In the 1970s, having gained a university degree, I lived and worked in London, enjoying life, doing various jobs but not really knowing what I wanted to do as a career.
Then in early 1976 I went on a 6-month government sponsored retraining scheme to learn computing. This completely changed my life and I never looked back.
This is a very boring and very detailed history of the hardware and software that I worked with from 1976 to my retirement in 2003 - from punched cards to PCs.
Apart from my own photographs some of these pictures are copyrighted. All copyrights are acknowledged. This is not a commercial website.
You probably need to be some sort of computer nerd to want to view my website.
If you really want to you can find my early autobiography here: www.geoffnewland.me.uk/
Within my lifetime the word “COMPUTER” changed its meaning from a person who does arithmetic with numbers to an electronic machine which handles data. The term “DIGITAL COMPUTER” came into being some years later.
I started work in "Data Processing" (DP) in the 1970s. This rapidly became "Electronic Data Processing" (EDP) and at some point was overtaken by "Information Technology" (IT).
When I started working in the computer industry computers were built using transistors and were called mainframes. They were large and along with their peripherals required large rooms to house them.
Into the 1980s I worked on minicomputers which were smaller, much more powerful and were built using microchips.
ICL core memory module.
A magnified image of core storage.
The memory made use of tiny magnetic rings called cores each about a tenth of an inch in diameter. The rings had wires passing through them for selecting and detecting the contents of the cores.
Core memory was a common form of random access memory (RAM) from the mid-1950s to the mid-'70s.
The ICL 1903 computer I used on my course utilised core store and would have had a memory size of about 128 kilobytes.
This computer could only run one program at a time.
The ICL 1906S I used at Plessey had RAM based on Plessey Plated wire memory, a later version of core storage technology and would have had a memory size of about 512 kilobytes.
With the later introduction of memory based on semiconductor technology, core memory became obsolete, though some people still call the main memory of a computer the core memory.
A reel of half-inch magnetic computer tape.
A group of computer tape decks.
The first programs I worked on were magnetic tape based.
A magnetic computer tape in the 1970s might have been 3600 feet long and half an inch wide with a recording density of about 750 characters per inch meaning that one tape could hold about 32 million characters (32MB).
The data held on magnetic tape could only be accessed serially in one direction, from start to finish. It then had to be rewound. For a large system a single tape would not be able to hold all the data.
A typical batch program would
• read in data on one or more magnetic tapes forming the master file
• read in update data from a paper tape, punched card file or magnetic tapes
• write out data on magnetic tapes forming the new master file
• write data to a printer, paper tape or punched card file.
It was not possible directly to update a record on a tape. For example, on a payroll system the record for John Smith, employee number 12345 was on magnetic tape number 9876 about 250 feet from the end.
As an aside:
The Driver and Vehicle Licensing Centre (DVLC), the forerunner of the Driver and Vehicle Licensing Authority (DVLA) was formed in 1973. Prior to 1973 all driver and vehicle details were handled by local authorities and the details would have been held on paper records. All the paper records from the local authorities would have been passed to the DVLC.
By my calculations if these records had been transferred to computer at that time they would have needed about 1,150 tapes.
My first encounter with computing was in the mid 1960s when I was at Sussex University gaining a BSc in Maths and Physics. I attended a one-term course learning about computer programming. I wrote a Fortran program to solve quadratic equations using a formula.
The university didn't have its own computer and so our coding sheets were loaded into a computer elsewhere. I don't know where.
In the early 1970s when I was doing temporary work for the Conduit Bureau I did clerical work for Great Universal Stores who were installing a Honeywell computer. They were transferring accounting data from a ledger card system maintained with manually operated machines. I don't remember exactly where in London this was.
In 1976 I joined the newly-formed government sponsored Training Opportunities Scheme (TOPS) which was run by International Computers Ltd (ICL).
ICL was the leading computer company in the UK and was formed in 1968 by the merger of three smaller companies: International Computers and Tabulators (ICT), English Electric Computers (EEC) and Elliott Automation. This was as a result of intervention by Harold Wilson who was Labour Prime Minister for eight years.
There is a detailed chart at the end of the page which shows the full history of ICL from 1950 to 1990.
The government had a 10% stake in ICL. In a famous speech in 1963 Harold Wilson warned that to prosper, a “new Britain” would need to be forged in the “white heat” of scientific revolution.
It is nice to know that I was involved in a scientific revolution.
This retraining scheme completely changed my life and my career path and I subsequently worked in computing until I retired in 2003.
I started the TOPS course on 16th February 1976. I had ten weeks of training at Beaumont College, Old Windsor in Berkshire, ending on 23rd April. There were just eight of us on the course.
Beaumont was at that time a training college owned by ICL. I stayed there during the week and went home at weekends. The living accommodation was good. The training and catering facilities were awesome. All meals were silver service.
The modules of the course were:
• Introduction to Computers
• Computer Operating
• Computer Programming
• Business Systems Analysis
• Systems Design
The computer that we used on the course was an ICL 1903.
ICL Beaumont College, Old Windsor, Berks in 1976.
More views of the college.
My fellow course members.
My room at Beaumont.
Evening meal at Beaumont.
In the 1970s systems analysis and design was all about building computer systems for companies that didn't have any computers. This meant looking at clerical systems that were paper based. So systems analysis meant learning how data moved around the company and how it was processed manually. In the early days there were few bespoke packages.
The business systems analysis and systems design courses that I took ended with a couple of days of role playing.
The eight of us on the course were divided into two teams of four. Various members of the ICL training staff had offices around the college and each played a role as a staff member of the “company” we were investigating.
Members of our team had to go and interview them in their roles. In this way we got an understanding of how the “company” worked. This was all done to a very tight schedule.
We in the team then had to put together a full analysis of the data flow within the company and design a complete computer system.
This meant brainstorming sessions into the night on a couple of occasions to prepare a “presentation” of the proposed computer system to the “company” management who were all in role.
For the presentation we were working at the front of a lecture room with blackboards, flip charts and an overhead projector with acetates. We were dressed in our best suits, facing the management staff who were sitting in the front row in front of us. The other team of four were sitting at the back of the classroom, watching the proceedings.
It was a very stressful situation even though it was all “acting”. I don't remember how long it took but it seemed like a lifetime!
I vividly remember how at the end of the presentation when we had finished our sales pitch and when the managers were about to get up and leave, they wondered why we were suddenly trying to suppress laughter.
They turned round and there at the back of the classroom were the four members of the other team holding up voting cards with numbers on.
Within the course, twelve weeks of work experience was to have been at the Tesco Computer Centre at Cheshunt in Hertfordshire. Because I wanted to move away from London I managed to get a placement at Plessey in Sopers Lane, Poole.
This started on 26th April 1976. I stayed in Poole during the week and returned home to London at weekends.
On 9th August 1976 I went back to ICL Beaumont for a final week of training and passed the National Computing Centre (NCC) orals on 12th August 1976.
Plessey Poole 1976.
The Plessey building is the long building in the centre of the picture.
At the end of December 1977 I moved house to Upton in Poole. I was able to walk to work across Upton Heath to Sopers Lane.
Plessey Poole 1976 showing Holes Bay in the background.
I started work officially for Plessey in Sopers Lane, Poole on 16th August 1976 sitting at the same desk and doing the same work as I had been doing two weeks previously but now receiving a salary. My telephone number was Poole 5161 extension 2081.
The computer centre was at Addlestone in Surrey. I never actually visited Addlestone so I never saw the ICL 1906S computer which we used.
[ Incidentally, Plessey at Addlestone ran one of the first computer installation laser printers and amongst other things they printed football pools coupons.]
Live running of batch programs was generally overnight. When I was responsible for a batch program such as payroll the computer operators would have my home telephone number. I can assure you that it is no fun being woken at three o'clock in the morning by a grumpy voice at the other end of the line saying “Your program has just halted with the message 'stop 1234'. What do we do?”.
I very soon learnt structured programming and how to make programs have far fewer error stops and give meaningful error messages.
The first Cobol program I was given to work on was totally unstructured - what we called monolithic programming. It had lots of error stop points. The error message was “ZZ” at all points.
Plessey Poole, when I started work there, produced traffic signals. Old style traffic lights were still being manufactured there known as “Tin Sigs”. They were still having to support those which were operated by clockwork.
A teletype machine with keyboard and printer roll.
A modem to connect to a computer over a copper dial-up phone line.
There were thirteen of us in our office. We had no computer access at our desks. My main equipment was pencil, paper, sharpener, rubber and Cobol coding sheets. We also had a couple of hand punch card machines.
Next door was the punch room where a number of girls operated card punch machines. There was also a machine called a verifier which printed the actual readable characters across the tops of the punch cards.
Our only access to the computer was via a teletype machine. This was connected to the mainframe using a telephone and acoustic coupler and modem. The speed was 300baud i.e. just under 40 bytes per second. We could not run live programs on the machine from our office.
ICL Cobol coding sheet and punched card.
An 80-column punched card. Each column represented a single character.
An ICL card punch for punching a card by hand. Up to four keys had to be pressed down at the same time to produce a single column.
A card punch machine for typing batches of punched cards.
An ICL EDS60 disc pack.
With the advent of exchangeable discs it was possible to have files with indexes so that data records could be accessed directly. I worked with various indexed file structures and then with database systems before relational databases became available.
With basic indexed files it was necessary to structure the file to minimize head movement on the disc which was done by “striping” i.e. arranging the data in “cylinders” so the file was arranged on all surfaces of the platters in a cylinder which could be accessed without head movement.
In an ICL machine files the units of data storage were “blocks” and “buckets”.
An ICL EDS200 disc unit - the size of a washing machine.
Projects I worked on at Plessey:
The first system I worked on was a Cobol Bill of Materials (BOM) system. This was a serial processing magnetic tape system which we converted to use indexed files on disc.
The data held was the structure of the traffic management machines and the system could generate shop floor pick lists based on the quantities and styles of equipment for a particular order.
Program names under ICL George III (“George Three”), the operating system we worked with, were only four characters long. We named our programs “General Resource” programs hence GR01, GR02, GR03 etc.
I really enjoyed using George III and I became quite skilled at writing George III macros.
There is a great website here with lots of information about ICL 1900 hardware and software: www.icl1900.co.uk
Other systems I worked on both as a programmer and as a systems analyst included:
• A stock control system maintaining records of electrical equipment held at depots around the country for the maintenance and installation of road traffic control systems. The system handled stock delivery, usage reports and annual stocktaking. Using the method of exponential smoothing I wrote programs to report on predicted demand for equipment based on previous usage.
• Estimating and planned standards. I enhanced an existing system to give better costing information of electronic equipment based on historical costs of components and manufacturing processes.
• Cost forecasting and cashflow management. I was the main analyst/programmer involved in the development of the system.
• Payroll. The Company used the ICL package COMPAY and I developed a number of ancillary programs using Cobol. Reports were also produced using Filetab, a decision-table based software package which on ICL 1900 machines was called TAB-N.
In May 1981 I changed offices and moved to Plessey at Christchurch, a few miles away from Poole.
Plessey Christchurch 1980
As part of my training on the TOPS course I had to produce a report of my progress at Plessey. This is what I wrote:
G. R. Newland. Report on my first six weeks training at Plessey, Poole. 6th June 1976.
Some 2000 people work at the Poole site, the majority being employed in the Traffic & Instrumentation Division (T & I). My training is taking place in the Information Services Department, which employs a total of 27 people of whom thirteen work as analysts or programmers.
The Information Services Departments at Plessey sites are administered centrally from headquarters at Ilford in Essex (now Greater London). Although the DP department at Poole is only small, I have already received experience in a wide range of Systems Analysis activities.
There is no computer at Poole, but we are on-line to an ICL 1906S at Addlestone in Surrey via vdus as well as typewriter terminals and a card reader and paper tape readers. The computer runs under George 3 (now being upgraded to George 4) and uses the MOP system.
The main work of the Poole Information Services Department is concerned with T & I, and this is the area within which my project work comes.
There are a number of program suites which are used regularly in connection with stock recording, manufacturing and purchasing. However, these suites have a number of shortcomings and are rapidly becoming inadequate for the work in hand. In addition, the programs were written for use on an ICL 1903 computer and thus do not make full use of the facilities now available.
I spent some time familiarising myself with the User Specifications for the existing systems. Unfortunately the existing documentation is not up-to-date, and it appears that for one reason or another many proposed systems were never implemented.
Work has been under way for almost a year on a new 'Bill of Materials' (BOM) system, and this now involves all the programmers in the Information Services Department.
My project is part of the BOM system. However, it has also involved implementation of small systems within the existing set-up, in preparation for the BOM system.
The areas I have been working on are concerned with preselection, that is the selection of piece-parts or sub-assemblies so that the workshops can build assemblies required by customers. My work has included Preselection Labels, for use in identifying kits, Draft Variable Headers, for forward loading the shop floor, State of Preselection reports and Preliminary Works Order Planning reports.
These are mostly enhancements to the BOM system as it was originally planned, and as such have had to go through all the normal stages associated with new systems.
Initially I was involved in discussions with the Users to decide exactly what was required, and I also conducted my own interviews to agree document layouts and procedures. I then drafted out a User Specification which is now in the pipeline for acceptance by the Users.
Other tasks I have carried out include the designing of forms, preparation of punching documents (drawing up layouts with special pens for copying onto printing plates), writing a couple of small programs and writing program specifications. Whilst all this has been going on I have also been reading as much as possible and discussing the systems being planned, to try to get an overall concept of the Information Services organisation as a whole.
My next project will be concerned with the implementation of the new BOM system, and this has already started with the drawing up of forms and punching documents.
Prior to Plessey taking over the site in the 70s, it was the Ministry of Defence Signals Research and Development Establishment (SRDE). I worked in Z block (see picture - long building on left).
I took on responsibility for the Configuration Management system used to control the Ptarmigan Project (a large defence contract for which Plessey was the prime contractor). The system was a modified version of one designed for the System X project. System X was a digital switching system installed in almost all telephone exchanges throughout the United Kingdom from 1980 onwards.
The configuration management system was originally written as a pseudo database system using indexed files on ICL 1900 and was later converted to IDMS.
Before relational databases appeared, IDMS was a very useful database system but it required that you knew the size of the database when building it. Extending an IDMS database was extremely difficult. This meant writing special programs to unload the database and reload it.
I still have my little rubber stamp used for making Bachman diagrams. (Google it!)
I designed and wrote a number of programs as well as specifying and implementing major schema changes.
In 1982 I designed and wrote a suite of programs to interface with the manufacturing database, also on ICL 1900.
In the same year the Company purchased a Datapoint 1800 system with four screens, which was used for data capture for the configuration management system. I wrote all the Databus programs and set up the RJE emulator to interface with the ICL DME system running on an ICL 2900 machine.
Also in 1982 I took over the support of the Ptarmigan incident reporting system which recorded details of fault reports on electronic equipment both during testing and while in use by the Army. The system had been written in Fortran and I specified the conversion to indexed files using Cobol. With the help of another programmer we rewrote the system.
At the end of 1982 I took up some responsibility for another IDMS configuration management system which was being converted from ICL 2900 DME to VME. I wrote much of the SCL for the batch processing side of the system.
In 1983 Plessey Defence Systems purchased its first HP3000 minicomputer running the MPE operating system and this was housed at Christchurch.
The main application was a project control system to manage material movement amongst subcontractors on the Ptarmigan project. I was involved in defining the hardware requirements, supervising the setting up of the communications links across the site, and installing the computer and peripherals.
We rather stretched the capacity of RS232 connections because some terminals were a lot more than 50 feet from the computer.
It is amazing the think that an HP3000 computer was able to handle our multi-user system. HP3000 data buffering was very poor. Every time a key was pressed on a connected keyboard anywhere it caused a main processor interrupt.
The Company purchased the HP MM3000 materials management package and I was solely responsible for extensive customisation of the software and its final implementation. Over the following two years I designed and wrote an auxiliary database system in Cobol to run alongside the main MM3000 system.
In 1984 a second HP3000 computer was purchased with MM3000 and PM3000. As systems manager, I was again heavily involved in the implementation. I designed and wrote a Master Parts File system using Image and Cobol, which interfaced with MM3000, the Configuration Management systems on ICL DME and VME.
We also started to use re-badged DEC Vax computers (possibly Systime) and I worked on a data collection system and a CAD system on Computervision.
I also set up a data entry system using Entry, V3000 and Reformat which allowed data collection and manipulation for the Configuration Management system. It replaced the Datapoint system developed earlier and using a “black box” and DSN/RJE on the HP3000 machine the system automatically transferred files to the ICL mainframe.
In 1984 Plessey Defence Systems bought POWERHOUSE (QUIZ, QUICK, QTP) and I was alone responsible for the design, development and implementation of several systems using prototyping, principally:
• Documentation Control. The system allowed the recording of estates, reference documents and descriptions within a number of manufacturing baselines. Numerous QUIZ reports were used in the system.
• Repair Work Order Monitoring. The pricing and control of equipment repair work orders was facilitated by the system. The data was eventually transferred to the main ICL Configuration Management System.
• NATO Codification System. Data on magnetic tape produced by a government codification agent on an ICL ME29 minicomputer system was read on to an Image database and online enquiry and off-line printing was available.
• Government Equipment Accounting. The Company was responsible for maintaining a register of materials supplied by the MOD for use on military projects. The system replaced manual records and had to be subjected to a rigorous audit procedure before being authorised by the MOD as an approved system.
• Reference Centre Stock Control. The Ptarmigan Reference Centre, run by Plessey Defence Systems, controlled the distribution of electronic equipment as well as its testing and repair. The system allowed the tracking by serial number of individual assemblies and components as they progressed through inspection and testing phases. Daily stock receipts and issues were recorded and a number of printed reports produced.
• Contract Control. The system allowed the recording of contract details such as timescales and prices. It maintained an amendment register and cross-reference index.
In 1985 the Personnel department purchased an ICL DRS20 system consisting of a model 150, a 125 and four 110's. These were connected on a Microlan network. I was solely responsible for the setting up and customisation of the Personnel 20 package and the transfer of an existing ICL DME system. I subsequently designed and wrote a number of programs using CIS Cobol to provide additional facilities not available in Personnel 20.
In 1985 Plessey started to use ICL OPD (One Per Desk) personal computers. These were based on the Sinclair QL home computer.
For a shorttime I worked on the OPD.
I set up a company called Terra Nova Systems Ltd with just two directors – my wife and me.
From July 1986 until December 1986 I took on my first contract with Factory Mutual Insurance in Victoria Street in London. I worked as part of a team developing and enhancing a very large insurance system using Powerhouse on HP3000/68.
The system ran on several machines around the world, all linked to the London machine. I was mainly concerned with the loss adjustment side of the system and designed and wrote many programs in QUIZ, QUICK and QTP. The system was written totally in Powerhouse and incorporated some of the most sophisticated use of the package in the country.
I also wrote some report programs using PRO*COBOL.
The Company also had systems running on some very ancient Datapoint equipment using Datashare. I maintained the systems where necessary as they slipped slowly into retirement!
From January 1987 until February 1988 I worked at BOCM Silcock, a Unilever Company, in Basingstoke. I was a member of a small team designing and writing a system in ORACLE on VAX. The system controlled contract maintenance and scheduling at a seed and grain processing mill in Yorkshire.
I had been given the contract because of my good experience of using Powerhouse. By the time I got there the decision had been made to use Oracle relational database and not Powerhouse. I was handed an Oracle manual and told to get on with it!
Early development had to be very much revamped when version 2 of SQL*PLUS and SQL*FORMS became available about four months into the project, in order to utilise the many new facilities.
Broad Quay House, Bristol in 1988.
The Oracle office where I worked in 1988 with my jacket on the back of my chair.
Between February 1988 and July 1988 I worked for The Oracle Corporation in Broad Quay House, Bristol designing and building a package called Road Maintenance Management System (RMMS) which was to be marketed to local authorities for use by their highways and transport sections to manage road maintenance.
The project involved a fair amount of design work using SDD and the analysis of an extremely complex system specification produced by The Department of Transport using SSADM. My work was split between design and development.
The system was developed mainly on DEC Vax but was destined for running on a Prime computer.
From August 1988 until February 1989 I worked for United Agricultural Merchants (UAM), a Unilever company which deals in grain and other farm commodities, designing and programming a sales ledger system using Cobol. The system used the Image database management system and a number of sophisticated programming aids which mean that program development times were short.
The system ran on a number of HP3000 machines around the country. All were linked by a network system.
From March 1989 until July 1989 I worked for MVM Consultants Ltd in Bristol and was involved in designing and developing programs using SQL*FORMS as part of a package which was being installed in local authority sites up and down the country as part of the computerisation of the land registry.
Solicitors who had PCs connected to the main system by a national network were able to enter search details which were processed by the authority on a VAX computer. The replies were sent back via the network. I was also been involved in the introduction of SDD into the development and much use is made of CMS.
From August 1989 until December 2003 I worked for (Education Management Information Systems (EMIS) a software house near Bristol on a package which is marketed to Further Education colleges nationally.
I was involved in design and build of programs using SQL*FORMS, SQL*PLUS and SQL*REPORT. The company transferred to ORACLE version 6 in 1990 and I played a strong part in code conversion and training of programming staff. I used CASE*DICTIONARY for recording design work.
I worked in a team of about 15 programmers. I had a keen interest in maintaining good programming techniques with code written to precise rules. I was responsible for setting up the company's programming standards with strong emphasis on code reviews and quality control. Program source code control was achieved using SCCS and CMS.
In December 1991 I became a permanent member of staff at Emis and ceased to be a contractor. I closed down Terra Nova Systems Ltd.
In January 1992 all software developers received our first DEC workstations.
Three years later in August 1995 the DEC workstations were thrown out and IBM compatible PCs were moved in. Microsoft Windows NT arrived!
1980 I built a Sinclair ZX80 computer from a kit.
![Apple ][](images/apple2.jpg)
1983/1984 Apple II
1987 Amstrad PC1640 which cost £1200 + VAT. Software included Gem and Fleet Street Editor.
1991 Amstrad PPC1640. My first “laptop”.
1997 Tiny PC.
1989 using my Amstrad computer and 300 baud dial-up modem I did some moonlighting for UAM working from home.
• 1980 November: COMPEC Computer Exhibition at London Olympia.
• 1995 October: I bought my first mobile phone - a brick!
• 1998 January: I first connected my home PC to the internet by dial-up modem.
• 1990 11th October: I went to my first Computer Conservation Society meeting in London.
On several occasions I went to the Oracle User Group Conference in Birmingham.
• NCC Basic Certificate in Systems Analysis - August 1976
• JCL Job Control Language - October 1980
• IDMS database - June 1980
• Digital DEC VAX/VMS Utilities and Commands - October 1984
• Digital DEC VAX/VMS System Management - November 1984
• SQL*Forms V2.3 to V3 Transition - July 1991
• Application Tuning under Oracle V6 - January 1992
• Oracle Transition to Version 7 (Forms 3) - November 1992
• System Design with Oracle CASE - January 1993
• Oracle 7 New Technical Features - April 1994
• Oracle Developer/2000 Core Technology - March 1996
• Oracle Developer/2000 Forms 4.5 Essential Techniques - March 1996
• Oracle 7 Procedural Capabilities - April 1996
• Oracle 7 Administration - December 1997
• Sysdeco Dynamic Systems Development Method (DSDM 3) - February 1998
• Oracle on Windows NT - June 1998
• Oracle Developer 6 New Features - March 2000
• Installshield Development - November 2001
• QA Application Development using Visual Basic V6 - March 2002
The history of International Computers Ltd (ICL)
