought seventy was a respectable age to finish, so I stopped.  And since then I've travelled around the world with my wife.

  I have, I have two children from my first.

[DP]:  Did they become engineers?

[MM]: No, not at all.  My wife has one from her first marriage. Her son is a doctor, a consultant, radiologist.  My two children are working,   one went into the TV industry, directing and producing TV things with her husband, and then, having done a course in sociology at University Leeds, and one day turned round to her husband and said, ‘I want to become a consultant psychologist’.  So she gave up, he worked and she went off to university, and she's now working as a consultant psychologist.

  And the other one  looks after homeless, people with mental problems.    Not very much to do with computing I'm afraid.  Yes, I've moved away from technology.

  Technology and I had to give it up and realised that you had to give it up. You're either in it or you're not.

[DP]:  But you were in it, right at the very beginning.

[MM]:  You might say that except I was at the LEO Computers reunion in the city a few years back.   And I was introduced to a guy who retired from LEO before I joined them, and that's what I call somebody who is a pioneer.

[DP]:  Nevertheless, you contributed to the development of the  British computer.

[MM]:     I've been very fortunate in my career.    And the nice thing about getting into computers when I got into computers was nobody knew anything about it so people couldn't challenge you.   I mean you did the best you could and  I made a thousand and one mistakes but managed to cover my tracks.  

[DP]:  Just tell me if, if we can pop back in time a bit, the relationship between engineers, programmers and management?

[MM]: We lived in different groups.

  So the operators, for example the operators in LEO II/5 used to play poker.  And we knew it, we came in one Monday morning and there was silence.  And we knew something had happened.  And they'd played poker the whole weekend, 
  and they’d lost, I'm sure by modern standards nothing, but they'd lost more than they expected and it was a very strange atmosphere  you got.
   And one of the operators   bought himself an old fashioned Rolls Royce, not on this, that was another occasion, he was very proud of his Rolls Royce which I should think is worth a fortune today.  

The programmers we saw occasionally.  A lovely story about Mary Blood. She did the pay for, the London Bureaus. It was a combined thing.  And, when they calculate the pay slips they pay people cash or some people do and so the last thing the programme did at the end was to calculate how many pennies, shillings, pounds, you needed.
 Cash analysis, and it was never right.  And she used to tear her hair out going through this and re-wrote the program many times.  One day she went down to county hall or wherever to follow the money, to see what happened and they looked at this. They took out the last few sheets of the payroll, put them on one side, and said, ‘well those are pensioners, we pay them through the bank’.  And so she’d done all this cash analysis but some of them never got paid cash.

  And another time I’d fiddled with the computer, as was my wont, to try and make it better.  I know what I wanted to do. Today I could tell you what to do but  because I only had limited time on the computer, because I was only supposed to be repairing it, it didn't work.    I told you that when the machine stopped they had to read the register.  Now, the current instruction was up on a series of lights which just flashed, but there was a whole line of lights.  And the silly thing about the machine hardware cycle was immediately before it stopped it cleared down the registers.  So when the machine stopped this line of lights was always black.  And what I wanted to do was to change the machine’s central cycle, fetch an instruction, do the instruction, get the next instruction, so that   when there was a stop instruction we stopped with the lights showing you the current instruction that was causing you to stop (because that would then give them, in binary,   the reference number and they could see it from across the room).  I never succeeded in doing that and I'd screwed up and I noticed that they were all running round,    you know, programmers, not getting the right results and I quietly went and reset it back to where it was and carried on. So I think that some people got paid rather strange sums of money, probably.

[DP]:    I didn't bring the form of words that we like to use at the end which is   ordinary to say that not only are we grateful, Michael Mills, for your time and those terrific stories but we would like you to understand that these are now LEO’s and we will transcribe them, no doubt you'll get a copy in due course.

Copyright
The video above may be subject to copyright of the original author and is present only on this website for non-profit educational purposes only. If you are the original copyright owner and wish to have the media removed from this website please contact us using the relevant contact email address listed on our contact page.