Home > LEO Computers > LEOPEDIA > Oral & Narrative Histories > Patrick Blakes: Inter ... 27th July 2016 53385

Patrick Blakes: Interview 27th July 2016 53385

 Home > LEO Computers > LEOPEDIA > Oral & Narrative Histories > Patrick Blakes: Inter ... 27th July 2016 53385

Patrick Blakes and LEO Computers Society

Digital audio of a recorded interview with Patrick Blakes, who worked as a commissioning engineer on LEO II/7, ultimately becoming Chief Engineer on Ilford Films' LEO III.

Interviewer: Cyril Platman
Date of interview: 27th July 2016
Length of recording: 27m34s
Format: 3 original .mp3 recordings 39.36MB (transferred to .mov video for presentation on YouTube 114.12MB)
Copyright in recording content: Patrick Blakes and LEO Computers Society

Transcript editor: unknown

Abstract: Received background training in Electronics doing his National Service in the Royal Air Force. Decided on career in that field after seeing an advert for trainee engineers from LEO. After training worked first as commissioning engineer on LEO II/7 than became site maintenance engineer on LEO II and LEO III sites including acting as chief engineer on Ilford’s LEO III.. Retired from ICL in 1969 to join an independent computer maintenance company working on LEO machines.

Date : 27th July 2016

Physical Description : 3 digital files, audio

Transcript :

Patrick Blakes (PB) interview 
Interviewed by Cyril Platman (CP)

The Computer Engineer’s Story

Early Days

I started work with LEO Computers in 1958 as an engineer and left what had by then become ICL in late 1969.  In that period I crossed swords with LEO II, the basic LEO III and the later, faster version the III/F.  And it was all entirely fortuitous, unplanned and the product of random chance.  In Shakespearian terms, ‘life, and in particular LEO Computers, provided the stage and I just stumbled about on it’.  
I was born in 1938 whilst Hitler was busy marching into Austria.  For background, I suppose you would say I was definitely working class.  Council house, one cold tap, outside loo and only allowed a fire in your bedroom if you're ill.  A few memories from the war years, we had evacuees, who were worse off than us, billeted with us.  I remember us all rushing down the air raid shelters at primary school when the sirens wailed.  If we were at home it was under the kitchen table quick.  I remember flights of towed gliders going over, that must have been around D Day.  But the nearest we came to the real war was when a V-1 flying bomb, doodlebug, came down near enough to blow our windows in.  The bitterest memory, however, was the VE Day street party.  I could hear everyone enjoying themselves but I was confined to bed with measles and feeling rotten.  At the local primary school I enjoyed what was on offer.  Officially, reading, writing and arithmetic; stories and sums as far as we were concerned.  And I suppose, because I enjoyed it, I did well enough to be regarded as ‘academically inclined’ and passed the exams to go to grammar school.  At Hertford Grammar School I took to ‘stinks and sparks’, that's chemistry and physics, and a heap of ‘O’ Levels followed.  I think it was hoped and expected that at this point I would transform from a mouth to feed into a provider, i.e., leave school and get a job.  The headmaster, however, had other ideas and persuaded my father that I should stay on for ‘A’ Levels.  ‘A’ Levels in maths, further maths, physics and chemistry duly followed, although by this time my interests were definitely more practical than academic.  

National Service
In any case two years’ National Service now loomed.  I was admitted to the Royal Air Force, spent the first year of my prescribed two undergoing assorted training courses.  From initial basic training, essentially running, jumping and standing still, on to weapons training, then on to basic electronics courses, radar techniques and practices and then on to the navigation bombing system fitted to the V bomber squadrons.  These airborne radar and calculating systems took note of where you were in relation to the target, how high you were, how fast you and the wind were going and in what direction and calculated where to go and when to drop the bomb.  The people flying them didn't really expect to come back and didn't know if there would be anything to come back to anyway.  Be that as it may, after a year I was a fully trained specialist air radar fitter.  However, the RAF was at this time in a state of mild internal disarray, and accordingly I was posted to a station that had none of the equipment that I had been so expensively and diligently trained on.  
No-one seemed to know what to do with me, so after a few weeks of enforced sloth I set about training myself on some of the stuff that they did have.  After a few months I think I may have been at least useful, whereupon I was re-posted to another station that did have the kit that I had originally been trained on but had now forgotten about of course.  I probably delivered about two months useful work from my two years’ service but it was all useful to me.  The grounding in electronics was thorough and although the computing was all analogue rather than digital it had been clearly established that electrical machines could do sums and very complicated ones at that.  [Technical Note. The radar circuits were pulse technology which is the technique used in digital computers.]

And So To LEO
When I left the air force I had two job offers, one from LEO as a trainee computer engineer and one, on the strength of my ‘A’ Level maths, as a trainee actuary.  LEO promised hands on engineering which appealed but the whole enterprise seemed an unknown quantity; was slightly weird and would lead who knows where.  The actuarial opportunity on the other hand seemed a steady, secure introduction to a recognised profession.  How did I decide?  I tossed a coin and was, I think, secretly pleased when it came down LEO.  So off I went to Minerva Road at the tail end of 1958 as a trainee computer engineer, for nine pounds, fifteen shillings a week.  

The LEO II training course at Minerva Road was run mainly by a gentleman named Colin Lewry  and was essentially about how a machine could turn an English statement such as ‘add the contents of memory location ‘x’ into register ‘a’, into an unambiguous series of discreet logical steps expressed algebraically, and then implement that boolean algebraic expression in electronics’.  On the practical side we undertook testing of various sub-assembly units as they were delivered from sub-contractors before we were let loose on, or allowed anywhere near, a complete machine.  In my case I went on to the build of LEO II/7 in Minerva Road before taking it to pieces and accompanying it to the British Oxygen site in Edmonton, helping re-assemble it all and lending a hand during commission and acceptance in February 1960.  

Looking back I don't think any of us had any idea of where it was all going.  We were too tied up with doing what we had to do at the time.  We knew it was something different and out of the ordinary and viewed with a mixture of awe and suspicion and plain disbelief in some quarters.  One of my colleagues came in by train every day and tried in vain to explain his job to a fellow traveller who flatly refused to accept the possibility that a second could be divided into a million discreet intervals, never mind that something could be accomplished in that time.  Again, looking back with hindsight, I realise that memory can play tricks on us but I don't remember any of the engineering staff at Minerva Road being what you might call ‘completely normal’, except me, of course.  We seem to have been a right bunch of odd balls, perhaps I have just forgotten the normal ones. 

The Job
As was customary at the time, British Oxygen employed their own programmers, operators and systems staff but LEO supplied the maintenance team under contract and I was one of them.  The chief engineer was George Manley and we had, at various times, two or three shift engineers, depending on the work schedule and specialist mechanics for the tape and card readers and punches and printers.  The Samastronic printer in particular was a work of Satan.  If you wonder at this profusion of maintenance staff you have to bear in mind that LEO II was made of valves, controlled light bulbs really, glass vacuum tubes that worked by electrically heating a filament of metal to white heat and trying to control the flow of electrons from its surface.  They wore out, changed their performance and characteristics over time and were prone to sudden, unpredictable failure.  And the same was true, to a lesser degree, for the vastly greater myriads of other components involved.  At this time your domestic television receiver came with user adjustments for at least brilliance, contrast, line hold, frame hold and tuning, to accommodate component ageing and you expected it to fail completely now and then, and it might have ten valves in it.  A full complement LEO II had around eight thousand and consumed thirty five kilowatts.  In short, maintenance of a LEO II was something you had to keep on top of but with eight thousand valves and hundreds of thousands of other components, soldered joints, connectors, plugs and sockets, wires and cables to contend with it didn't do to go charging in and disturbing things.  As the saying goes, ‘if it ain't broke don't fix it’.  

A fine balance between routine maintenance and intelligent neglect was called for.  If we could keep things running for ninety five percent of its scheduled run time for the week we were more than happy and it was considered good programming practice to write things so that you had re-start points at no more than half hour intervals.  If something went wrong and your job lost more time than that it was your own silly fault.  To help stay on top of things and diagnose faults we had test programmes which exercised the memory and ran through all the logical operations the machine was capable of and a system called ‘marginal test’.  Faced with this vast machine with torrents of electrical pulses rushing around where each pulse of a voltage above a certain level represents a one and their absence, i.e., any signal below a certain voltage level means a nought, marginal test system was essentially a stress test.  It provided a means of slightly varying the machine’s clock speed and making the ones a bit smaller and the noughts a bit nearer to ones.  [Technical Note. Marginal testing varied circuit voltages. The machine’s clock speed was carefully controlled to compensate for temperature change altering the mercury storage tube length.]  If we could run a suite of test programmes with the marginal test system in operation we knew that nothing was on the brink of failure through gradual degradation and we had a reasonable chance of getting through an eight hour shift, unless of course a sudden catastrophic failure stepped in somewhere.  
If a problem was indicated under marginal test conditions it could be narrowed down.  The marginal test stresses and strains could be applied and/or withdrawn across a selection of a few hundred individual points in the machine.  

With all of these engineers floating around you would think there was never a shortage but not so, although there were just about enough to go around.  I once did a shift on LEO I in extreme circumstances but mercifully, because I knew nothing about it, nothing serious went wrong.  I also managed a continuous forty two hour shift on LEO II/5 at Hartree House in the middle of a flu epidemic.  

Modern computers, thanks to their intuitive and user-friendly software, can be annoying, frustrating and infuriating.  But for an engineer on LEO II things were different. LEO II could kill you.  The direct current power rails were plus 150 volts, plus 250 and plus 450, and the minus 50 and minus 200.  And once you had the doors open it was all there waiting for you to poke about inside to find out what was going on with your oscilloscope probe.  Old radar disciplines came in handy, wherever possible keep one hand in your pocket so you don't get a shock across your chest, and remove things like rings and watches.  On peripherals and all rotating machinery, one had to control loose clothing, ties, long hair and loose spectacles and strings. 

 The memory on early LEO IIs was acoustic delay lines, steel tubes filled with mercury. We cleaned and re-filled them where necessary.  The emphasis was on cleanliness for the procedure but with no particular precautions that I remember, and I don't recall it being regarded as hazardous.  The mercury had to be of very high purity, triply distilled mercury came in glazed stone jars, but quality control was not what it might have been.  I remember breaking the seal on one jar and removing the stopper only to find a fag end floating inside.  Mercury was funny and fun, we handled it and probably breathed it.  Did I mention that we seemed to have a lot of oddballs in the engineering department?  I doubt that a LEO II could be built these days, health and safety legislation would stop it dead in its tracks.
The LEO II at British Oxygen was not a one hundred percent success.  The machine was okay but the systems and applications side was not.  We ran payrolls and other stuff but the main application never got completely off the ground.  Whether responsibility lay with LEO or British Oxygen is moot.  And if you want to know more, but probably not much more, read Georgina Ferry’s book called ‘A Computer called LEO’.  

I moved on to serve as chief engineer at LEO II/9 at Ilford, and then on to admin and management, of maintenance generally, based at Hartree House.  My lasting memory at Hartree House was that the boss, Mr Gibson, had a habit of catching people as they left at five or six o'clock for long discussions of various matters of interest – to him.  He managed this because the office doors were half glazed so he could see people leaving and this led to the interesting spectacle of quite senior staff staying out of sight by crawling past his office on their hands and knees at going home time.  Happy days.  I stayed in maintenance management right through the upheavals of the English Electric mergers involving English Electric, Marconi Elliott, ICT.  

Things that stick in my mind, including being on site at LEO III/3 at Dunlop when President Kennedy was assassinated in November ’63, when no-one knew what on earth was going to come next.  I thought the Dunlop management did a fine job; married staff were sent home to their families and the installation carried on running - manned by all the single people. [Historical Note. That evening I had been asked to call in at Dunlop on my way back from Kidsgrove to look at a fault. An engineer’s wife phoned in to give us the news and I got my car portable radio into the engineer’s room so we listened to the BBC. Driving home down the motorway at 3am I seemed to be the only person around. It felt like the world had ended. Tony Morgan]
After The English Electric Merger
My other introduction into other forms of computers at English Electric followed.  I had moved to Kidsgrove and somehow, once again, ended up in charge of things I knew nothing about, but it was all very interesting.  My experience of data processing involved trolley loads of punched cards and tapes and more printer output than you could push, but the KDF9 at the Nuclear Power Group in Knutsford was different.  A boffin would wander in with perhaps a foot of paper tape, feed it in, the machine would grind away for a couple of hours and spit out another foot of tape.  The boffin would look at it, mutter, ‘no, that can't be right’, and wander off.  Similarly the industrial side of things was an eye-opener.  

I was dragged out of bed in the middle of one night for a serious problem at the Shelton Iron and Steel Works in Stoke.  I knew nothing about the equipment but my duty was to support the chaps who did.  The problem was reported as ‘an intermittent fault on a digitiser’.  I had no idea what a digitiser was but I soon found out, it sets the gauge of a rolling mill.  When you have a twenty-foot billet of white hot steel, seventeen inches thick coming down the rolling bed to be rolled to sixteen inches and the roller settings suddenly change from sixteen to 1.6 inches your choice is to run away or die.  I ran away, together with everybody else. 
I went hands on again for a bit in the early ‘60s.  The British Oxygen machine, LEO III/13, in Worsley, Manchester was installed, it passed its acceptance tests but it then proved somewhat sickly.  I was dispatched to sort things out.  The message was that what was required was ninety eight percent serviceability for four consecutive weeks, and soon, or there would be big trouble.  Apart from the engineering work that was necessary I organised a sweepstake amongst the engineers and all the British Oxygen staff on how long it would take.  We made it on weeks four, five, six and seven.  I won the sweepstake but was accused of fixing it, which I considered ungrateful.  This triumph was brought swiftly back into focus at the party that followed.  It was a British Oxygen organised affair and I was explaining to a guest that British Oxygen had bought their LEO III because of their LEO II when I was overheard and corrected by the British Oxygen big white chief.  ‘No Pat, we bought the LEO III in spite of the LEO II’.  

I was then involved in improving the serviceability of the first of the fast versions of LEO III to be installed.  They had been installed and commissioned but were proving a real headache to keep going reliably.  The central processing unit and memory of these machines was completely different to that of the original LEO III and was right on the limit, and sometimes past it, of what was technically possible with discreet component technology, i.e., individual transistors and resistors.  The machines had been sold on the basis of job timings that demanded top-notch performance and whilst they had all been made to work they all did so in a slightly individualistic way that proved a bit problematic for all sorts of reasons.  The situation wasn't helped by the fact that whilst our stand-alone test programmes tested all the logical functions of the machine it was after all an asynchronous time sharing multi-tasking machine, and the users were capable of pushing everything much harder than we could with our test programmes.  It took months and months of mainly weekends because that was when we could get at the machines, but we restored some order eventually. [Technical Note. The LEO IIIF was right on the limit. Major earthing bars were installed in the main frame and there were many other technical problems which got worse as transistor speeds got better through improved production techniques. These problems were still bugging the machines ten years later when I spent 6 months at Post Office, Charles House.Tony Morgan]

On one of the weekends in 1966 England won the World Cup; I was oblivious to this, I was in Charles House, Kensington on one of the Post Office machines and came out in the evening to wonder what on earth was going on and what the celebrations were for. I stayed in maintenance management with what became ICL until I left at the end of 1969 to become a founder member of an independent maintenance company where I became re-acquainted there with the British Oxygen LEO III/13 machine and the original Dunlop LEO III/3 machine, which was being moved to British Oxygen to boost the capacity, but that is all a new chapter of a different story.  

Closing Thoughts
How I would have got on as an actuary I have no idea.  I suspect it would have been less interesting.  
My LEO story took place in the middle of the cold war during a period of great social change and the computer industry contributed to that.  It generated new types of jobs that turned into careers where background and qualifications were secondary to how well you could actually do your job, and that was pretty transparent. It was an on-merit industry, and all the better for it.  If you want to see what England looked like in 1969, watch a film called ‘Withnail and I’.

The LEO Computer Society has recorded this interview with Patrick Blakes. The Society would like to thank Patrick very much for his time and reminiscences.

Suggested page footnotes:

Page 4
KDF9 was an early British computer designed and built by English Electric. The first came into service in 1964 and the last of 29 machines was decommissioned in 1980

Page 5
Asynchronous multi-tasking, time sharing was a LEO III innovation. Unlike LEO II that ran a single programme with operations occurring in a controlled sequence, LEO III introduced multi-tasking that enabled many programs to run concurrently. This was achieved with a Master Routine controlling operations via         "Interrupt " orders from the different individual programs demanding priority access to memory etc.  Although each individual program still ran according to its defined sequence, the overall sequence of operations across all the programs concurrently running, or time-sharing, was not repeatable or predictable. The programs thus ran asynchronously.

‘Withnail and I’ -  a 1987 British black comedy film written and directed by Bruce Robinson and based on Robinson's life in London in the late 1960s,

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

Archive References : CMELO/LS/AV/BLAKES-20160727 , DCMLEO20221230012-014

Related Topics:
This exhibit has a reference ID of CH53385. Please quote this reference ID in any communication with the Centre for Computing History.

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.





Click on the Images
For Detail

Help support the museum by buying from the museum shop

View all items

Founding Sponsors
redgate Google ARM Real VNC Microsoft Research
Heritage Lottery Funded
Heritage Lottery Fund
Accredited Museum