Transcript :

LEO COMPUTERS LIMITED  -  Oral History Project
Simon Benedictus  17 April, 2012.

David Philips: It’s the seventeenth of April 2012. I’m David Philips and I’m interviewing Mr Simon Benedictus and asking him to give his story and reminiscences of his involvement with LEO Computers from its earliest days. Now Simon we’re recording this interview as part of The LEO Computer Society Oral History Project and the audio version and the transcript will be lodged at a central archive and made available for researchers and members of the public, and I’m grateful for the time that you’re giving me this morning and I’d like really to ask you to introduce yourself? Give me your name and maybe tell me a little bit about your family history before we start talking about your career and how it led you to LEO.
Simon Benedictus:  I was born in 1933. 
DP: Indeed. 
SB: And I was always good at Mathematics to the exclusion of, I couldn’t spell so English subjects were pretty well out for me. And I went to Imperial College and I trained as a mathematician and doing the various branches of Mathematics including Statistics, which we’ll come to later. 
DP: Well, before we get too involved in your academic career can you just tell me perhaps a little about your parents? Where did they live and where did you live to start your schooldays and so forth? 
SB: I was born in Kensington and when the war started we moved to my grandparents’ house near Marlow in Bucks. I then went to a series of local schools. I went to a public school, Bryanston, but was ill shortly afterwards and missed nearly a year of schooling and thereafter it was thought better that I stayed with my family and went to a day school. 
DP: Your parents were relatively affluent were they to be able to send you off to a public school? 
SB: My father became the Managing Director of Lillywhites. So yes, we were quite affluent.     I lived with my mother because my father was in the RAF.  He had been determined to get into The RAF. In the First World War he was too young. In the Second World War he was really too old but pulled a number of strings, and was in the end in a place called Bletchley, but he wasn’t doing the sort of work you expect to hear from there. He was listening to the pilots, German pilots coming over and giving them false information because he was fluent in German. 
DP: Why was he fluent in German? 
SB: Because he learnt it at school. And he was one of these people that really learnt a language. And there’s a lovely story in one conversation with one pilot. The pilot broke off and  he said ‘I know who you are, you’re a bloody Englishman’. Reply was ‘And I know who you are, you’re bloody German’. But basically he could disguise himself, and that’s a whole different part of the... 
 Of the war effort. But after the war he was involved in the de-Nazification in Germany and joined us in about 1947 I think when he was demobbed. 
DP: Meanwhile you were at school showing some academic promise? 

SB: Only Maths.  I could do any Maths. While I was at Bryanston,  I and one other boy made up a class of two amongst the mathematicians who could really do Maths, but in any other subject I was okay but limited. Anyhow after my illness I went to a local grammar school, which was a very early grammar school, it had been a fee paying grammar school until the 1944 Education Act, Sir William Borlase School in Marlow, and it used to always come top of the list The Times used to publish. It was a very small school, about two hundred. Well, I made the two hundredth pupil when I joined. And I again specialised in Maths. I didn’t have very good advice because I should have been a Pure Mathematician, and I went to Imperial College where I did a Maths degree, but it was Applied Maths and there was various parts of it which I was not altogether happy with. I did a Statistics course but I didn’t really understand much about it. It was very badly taught. 
DP: This would be the early ‘50s wouldn’t it? 
SB: Early ‘50s. And  I then determined to be a teacher and I did a teacher training course and then I had to do my National Service, I couldn’t put it off any longer. I didn’t particularly want to  and I expected to be put in The Education Corps but I wasn’t, but marvellously I was sent to The RAE in Farnborough as a statistician. I had no statistics other than this rather bad elementary course. 
DP: Had you been in contact with computers or learnt anything about computers whilst at… 
SB: Not until we got to Farnborough. 
DP: Didn’t appear at Imperial, there was no mention of computers? 
SB: Yes, it was being built by two teachers who used to teach us. Dr Morley, and I can’t remember, Michelson I think, their famous computer. But that wasn’t taught to undergraduates, I mean that was a side issue. At Farnborough I was shoved in the deep end as doing evaluation of missile, a missile which was being tested in Australia and my job was to try and design an analysis system for use for very few missiles to find out if it was any good over the full range of what it claimed to be. 
Well, I designed such a system and I had help from two, The Statistical Advisory Unit, which was a part of The Ministry of Supply, otherwise I could never have done it. They did it in fact but, you know, I would have to talk to them in terms of ‘I’ve got three strings that I want to test for length’ rather than the missiles because they were not cleared for the security.
I designed such a system, which was a multivariate statistical analysis. And when we did it the conditions that were actually flowing  were so different from the design that we had to use regression analysis on it. Now regression was not a new technique but it involved quite a lot of computing power and a multivariate analysis. The theory was known but hadn’t been really, as far as I know, practically used and I built with the statisticians, with the computer people. I was not a computer person at all but I could specify the Maths and we built on ‘Gert’ and ‘Daisy’, which were the two computers there. 
DP: Is this at Farnborough or is this at The Ministry of Supply? 
SB: No, Farnborough. 
DP: So what were you running on the computer, what were you asking it to compute? 
SB: Machine code. 
DP: You weren’t coding? 
SB: I wasn’t coding, I was specifying. I had found out that I had to know a bit about machine code in order to specify the program in something which was as near machine code as I could make it so that  we eliminated most of the mistakes and that was my philosophy throughout my life with computers.  I wasn’t a systems engineer, I was a statistician that we had to work in the language that was being used by the programmers as near as possible. 
It was much too big for ‘Gert’ and ‘Daisy’, the regression. We managed to make it work but one  time we had the marvellous idea of joining the two up to be one machine. The only problem was that their timing was set by one of the disc’s rotation and of course, there was nothing because  you couldn’t really use a different disc to set the timing up so there was no way you could tie in the timing. 
And then while I was there I did a Statistics course so I really learnt quite a lot of Statistics at The Chelsea College Department of Technology, and I got an MSc. No I qualified to take an MSc  and I was then demobbed and I went in to teaching. 
DP: Where were you teaching? 
SB: Harrow Weald Grammar School, which is very near Harrow but is nothing to do with Harrow. I hadn’t got a good degree, I got a 2:2, which didn’t qualify me for the top rate of pay as a teacher. And having qualified, you know, done my two years and I think another year I left teaching in order to qualify to get a good degree in order to complete my statistical work. And I went to The  LSE and had all the top teachers in England who were there at the time. It was one of the hotbeds of Statistics. 
DP: Can you remember who taught you? 
SB: Now the main person who taught me was Kendall. He became later Sir Kendall. And Stewart, who did the main mathematical statistical book, and various other people. I specialised in design of experiments because I’d become exposed to that much earlier on. And my job while I was a trainee was to, I did it by examination because I was going to do it in one year, I’m the only person ever to have done an MSc in one year at London as far as I know, I was then anyhow. 
And my job was to read Statistics - Kendall and Stewart had really picked holes in it and find out   where it was weak and do every single mathematical example, which they were very advanced, it was not an easy option. Anyhow I got my MSc and then I had the problem of whether to go back in to teaching or to go in to industry and I decided the only thing to do was put it in the lap of the Gods and apply for jobs in both. Well, I think I had decided to go in to industry because the jobs I decided to apply for in teaching were really above my standard.
DP: How were you funding yourself during all these jobs? 
SB: I funded myself on the savings that I had made while I was teaching. I’m a bachelor and I didn’t, I was very much involved in the East End of London Junior Youth Club Settlement and because of that I lived there free so it cut out a lot of my costs. 
DP: So what sort of advertisements were appearing in this time? This must be what? Mid to late ‘50s? 
SB: Early ‘50s. 
DP: And who was advertising for statisticians with your sort of Masters qualification? 
SB: Well, I don’t remember but I applied for a job that was advertised for Lyons and they interviewed me and asked me what I wanted to be paid naturally. 
DP: What was the job that was on offer? 
SB: As a statistician. To look at a computer system they were building and to work on weather forecasting, both of which I was very well qualified to do. 
DP: Why weather forecasting? 
SB: Well, the job was within the Ice-cream Division. 
DP: Well, what was that called that division, can you remember? 
SB: Lyons Maid. And I worked with a Peter Shaw who was the meteorologist and a lot of other people. Oh no, not yet, I’m jumping the  gun. Anyhow I started doing this job and it soon became obvious that there was no way the Ice-cream Division wanted a statistician to assess a computer system that another department was writing because if I was going to make any useful input I needed to be in the team that was writing it. I could still assess it, or at least other people could assess it, but I picked up enough nonsense to realise that it was not a goer. It wasn’t a goer anyhow, but that’s another matter. 
DP: What sort of assessing were you doing? 
SB: Well, I was employed really as a statistician to look at the system so I was working as a systems analyst in modern terms. 
DP: Could you tell me what that involved? 
SB: Well, it involved looking at the code and seeing if it made sense, and where there were blunders of a mathematical sort, although it stretched over. It was an extremely complicated job. The ARDOC which it was called. 
DP: What does ARDOC stand for? 
SB: Alpine Refrigeration mmm Organisation Control. I can’t remember what the ‘D’ was. 
DP: And this was a computer? 
SB: This was a computer. 
DP: A primitive computer  or sophisticated?
SB: Very sophisticated. 
DP: Built by Lyons? 
SB: Built by Lyons. It was designed to produce every day a loading schedule for every van taking ice-cream to every depot that was getting ice-cream based on their sales of ice-cream in the previous period. Not always the depot, could take every van. Not every, the demand from ice-cream varied an awful lot depending for example if there was cinemas in the area or not. And we knew that the ice-cream, the sales varied dramatically with weather.
I, in fact did a spin-off on that which did work which was to provide every dealer with a stock for  their freezer cabinet. It didn’t order them or anything so it was a one, each one was separate and therefore it was a much more containable element. And the philosophy was that you want to keep every freezer cabinet full because that’s where your best stock is, it can be sold immediately and anyhow  that did work and I had great fun implementing that over the country.
DP: And would there be the reverse side of that, that the manufacturing operation would…?
SB: Yes. Well, this was all in ARDOC, which was just too big and the forecasting systems were too poor to get it really working. 
DP: And ARDOC was it linked in any way to LEO or was it a precursor? 
SB: Yes, it would depend on LEO. 
DP: Ah, so it was separate from? 
SB: Both of these jobs that I’ve mentioned. 
DP: Yes. 
SB: The one that I got working, which was the freezer, freezer stocking job and ARDOC were big computer programs, their object was to run overnight. Oh, ARDOC was to run overnight and produce a stock, a loading schedule for the morning. Of course, the problem that had been overlooked was if ARDOC, if LEO failed to produce a loading schedule everybody was in stuck because there was no way of substituting, unless we had a secondary loading schedule produced a day ahead and that would have doubled the size of the job. 
DP: May I just stop you for a moment and, and just go back a bit so that we see how LEO became part of the story? When you joined Lyons Maid LEO was in existence at that point? 
SB: Yes. 
DP: And so it was decided was it that you could take advantage of the computing power of LEO? 
SB: Before, yes. Before I became involved it had been decided that LEO should do things more than just payroll, which it could do quite easily, but so could a lot of other people and do some things that other people couldn’t do. And ARDOC was the first one. Later on I tried to build a similar system for the bakery called BROVAD. 
DP: Could you remember what BROVAD stood for? 
SB: Bakery Replenishment Of Vehicles And Depots. 
DP: So you were able to use LEO. Can you recall which LEO you used and  where were you located , and where was LEO located at the time? 
SB: When I was with the ice-cream company we were in Hammersmith. Just opposite the station we had a big building which was called in those days Lyons Maid House. The, Shepherd and Manning, which was the other company, the consultants that were involved, but I mean they were internal consultants, was opposite side of the road to Cadby Hall- in a house called Hartree House and very near to Cadby Hall. 
DP: Right. 
SB: No Hartree House was the computing. 
DP: Yes. 
 SB: And very near to Hartree House which is where the LEOs were. 
DP: This is LEO I and LEO II? 
SB: LEO I, LEO II and LEO III.  And also the 370, whatever it was when we moved away from LEOs. But that was after my time really. 
DP: So can you just tell me a little bit about  your own working environment when you were at Lyons Maid? You were in a large office with other people ?
SB: No, I was nearly always in a small office. I shared with Peter Shaw. I shared an office at one point. That was  when I moved, when I was in Lyons Maid House, I was in an office of three or four. 
DP: All working on this same project? 
SB: All working on the computer system. 
DP: And how were you trained in using the computer codes and programs? 
SB: Well, I went up to Bayswater and went to the various courses. 
DP: This was in Queensway? 
SB: And I could always call those in and make full use of them. 
DP: How did you find the programming, was it complicated, difficult? 
SB: I was never a programmer. I once worked with quite some complicated programs. I also was, used a different, a whole series of different computers for my own purposes. LEO was not entirely suitable for a lot of my own researches on ice-cream sales because it was much more. Take your handling machine, and I used Elliott 803 and 503, both of which I used by night. I used to go up to North London and they used to hand over the whole machine basically to me. 
DP: Why would they do that? 
SB: Lyons paid them.  But a very low rate because it was useless to them, they wouldn’t be using it otherwise. And some of my regressions for example that I did, were very big number crunching exercises and LEO, I’m afraid to say, was not the best machine in that area that you could find. 
DP: Which LEO were you using predominantly for your Lyons Maid work?  
SB: I think LEO III. 
DP: Yes. 
SB: Yes, but I may have used, I think I went in to LEO II initially but it was really LEO III because I wanted to program it myself, or at least write a pseudo code for it so that somebody could take it apart, straight on to do a bit of tidying up. 
And therefore until CLEO came along, machine code wasn’t suitable for that purpose because by  the time you wrote a machine code, a better machine code, you were away in to the arcane what ifs, you know. I mean I had the great ability, and it really is a great ability that I had, I wrote a general program for forecasting which took in data, transformed it because I wanted to know if I, if it was the log of the data that behaved in a certain way or the, or whatever it was. So I used to, I wrote a pre-program which enabled me to just take an array of data and transform it by taking the logs or take the square root or taking whatever, antilog or the expenditure is that it or whatever. 
And the first time I tried it on LEO II, this was definitely LEO II, I had a sub routine to take a square root, which was one of the options. And it’s the first test corrector and it didn’t work and, very typically for me, what had I done? I had taken the floating point square root of two and used that. And there was a device  in the machine code for going if it was greater than square root of two, or lower, less than the square root of two nobody had thought of it being the square root of two, but that was, you know, time and time again I was able to find out everything in code, and that is probably the most famous.  
DP: And what did you do with that information, who would you be dealing with at LEO? 
SB: Well, with that whoever was responsible for it for me. The programmer had to produce a modification of it to deal with the equality sign. I had various people. We had a team working in LEO and I can’t remember their names, I’m very bad at names. 
DP: When you say team working in LEO, you mean separate from the team at Lyons Maid House? 
SB: Yes. 
DP: Why, but they were Lyons Maid people? 
SB: No, they were LEO people. 
DP: LEO people, ah, dedicated to working with you?
SB: They were writing the program.  To our specification. And it was a fairly large team. Greg comes to mind as a name. 
DP: And you were running these when, during the day or during the night? 
SB: Nearly always during the night. During the day the machines were very well occupied by payroll and other such jobs, therefore we could use it after they had finished, which meant we usually went on at about midnight and then had it through  with a skeleton  number of operators. 
DP: And were you there working with them at night? 
SB: I would work with them, yes. 
DP: And what might happen, you’d find some, an error would come up? 
SB: An error would come up and  we would pack up at that time as a rule, although I might be able to see ‘Ah yes, that’s what the problem is’. I’d usually spend about twenty minutes trying to find the error and if I couldn’t… We didn’t have systems people with us. 
DP: But you were the systems man? 
SB: Sort of.. 
DP: How were you getting the data in to the LEO? 
SB: In various ways. Usually in a realistic way. We used the papers with the... 
DP: Punched, no? 
SB: No, not punched. 
DP: Ah. 
SB: We were really, it was really after punch. They had developed a system where they had a, a line of data divided in to four bits, and it wasn’t two, four, eight but it was what? I can’t remember what it was but essentially then twenty, forty, eighty but it was like that. And they had dips which you could join up, the paper across, and there was a machine reader which would read the data in. 
DP: What was that machine called, could you recall? 
SB: There were two of them, two generations of them. 
DP: What about a Lector, did you come across a Lector? 
SB: I really can’t remember. It wasn’t so important to me. 
DP: So you would run the program at night. You’d find errors or not and then you would take the information that you’ve gained back to your offices? 
SB: Well, it wouldn’t all, I normally tried to have an output which was printed out on the, it wasn’t printed out by card it was printed out directly on paper either by the infamous machine later on that caught fire, or you know about that don’t you? 
DP: Tell me about that? 
SB: Elliott, LEO III had a, I think it was called a Lector which printed paper, printed the results ready for loading up on the van or the payroll slips, they, and this, or indeed filling in because the idea was that in our case was that  we produced a schedule of what should be loaded on to the van and then what was loaded on the van might well be different so they needed to be, the form both contained data to be ready as well as data being printed out. 
And this was printed by… No there was a Xerox copier system. Now it worked at about, twenty feet a minute, or faster than that, and it used to, the Xerox method of, there were two things put up on to a screen essentially. One was the background and then was anything to be printed on the background. The background was done by a slide and the foreground was done by a computer which put it on the same thing and then they were merged and printed. And it was a very, very good system. Unfortunately it caught fire. Xerox ran at a high temperature and the paper caught and there was nobody in the computer room and the LEO III was virtually a write off. You didn’t know about this? 
DP: Tell me what happened and how you learnt about it? 
SB: I learnt about it the next morning, everybody knew. We had awful problems because every single LEO job was kaput. All the data was on the computer but it was usually greased up by this, the smoke and what have you so, and we used somebody else’s LEO for a couple of months I think. But it wasn’t only that it was that the data loss, everything had been lost. It was, it was the death of LEO III. 
DP: This was in the Whiteley’s Building was it in Queensway? 
SB: Hartree House, on the first floor.  No, it didn’t burn the building down but it did make grease and smoke and what have you got in to the computer room and computers don’t like that. I’m surprised that that has died a death because it really was a major, the... 
DP: Were the van men waiting for it? 
SB: Everybody was. 
DP: So what happened? 
SB: Everything ground to a halt. Payroll was paid by the previous week for week. 
DP: What about the difficulties...? 
SB: And every, every back up system that was available had to be brought in to use. There were, there was data, back up data that had been stored out of the computer room but anything that had been stored in the computer room, which was the majority of it, was damaged by the fumes from the fire. 
DP: This would be what? Cards, paper tape, magnetic tape? 
SB: It was magnetic tape mostly. Cards were, were out at that time. 
DP: So what did this do for your work in particular, how far were you set back by this? 
SB: I was off it, that area at that moment. I was working on another computer which Lyons had. A portable, the first ever portable, IBM portable. I can’t even remember what it was called. I was working on a thing called the Company Model, which I have a printed set of printouts which I can show you but they’re not here. 
DP: And they’re not LEO? 
SB: It, LEO wasn’t considered safe enough. 
DP: In what way safe? 
SB: Well, it took every department, their plans and money that was going to be invested in them and it was a very considerable risk. 
DP: Ah, right. 
SB: It was a very considerable risk for... 
DP: Commercial risk? 
SB: That say LEO Computers found out that we were not going to invest in them next year, you know, and it would have a, quite serious consequences. 
DP: So you wouldn’t run this on a LEO computer? 
SB: So we did run it on a LEO computer. 
DP: Ah. 
SB: But they, soon afterwards I ran it on an outside bureau machine. 
DP: Right. 
SB: And then we got a machine in to run it on our own because we were worried about… In fact when I started running in to trouble, we were very, very worried about the, even taking this down to a bureau, the bureau see hard printout. The first thing we did was we stopped having the printout and we, or the printout backed to ours so there was no immediate thing to see. But the end product was it ran on a huge, well a small computer but it was very powerful. It was the first computer, portable computer that was built. 
DP: So can I just take you back to your original project, which was in the weather forecasting area? 
SB: Yes. 
DP: Which you were using the LEO computing power to assist you. What was the outcome of that? 
SB: Well, it was needed for ARDOC. 
DP: And was it successful in the end? 
SB: Well, ARDOC was given the boot because ARDOC was no, couldn’t deal with the fact if we didn’t get a printout one night we were really in an awful problem. And however I looked at it I could not see a way of getting out of the fact that we had various people who were trained up to deal with the manual system, if we got rid of them we would have no back up if, for some reason, the computer failed to produce. And in those days computers did fail to produce, they went down. 
DP: Simon forgive me, I haven’t quite understood the ARDOC/LEO relationship?  Is ARDOC a program or a system? 
SB: It was a system  which ran a group of programs. 
DP: A group of programs that ran on the LEO? 
SB: Yes. 
DP: And what about your modification then? 
SB: Those were ARDOC  until I moved across to the bakery rota, similar but easier system for the bakery. But again that didn’t work, same reasons. I had programmed it so we produced two days work and the second day was scrapped under normal circumstances, but even that was not really sufficient for down time on the computer. 
DP: So you did two days so the second day would be the back up? 
SB: Yes. 
DP:  And that would give you then a day to put something right? 
SB: That’s right. 
DP: Was that enough? 
SB: It depends what goes wrong with the computers. 
DP: Well, what might happen? What, what sort of call might you get to tell you that something was wrong? 
SB: Well, it was usually something quite simple like a valve burning out or, well, they didn’t use valves. 
DP: This would be LEO II presumably? 
SB: That would be LEO II, but with LEO III there were... 
DP: They were now transistors? 
SB: Yes. 
DP: And they would fry as well would they? 
SB: Well, they shouldn’t but if you rely on it every day of the week for every week of the year it doesn’t take much for one day to… There was also the problem that, and there was a concern about this, of the head office of Lyons putting all their eggs in one basket and that these programmers, the operators would really have the power of God over the machine, everything, and this was a serious consideration. 
DP: Was it a neurotic consideration or was it a real consideration, a real problem?
SB: I think it was neurotic, but Lyons did have its problems on personnel of one sort or another from time to time.  I think the idea of data getting out was neurotic, but the idea of the programmers failing to produce the stuff at the end of the night was by no means neurotic. It could happen for a variety of reasons,  some of which were not in their control.  
DP: Give me examples?
SB: Well, just that computers today, we assume that we turn a computer on, we turn that thing on and it works, or occasionally it doesn’t work, but if it doesn’t work you pick up another computer and use that. In those days, although a lot of work went in to having back ups that if one computer goes down, if one LEO III goes down, there was an arrangement with another operator to, to back it up but the data might be or not be there and then would have to be transported there. The data, the output would have to be transported back. It was a major consideration. 
DP: How were you treated by other members of the organisation? You were this group of mathematicians, statisticians? 
SB: I was a, well, I was an online line manager. I had no responsibility. You see I was doing other work as well you must remember. And I can explain it better there. I might be consulted by somebody, now the agreement right up to the top was that anything I did I would get the pat on the back but so would the person who had asked for the work to be done. And it wouldn’t be, their pat on the back wouldn’t be diminished because I’d put, had some input. 
Mm, now I’ll give you an example, nothing to do with the computer although it has side issues. A  repair, the company that repaired the fridges, both our fridges and customers’ fridges, was very worried it was growing in to a very big rate and was very worried about where it should build its depots. And I was consulted and produced a plan, a notional plan to go up from six depots to twenty depots over the next fifteen to twenty years. And five years later he came up and said ‘Well, you know, we’re going to have to update this because we’re increasing faster than we anticipated and we’re nearly at the end of your schedule’. And they were sticking to it absolutely bang on, which was not what was intended, it was to give a… I mean it didn’t say build a depot in the middle of the Thames in London but, you know, it wasn’t exactly able to say build at, look where a site is available for a depot in South-east London and use that. So anyhow, but that is an example.
 Now I remained best friends with them and I put a lot of effort in to this relationship with the person, and that I was working for them and not for Lyons. I was of course, working for Lyons but effectively they got the kudos of any development we did. And in management services today I wish that was still the case but it’s not always, they often want the kudos for themselves. 
DP: But today the IT Department of any company... 
SB: Yes. 
DP: Is just, just another department.  
SB: Yes. 
DP: But in your day was it viewed as something quite significant? 
SB: It was... 
DP: Possibly even with some suspicion about what, what it did and why? 
SB: They didn’t know what it did, but then they didn’t know what a lot of other parts of the company did. In the early days LEO got a lot of bad publicity, because it was LEO III before the first job really paid for its keep as opposed to the pure costs of the LEO establishment with all the hardware and so on compared to doing it manually, the manual was always cheaper. Payroll for LEO One, PA…, payroll for LEO II, payroll for LEO III were always expensive options compared to doing it manually. 
DP: What was the advantage, speed? 
SB: The advantage was that Lyons management believed in it. 
DP: But they were right to believe in it weren’t they? 
SB: Well, cost wise they weren’t. It did a marginally better job. People who got, more often got the right pay than with the manual system. The manual system can crack up. Whether it’s better to have done it that way, it was a terrific gamble and I mean the number of times you see photographs of computers, people sitting at a, in a huge room all with their own little desk doing exactly the same thing, and it’s thought to be today quite obviously a non-sensible thing. In those days on a cost basis it was the right thing to do. 
DP: What was your impression when you first went to Lyons as you went in to the LEO, probably  LEO II computer room? Were you impressed by what you saw, were you overwhelmed by it? 
SB: I was amazed by it because I compared it to what I’d seen before and used before. Because remember I had been exposed to ‘Gert’ and ‘Daisy’. Now ‘Gert’ and ‘Daisy’ were in a room about four times this size and they were two computers. Now, you know, it makes LEO rather massive. The thing that I found inspiring was the add ons, not the computer itself. The computers are fairly, even then was a fairly small element. But you had card readers. You, or you had disc readers. You had printers, you had things for the operators to type at, you had air conditioning, the whole thing was a massive system which worked as one, and usually quite well. It still had an awful lot of people around. 
Okay, they were operators, they were programmers and they were systems analysts and they  were heating and ventilation engineers and they were, but there were an awful lot of people involved. And the system that was important really was the, the system that controlled those people so that they weren’t all doing the same job. 
And now to me the Eastern idea of various computing engines were quite, you know, the beads  and so on, I mean we were up and down, you could get fantastic rates but the important thing was interpreting that and getting that back to mean something. And I found really that the payrolls came out with a name at the top and their details, which could be given to the person in an envelope, and the bank could be notified of the same thing. That is the particularly clever thing. 
The mathematics I don’t think is all that clever. I mean even today I mean one looks at something like that, it’s very clever in a purely mechanical way but it’s not earth shattering. LEO III, LEO I was earth shattering in that it was showing another approach to a common or garden job. Things like ARDOC were trying to use the things that we had to do things that couldn’t be done manually or,  didn’t need to be done manually but could have an advantage if we could do them. 
I mean the steel companies or the oil companies were developing computers to do things that LEO I, II or III could never even attempt. Fractionating oil for example. Lyons were lucky in that they had an engine, a system which had a fairly simple need which they were worried about, because as they expanded, the number of office people expanded and expanded and expanded and would continue to expand and they couldn’t see an end of it and they didn’t like having all those non-productive people around. I mean doing payroll is productive but it’s not productive in the sense, that the end product isn’t there. 
DP: It’s not like making Swiss rolls? 
SB: Well, it’s a support service essentially. It’s to support the people who are actually selling the product. 
DP: Did you get any response from the people on the ground as it were? Did you ever hear directly from the van drivers for instance? 
SB: Yes, I did a lot of work with van drivers. 
DP: What was their feeling about the processes? 
SB: Well, again I mean my first, if we can take programs this is the best example, the Bakery Replenishment of Vans and Depots. Now they got, initially they would be, in the morning they would find on their van certain products and they would look, have a look through and this was the existing system and they were very unhappy with it basically because it limited, it didn’t always give them what they wanted. 
Now I first of all studied the problem from their point of view. I would spend weeks going out with the drivers as their assistant, because they sometimes had an assistant in a white coat. Having to get up at five o’clock in the morning wasn’t an exact joy but I would go out and examine what they were doing and ask them politely questions, and sometimes you came out with very curious answers. I mean take an example, which is a straight theme possibility of a computer program. Lyons Little Rolls. You remember them? Made in four colours. 
DP: Yes, I remember the chocolates. 
SB: And I asked them very simply ‘What happens if during the day you find you’re running short of one’? And one driver said, ‘Well, I will stop selling them because I will only sell three instead of four’. The other said ‘I will take the orders as if it’s four and deliver four, although one, two of them will be the same colour’. Now I can’t say which is right. And I mean even now, even then I couldn’t say which is right. What I did know was that I couldn’t program for both. 
DP: What happened? 
SB: Well, they continued with the same system and I ignored the problem and basically decided on how many mini Swiss rolls I should put on the van and then split it in four, it ducked the question. It shouldn’t have ducked the question but I, finding out what every van driver did in that situation was not a practical situation. 
DP: Did the van drivers become aware that they were being given information from a computer rather than from a team of individuals? 
SB: I didn’t use the word ‘computer’. I said ‘Rather than doing it at the depot and then deciding on a loading by the depot and not telling you, we can’t tell you because there isn’t time but it might be done centrally. And we have a, and I’m working to see if we can find a system which will be fairer  to all of you’. Because it was always said when they got the bad load it would be fairer to, but I avoided it. Well, computer wouldn’t have meant anything in those days but centrally it did. No, I was very much a hands on. 
DP: How did you get on with the people actually at LEO Computers themselves? Was it a good relationship with them? 
SB: Reasonably good. I was a user and I’ve always kept the idea that I was a user, I was using the computer. I had very little to do with other programmers. I had a lot to do with the operators. They would be quite interested in what I was doing obviously, and there are always odd bits of paper that it produces that you could show them and say ‘This is what we’re trying to get the computer to do, wouldn’t it be good if we could break in to that area’? 
And it’s much more interesting than payroll to them, payroll’s rather dull. I mean it is necessary  to think the job through, not to lie. I mean if I had a computer, if it was working on the computer, the computer was producing a loading schedule I would quite often go through the loading schedule with the driver and talk about the different colours of mini Swiss rolls and how impossible it was to cope, and how both of those decisions were right because it was never my job to tell them how to run their business. 
DP: Can you remember the move from computer coding to CLEO? 
SB: Yes. 
DP: I mean was that a sort of tremendous breakthrough as far as you were concerned? 
SB: I had already been using Elliott Intercode. I’ve got one up there, I saw it the other day. I had already been using intercodes of one sort or another. I didn’t see CLEO was a particular breakthrough. I was very intrigued that they’d decided to produce their own. Okay, it was in advance of the intercode, most intercodes worked. It had some areas, the writers which I had greatest respect for. For example, it could run at, a drive forward or backwards looking for a particular entry. Most systems of that period only would run forwards and being able to run backwards was a big, big advantage and having it happen automatically.  I got fed up with CLEO because in the sorting it was very, very stupid, there were many better systems.
DP: Could you elaborate a little on that? 
SB: Well, if you were sorting things, if you’ve got a series of letters, say, and you’re sorting them, what do you do personally when you’re sorting them? You, you probably divide them in to half, there was the ones before ‘M’ and the one’s after ‘M’ and beyond and then you divide those. But there were better ways of doing it than that. And a computer system should be capable of using one of those better ways. It is a very difficult job, it sounds very simple.  I’m not very much in touch now with, I use computers. 
DP: Oh how long were you with Lyons in all do you think? Were  your entire career there? 
SB: Forty-seven years I think. 
DP: Really? So you retired from work from Lyons? 
SB: No. After the merger.  
DP: With? 
SB: With Allied Foods. 
DP: Oh yes. 
SB: I, my Company Model which was my job at that time. I mean I, it was specifically my job and my only job came to an end. The new management wasn’t interested in computers per se. They weren’t interested in, and it’s wider than computers, they wanted to run the business from the seat of the pants and they could run a pretty good business from the seats of their pants, they were good businessmen. 
But I had no direct  input in to that system. I looked around a bit. When I came up with a warning of a system and everybody, I can tell you this now, it’s nothing to do with Lyons, the computers but it’s a good example. We ran a system which was computer based but not LEO based, where we realised that we had to handle a whole load of different currencies. We bought tea in Rupee’s and we had at that point a lot of companies in America and a lot in Europe and they all had their own particularly these. And we totted up and I, I don’t take any credit for this a part, we totted up the total and said that in fact when you look at the system you don’t, or didn’t, buy Euros with pounds, in those days you bought Euros with Dollars and Dollars with pounds and the two matched up to give you the total transaction. It’s not quite the same now, that’s why I can tell you that. 
Now we found that we could work out what was happening, and the exchange rate tends to go on on a trend, the breakdown will go on a trend or go down or up, and we could make an awful lot of money at Lyons by following, doing, going ahead or not going ahead according to what this told us to do mathematically. And in fact for three years Lyons made more profit from doing that than they did from the rest of the business. 
 DP: Currency dealing or trading? 
SB: Mm, not really because we had to buy them. You see the idea, we, when we bought Rupees now or Rupees in six months when we needed it. We had to, we had to gamble. 
DP: Are you saying that you produced a program that facilitated that gambling? 
SB: I didn’t. A program was produced, well some mathematics was done, the program wasn’t important. Now we were able to make money hand over fist because we were the only people in it doing it. Then a small exchange in Chicago opened up which was doing the same thing but enabling people to gamble. Now I immediately said ‘Hang on, the game has changed, the statistics no longer apply, beware’! And we lost money for the first two weeks, three weeks. And I said at this point ‘We should scrub it, forget it, not continue doing it, the money’s gone out of it’, and I was overruled by top management who went on gambling and lost a fortune. 
DP: When did you leave Lyons? 
SB: I was sixty when I retired. 
DP: Sixty, ‘93? 
SB: I was sixty when I retired completely. I was fifty or about when I retired, so that’s’83. 
DP:  What did you do for the next few years then, you were still a young man? 
SB: I went back to my teaching. 
DP: Ah, did you enjoy that? 
SB: Yes. I taught in Brazil, which was very interesting.  
DP: [Laughter] Why Brazil?
SB: Well, because I’d taken a number of holidays in Portugal and learnt a bit of Portuguese, I liked the Portuguese. 
DP: You have your father’s facility for languages? 
SB: Well, I’m very bad but I thought ‘Well, if I’m shoved in to a Portuguese speaking world I will learn Portuguese’. And I taught at an English speaking school so I didn’t learn much Portuguese. I did some work with an orphanage in, in Sao Poulo but I had a very good time. 
DP: But looking back on your days with Lyons, does anything really stand out that was an interesting career for you? You were involved with computers at their early stage.
SB: Yes. 
DP: How do you reflect on that? Was it a happy time, exciting time, worthwhile? 
SB: I think it was a challenging time. Oh definitely the word challenging I would say. I wasn’t able to rest on my laurels, I had to do something. Some computer jobs and some of these jobs were very long in terms of years but equally they break up in to stages, and the stages are not necessarily very long and they have different…
I got involved in an orphanage and had been involved earlier on with the youth club community centre. I knew,  strangely enough, there David Lowie, who was a programmer with Lyons. Our lives have crossed endless number of times. 
DP: What was he doing there? 
SB: He was a systems engineer , and he was a manager of the youth club when I was in charge of the youth club. 
DP: Well, I’d like to thank you very much for what’s been a very interesting conversation. Apologies for the telephone call that I didn’t take, but I’d forgotten to switch the machine off, I’m afraid that happens, I hope I can edit that out in due course. And I’d like to just end the interview by saying that the interview with you, Simon Benedictus, has been recorded by The LEO Computer Society as part of an oral history project to document the earliest use of electronic computers in business applications and any opinions expressed are those of you, the interviewee, and not of the Society and copyright of this interview is recorded, in recorded form and in transcript remains with The LEO Computer Society, 2012. 

Provenance :
Recording made by the LEO Computers Society as part of their ongoing oral history project.

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.