April 1969: 100,000 missing passengers at Dublin Airport
Easter fell on the first Sunday in April 1969 and the disaster struck just before the holiday weekend. It occurred in the Aer Lingus computer department and was caused by a mistake in the allocation of data storage space on a bank of disk drives. The error was discovered late at night after all of the airline’s passenger name records – about 100,000 in total – had been overwritten with system messages. This process could not be reversed.
The Aer Lingus reservations system, which was known inside the company as Astral, required not just one, but two IBM mainframes. It had recently gone live following the largest computer investment that the country had seen to date.

Sales agents Dympna O’Flynn and Mary Lennon using Astral terminals.
As a rule Irish businesses in the 1960s were late adopters of computer technologies. Even the largest organisations made do with machines from the low end of the manufacturers’ product ranges and seldom looked beyond accounting applications. Computing was still something that engineers and programmers practised out of sight of ordinary folks, entering arcane messages into typewriter-like devices in order to converse with their systems.
Aer Lingus, in contrast, had introduced leading-edge data communications and become one the first organisations anywhere to implement IBM’s IPARS software – a sales and passenger records application. Its frontline employees could access the system from more than 200 user terminals whose screens displayed up to 1960 characters. This style of computing was exceptional because it transformed the everyday work of staff who dealt with customers.
The airline liked to show off its data processing muscle. Visitors to the Aer Lingus head office at Dublin Airport could view its IBM System/360 computers and their rows of disk drives through a glass wall in the reception area of the building.
IPARS was an enhanced version of a reservations system from the US – the ‘I’ in its name signifying ‘international’. In 1967 Aer Lingus was one of five European airlines – along with Alitalia, BOAC, KLM and Swissair – that committed themselves to implement the system. IBM relocated about 200 software developers to an IPARS project office in London and began to ship its new generation of mainframe computers to the airlines. A pair of System/360 model 50s arrived at the Aer Lingus headquarters before the end of 1967.

Aer Lingus showed off its computer room through a glass wall.
Dick Cahill was the manager responsible for this project at IBM Ireland, where Joe Cunningham oversaw a local technical team that included Kevin O’Brien, Declan O’Riordan, Robert Poynton, Paul Kilcullen, John Holland and Stephen Ellis. After IPARS went live at Dublin Airport IBM delegated some of its hardware engineers and systems software specialists to keep operations running there.
David Kennedy was the systems manager in Aer Lingus. He had persuaded the company to implement the IPARS software and he also came up with the Astral name – an acronym from ‘Advanced System of Telecommunications and Reservations for Aer Lingus’.
Denis Behan became the airline’s main liaison with IBM. Paddy Byrne was in charge of the Aer Lingus data processing department and Liam Cassidy was its operations manager. Brian Ennis and Emmet Wilson held senior software development roles. In total the airline employed about 30 programmers and system analysts, who were based on the fourth floor of head office.

John Cronin seated in front of the operator’s console on one of the IBM mainframes.
Aer Lingus looked to the United States for computer room expertise, recruiting Americans Ron Chaffee and Ron Schlitt along with Irish expatriates Pat Cormican and John Cronin. It soon employed sixteen computer operators who worked in teams of four, supported by more than 40 punch card equipment operators and the communications engineers who managed the online connections. It also provided Astral training to the reservations controllers who were in charge of the flight schedules and to the sales agents who would capture passenger details for the system, keying in codes from a prompt sheet.
The new booking procedure was still not fully computerised. In the early years of Astral there was no way of printing tickets at the reservations desk. When the staff had recorded a sale through the computer terminal, they would write out the customer’s ticket by hand.
In autumn 1968 BOAC became the first of the IPARS users to start making reservations through the system. Aer Lingus claimed to have the second implementation, but it never held a formal launch event or switch-on celebration. No one knows who the first Astral-enabled customer was.
Some project veterans suggest that the reservations system was not fully operational until April 1969. Others point out that a first batch of sales agents at the airline completed training for Astral in the previous November. Some Aer Lingus employees, moreover, were still using an older reservations system – a ‘teleregister’ that indicated seat availability but could not capture passenger names – in the second half of 1969.
It would appear, therefore, that Astral was phased into service as one group after another completed user training and as one location after another was connected to the mainframes.

Aer Lingus marketed the expertise in its computer department to other airlines. This 1975 brochure features (l-r): Conor Campion, Maurice Wallace, Bill Thompson, Patricia Brennan, Pat Cormican, Gerry McGrath and Brian Kirby.
Each of these links was a landmark for the project. The Astral network connected the Aer Lingus headquarters with sales offices and airports in eighteen cities in Europe and North America. It employed special modems that made allowance for industry-specific protocols and national differences in telecommunications standards. Each premises required a hardwired computer to drive the local screens and every configuration was different. When IBM inadvertently installed the Glasgow machine in Birmingham, the company had to take it out and transfer it to the correct office.
The just-before-Easter emergency threatened to bring this entire infrastructure to a halt.
Aer Lingus and IBM both claim credit for resolving the problem. The computer firm called in operating system experts from around the world to examine what had gone wrong. The airline was able to retrieve the lost data from the very basic details held in a separate passenger name index. Eight members of its computing department managed to reconstruct the 100,000 passenger name records in a marathon session that lasted for around 24 hours.
The norms for hardware reliability and service uptime were rather different 50 years ago from today. All systems crashed more frequently in those years. Aer Lingus software veterans recall fixing programs on the fly, taking risks that would be unthinkable in the 21st century. Astral, indeed, collapsed on a second occasion about five years after the April 1969 incident.
As an early adopter of online reservations Aer Lingus not only gained prestige within the travel industry but also opened up new business opportunities. By the mid-1970s the company was developing add-on software applications and assisting other airlines across the globe to automate their booking systems. Its Cara subsidiary had also become one of the largest computer service providers in Ireland. Aer Lingus described itself in a 1975 marketing brochure as ‘probably the most highly computerised medium-sized airline in the world’.
The functionality of the Astral system is curiously familiar to modern eyes when it is compared with the intermittently live, teletypewriter-controlled, one-job-at-a-time applications that other organisations ran in 1969. Likewise, the rapid response to the Aer Lingus data loss was truly dramatic to those who were involved, but now sounds like a normal operating procedure.
What happened at Dublin Airport half a century ago was a foretaste of the future – an early instance of an always-on, network-dependent information environment. That is why the introduction of Astral was a major achievement – especially for an installation whose twin computers had a combined memory of just 512K.
Dates To Remember
04 September 1998: ‘The culture was always to “think big” and go for it’
24 November 1997: Inauguration@Cork
September 1997: Online banking – Ready if you want it
September 1992: Glockenspiel falls silent
October 1991: ‘This is text messaging’
17 June 1991: ‘The internet is now available for testing’
19 March 1991: Three software developers. One desk. No chairs.
09 October 1989: E-commerce pioneers assemble
26 January 1984: Hello Macintosh
18 January 1983: ‘Your friendly IBM Personal Computer’
April 1969: 100,000 missing passengers at Dublin Airport
September 1966: Paper-based computing comes to Guinness
December 1960: The first computer installation in Dublin
23 February 1909: A proposed analytical machine
27 July 1866: ‘It is a great work a glory to our age and nation’
Brian Dent
Aer Lingus computer operator
Two of us trained with IBM and British Airways in Boadicea House at Heathrow Airport before coming back to Dublin to run the Aer Lingus units. We then had to train the remainder of the operations people on the IPARS system.
I do not remember any big cut over day for Astral, it was a gradual process of cut over of central reservations and outlying offices.
In 1971 I moved to programming on Astral in 1971, working in the message switching functions of the system. I spent 1983 and 1984 rolling out IPARS to Kuwait Airlines and then returned to Ireland to join the Timas project, which enabled travel agents to make online reservations.
Pat Cormican
Aer Lingus computer operator
I had previously worked on IBM mainframes in New York. When Aer Lingus bought two IBM System/360s it recruited three of us to run them in shifts. I joined the company in 1968 as a senior console operator.
Another operator, John Cronin, had computing experience in the US as well. Aer Lingus asked us to train in local lads so that they could also do this work.
Later on I became a supervisor at the installation and worked in operations management.
John McEneaney
Aer Lingus programmer
The Astral development teams were organised around functional areas. My primary area was ‘inventory’ and my secondary was ‘teletype input and teletype output’. This handled inter-airline (and specifically inter-system) communications, emulating the AIRIMP messaging system, with which I had become very familiar when I worked in reservations control.
Aer Lingus had unbelievably forward-thinking ideas on training and development. In the IPARS project we had weekly afternoon training sessions where someone (in turn, and usually, junior on the team) would give a short talk on their secondary area of expertise. Since it was the area I knew less about, I had to do a lot of research before trying to impart knowledge the rest of the team, many of who would already know the subject far better than I did! It was really great training.
Michael O’Duffy
Aer Lingus programmer
I worked on Astral in a junior capacity. Ten of us were recruited from within the airline as programmers for this new online system and within the data processing department we were considered to be particularly fortunate. There was major interest among the staff in joining this new technology era and many hundreds applied. The selection process was a combination of aptitude tests and interviews.
This was followed by six months of intensive training on the fundamentals of programming with fortnightly tests where we were expected to achieve A level results. This education was handled by IBM personnel from the US.
The system architecture was particularly complex and had features that I did not experience since then despite working on many innovative projects. It needed to cope with a high level of temporary or short life data. So it introduced the concept of data pools with short lives and large and small data records with long lives.
The reservations function had been highly manual until then. I recall that it shocked me when I realised that 300 staff would become redundant. They did not lose their jobs, but were redeployed.
Seamus Gallen
Aer Lingus programmer
Aer Lingus generally saw the new system as an upgrade. Most people there did not realise how much of a jump it was in technical terms.
The general staff didn’t quite understand the difference between a system updated daily (at best) and an online, real-time system. With the old Bunker-Ramo system – which would barely qualify as a computer now – several telesales people could sell the last seat. That could no longer happen on Astral.
On one occasion when the operator forgot to cycle forward the date, I got a call from reservations to say that ‘the computer thinks it’s yesterday’!
John Fullerton
Astral user
Astral was the internet of the day. For example, it was possible for users to have what would now be called a web chat.
There were two types of screen messages. A general message was intended for later action, while a supervisory message was instantaneous and could not be screen saved. It was screen-to-screen only and was free script. Its purpose was to get something actioned immediately, such as a request to overbook a flight or to put on an extra flight due to demand.
But it could also be used just for a chat with a colleague in somewhere like New York.
Photographs courtesy of Nora O’Rourke, Tom Marum and Conor Campion
© Newsmail 2019
29 April 1968: ‘A stimulating environment in which to exercise the highest professional competence’
Ireland’s first software company had deep roots in the IBM tradition. System Dynamics was led by Paddy Doyle, a former systems manager at IBM Ireland, and its engineers and programmers had strong technical reputations for work that they had done at the corporation.
From its inception the Dublin firm developed bespoke applications for IBM mainframe computers and it was not long before it attempted to design and sell a proprietary software package. The primary aim of System Dynamics, however, was to siphon away part of IBM Ireland’s revenue stream.
The formal registration date for the new company was 29 April 1968. But the personal connections behind System Dynamics were forged five years before the Companies Registration Office logged its existence.

The cover of an early company brochure, featuring artwork by Anne Kennedy.
In 1963 Raymond Girault, the general manager of IBM Ireland, assigned Paddy Doyle to lead the development of a reservations system for Aer Lingus. This was a major undertaking for IBM Ireland, so he assembled a group with international experience. Technical manager Fred Kennedy and Joe Cunningham, who determined the airline’s functional requirements, had both worked in IBM installations in New York. Fred Kennedy selected Brendan Byrne (another IBM New York veteran), Tom McGovern (who had previously worked as an electrical engineer for Radio Eireann) and Kris Padmanabhan as his program design and analysis team.
Aer Lingus rejected the planned reservations system in 1964 after IBM announced its new System/360 generation of computer technology. The airline was no longer willing to accept software developed for an older platform.
IBM’s project group split up. Paddy Doyle turned to consulting and started lecturing at UCD. Fred Kennedy moved to IBM in France, where he led a project that developed the first fully electronic voice and data switch. Tom McGovern stayed with IBM Ireland. He also studied part-time at UCD, which had introduced the first MBA degree course in Europe.
The members of the software team had got to know each other, and their capabilities, when they worked on the Aer Lingus job. So they stayed in contact.
Software was not yet regarded as a commodity in the mid-1960s and applications did not yet have price tags. The big computer vendors were one-stop-shops that installed and maintained computers and storage equipment and printers, selected and trained technical staff for each installation and even delivered whatever stationery their users required. They developed customer-specific software or taught the customers how to write applications for internal use.
IBM routinely treated software and support as free giveaways for the purchasers of its very expensive machinery. It was common, therefore, for programmers to exchange code with their counterparts in other organisations. They knew the cost of a reel of blank tape, but an application stored on that tape had no apparent value.
As the years passed, however, computing professionals began to question these conventions. Rival hardware makers in the US lobbied against IBM’s effective monopoly on ancillary services for its systems. Independent software vendors and technical service providers began to surface in different parts of the world. Some even attempted to sell replacement software for programs that IBM supplied free of charge.
In 1967 Paddy Doyle, Tom McGovern, Brendan Byrne and Joe Cunningham decided to create a company that would offer an alternative to IBM in Ireland. Soon afterwards Tom McGovern travelled to Paris and discussed this plan with Fred Kennedy and Raymond Girault, who had moved to IBM’s European headquarters there. Kris Padmanabhan was also made aware of what was happening.

62 Northumberland Road: The company’s first address in 1968.
System Dynamics began operations in April 1968 with a board of non-executive directors and initial funding from a group headed by Professor Michael MacCormac, who had set up the MBA course at UCD.
Within three months the start-up had contracts in the bag from three customers – Irish Life, Irish Ropes and the Irish Management Institute. In August the company opened its first office at 62 Northumberland Road, where Paddy Doyle, Tom McGovern and Brendan Byrne were joined by Nick Spalding (another systems programmer from IBM) and secretary Maria O’Regan. Joe Cunningham, meanwhile, had decided to remain with IBM.
In the words of an early customer presentation, ‘The System Dynamics company was founded to create a stimulating environment in which to exercise the highest professional competence in providing much needed services to the information processing community.’ In practice it eked out an unsteady existence by hawking a variety of consulting and systems implementation services for existing and prospective computer users.
By now ‘unbundling’ the costs of systems, software and services was the talk of the computing trade. In December 1968 IBM announced its intention to introduce separate prices for the first time for the different components of its computer business. In the following month the US Justice Department filed an antitrust suit against the corporation. In June 1969 IBM announced new prices for its software products and for its systems engineering, equipment maintenance, customer education and custom programming services.
This changing industry environment improved the commercial prospects for System Dynamics. The company was now entering its second year, but the going was tough. It had met its annual sales target of £20,000, but Brendan Byrne had left the organisation and Paddy Doyle was preparing to depart as well. He had already stepped down as chief executive. Fred Kennedy and Kris Padmanabhan were drafted in over the following months.
System Dynamics overhauled its board in 1970. All of the original directors stood down. Tom McGovern, Fred Kennedy and Raymond Girault, who had retained an apartment and a Rolls-Royce in Dublin, replaced them.
Tom McGovern and Fred Kennedy became joint managing directors with equal shareholdings of almost 50% each. The company relocated to a new base at 72 Merrion Square. It now employed ten staff and its annual revenues had risen to around £35,000.
Tom McGovern was responsible for technical sales and Fred Kennedy took charge of project management. After Ireland joined the EEC he also set up a small System Dynamics office in Brussels to explore the emerging opportunities in European institutions.
This arrangement lasted until 1979, when Fred Kennedy left to launch Malahide-based Computer Applied Techniques. Tom McGovern remained managing director of System Dynamics, as well as the company’s majority shareholder, until his death in 1994. He always gave priority to consulting projects. The company also became known as a provider of technical staff to the larger user installations.
Hence System Dynamics is not remembered as a software product developer. During its early years, however, the company attempted to deliver the first Irish-made commercial software package. Its name was POEMS and it was intended to support factory management. This project was mentioned in company documents from about 1970, but it did not result in a product launch. It is possible, however, that a later version of the software supported a modelling exercise at the SPS International machine tool factory in Shannon.
System Dynamics envisaged POEMS as ‘a program for use in the Manufacturing Industry which is used to analyse production line organization and identify improvements which may be made under the application of Group Technology concepts’.
Judging from that description, the product was probably not easy to sell !
© Newsmail 2018
September 1966: Paper-based computing comes to Guinness
One of the least well known craft traditions at the St James’s Gate Brewery flourished for a brief spell from from the mid 1960s until the early 1970s. It involved cutting and splicing strips of paper tape with rows of holes punched through them. Guinness had to master this technique in order to operate its first computer: an ICT 1902 that arrived at the brewery 50 years ago.
The primary method of information input for the 1902 was an electro-mechanical device that read paper tape at 300 characters a second. Most installations in that era programmed their computers or entered data for processing by feeding stacks of punched cards into the systems. Guinness selected a model that was driven instead by paper ribbons. Modifying the company’s software, therefore, meant editing the paper tape. When staff needed to correct or amend their code, they had to alter or replace those perishable strips.
Half a century afterwards, programming a computer in this way sounds as impractical as mining coal with household cutlery. Guinness, nonetheless, assembled a team that became very proficient at the new paperwork.
The company had spent several years evaluating electronic computers before deciding that the technology was sufficiently reliable and affordable for its purposes. It finally placed an order with International Computers and Tabulators (ICT) in June 1965. The manufacturer promoted the 1902 as one of its small machines. But the system at the brewery was the most powerful model that ICT had sold in Ireland.
Much of the planning and preparation at Guinness had been done by a group of statisticians who reported to the head brewer. They had access to an Elliott Automation computer at An Foras Talúntais / The Agricultural Institute, where they had started to explore scientific applications back in 1962. These experiences enabled them to assess potential tasks for a computer inside the brewery.
It was, however, the Guinness accountants’ department that oversaw the introduction of the ICT 1902. The first computing manager at the brewery was Pat Stephenson, whose job title was ‘accountant-in-charge of computers’. He spent five years in this role, then went back to full-time accountancy, remaining at Guinness until he retired in 1987.
The accountants’ department provided accommodation for the new technology. It converted a group of small offices into a computer room, tearing down the partitions that had previously divided up this workspace. It installed air conditioning to regulate the environment, because changes in temperature and humidity could upset the performance of the equipment. It also imposed strict controls on access to the computer room. Mere programmers were seldom allowed to step through the door.
Fifteen months after Guinness ordered the ICT 1902, it installed the computer in September 1966, along with the paper tape reader and punching machines, a set of four magnetic tape decks for file storage and a floor-standing printer.
ICT’s Dublin office assigned one of its system engineers, Darragh McArdle, to help get operations started. But Guinness was soon technically self-sufficient, thanks to the in-house computing team that it assembled.
One of the new recruits was Don McInnes, who subsequently replaced Pat Stephenson as head of the group. He went to work at St James’s Gate shortly before Guinness took delivery of the 1902. He had qualified as an accountant, then moved into computing operations at Tate & Lyle in London. That company had implemented new sales and payroll applications while he was there. Guinness wanted to automate the same processes. Trade sales accounting came first, using the 1902 to record transactions. A payroll system followed. Both applications went live in 1967.
Chief programmer Donie O’Brien led the team that wrote the software. Other members of the computer group included assistant manager Joe Kinnane and chief systems analyst Ken Meates. Another analyst, Clive Brownlee, later became executive director of the Guinness organisations in Nigeria and in Dublin. Michael Ward was the head of operations, while Enid Wiltshire was in charge of the data preparation equipment.
The brewery phased out paper tape after it adopted punched cards at the start of the 1970s – a time when it was developing new software to facilitate the introduction of a pound made up of 100 pence instead of shillings and pence. It installed card handling equipment that was manufactured by IBM, but compatible with the 1902.
The paper-driven computing era at St James’s Gate was thus short-lived. Guinness veterans still speak highly, however, of the former colleagues who could edit their programs on paper.
(Guinness computer room photograph, featuring Clive Brownlee and Michael Ward in 1966, courtesy of Diageo)
© Newsmail 2016
December 1960: The first computer installation in Dublin
IBM Ireland commissioned photographer Dick Deegan to record a notable arrival at Dublin Airport in the closing weeks of 1960. His pictures show a set of bulky packages beside the cargo aircraft that had transported them. Each contained part of an IBM 650 computer that was bound for the department of the chief accountant in the Electricity Supply Board (ESB).
The 650 was not only the state-owned utility’s first computer. This was also the first computer that IBM had sold in Ireland. It trailed behind a rival manufacturer, International Computers and Tabulators (ICT), which had shipped similar systems to customers in Thurles and Belfast.

The first IBM computer in Ireland arrives at Dublin Airport. The IBM trio holding the sign are (l-r) Ehud Rubenstein, John Moriarty and George Connolly, who was responsible for orders and shipments.
(Photograph by Dick Deegan courtesy of John Moriarty)
The consignment on its way to the ESB offices in Fitzwilliam Street was therefore the first computer in Dublin.
IBM’s product literature did not refer to the 650 as a computer. It usually called the system an ‘automatic calculator’ or a ‘magnetic drum data processing machine’. By any definition, however, this was a stored program computer rather than a mere calculator. Indeed, the 650 was the first IBM computer that was sold for business use as well as for scientific applications.
It belonged to a generation of electronic devices powered by vacuum tubes – or, to be more accurate, by the electric currents inside those tubes, which were also known as valves. The IBM 650 housed an array of tubes that performed calculations with decimal, not binary, digits. It stored data in memory by recording each digit as a pattern of magnetised spots on a rotating metal cylinder. Data input involved converting the encoded information on 80-column punched cards into electrical impulses.
From the corporate IBM perspective, the 650 was a mature product in 1960. The company had started shipping the system to its customers six years earlier and had already announced a more advanced computer – the 1401 – which featured solid-state electronic components in place of the fragile vacuum tubes.

This illustration from an IBM product guide, dated 1955, shows how the 650 performed arithmetic calculations. The magnetic drum was the key component in this process.
For IBM Ireland, however, the 650 was a new technology. The company had opened its Dublin office in 1956. In the following years it delivered electro-mechanical accounting machines – IBM called them ‘unit record equipment’ – to several local customers. But the ESB installation was its first computing project. It brought in two new staff to support the system. Ehud Rubenstein was an experienced hardware engineer from IBM Israel, while Dubliner Noel Peare had acquired programming skills at IBM United Kingdom.
The January 1961 issue of the ESB company journal reported that assembling and testing of the 650 were underway. IBM had also delivered six new punched card accounting machines to work alongside the computer.
The ESB was well acquainted with older forms of data processing. It had used punched card technology to manage customers’ accounts since 1932. It ran card handling equipment supplied by an earlier incarnation of ICT. The workforce in its accounting machine group knew how to prepare, verify, file and retrieve large batches of punched cards.
There would be little change to these procedures when the computer went into service. The IBM 650 was capable of storing data on magnetic tapes, but the ESB chose to entrust all data input and much of its data output to punched cards.
Programming the IBM 650 involved specifying how and when to transfer processing instructions and transaction data to and from designated locations on the magnetic drum. Operating the system was a matter of feeding in one stack of punched cards that held the program, followed by a second stack of cards that contained transaction data. The vacuum tubes then did their arithmetic and the spinning drum recorded the results. This type of batch processing demanded rigorous accuracy.
The ESB entered its computing era with just two in-house programmers – Eamonn Halpin and Tom Bolton. IBM’s John Moriarty and Noel Peare assisted their initial coding efforts.
Electricity billing was the first application on the IBM 650. The programmers combined details of each account with the latest meter readings and the computer calculated how much the customers should pay. Inventory management for the utility’s central and district stores came next. This involved replacing local records with a centralised control system. All stock values and movements were logged on punched cards at head office.

The ESB’s annual report for 1960-61 contained this photograph of a technician working on the IBM 650. Appropriately for a computer on the premises of an electrical utility, he appears to be testing the system’s power unit.
(Photographer unknown. Photograph courtesy of ESB Archives)
By mid-1963, when the billing and stores applications were both up and running, the accounting machine group had accumulated more than one million cards – all stashed in boxes and stored in a single location. The company made little provision for data recovery, but never had to deal with a major calamity.
The ESB did, however, need more computing power. In 1964 it replaced the 650 with two IBM 1401s.
Dublin had more computer installations now. There were IBM 1620s on the campuses of TCD and UCD. The Revenue Commissioners, Jacobs Biscuits, Esso were using ICT 1301s and An Foras Taluntais / The Agricultural Institute ran an Elliott 803B.
Between 1954 and 1962 IBM delivered 650s to almost 2,000 installations around the world – an unprecedented sales achievement for the young computer industry. The IBM 650 at the ESB was the only one ever seen in Ireland.
John Moriarty
IBM systems engineer
The 650 used a magnetic drum for storage of both program and active data. Program performance could be optimised by calculating the execution time of each instruction and how much the drum would rotate during this period at its constant speed. IBM programmer Noel Peare, who worked on the ESB computer, was thus capable of positioning each instruction on the drum’s surface in a way that minimised the rotational delay.
The ESB continued to store files on punched cards until the reliability of magnetic tape was deemed acceptable!
Read John Moriarty’s testimony
George Morgan
IBM recruit in 1962
I have a particular memory from my first days in IBM. I was brought along to the ESB to see their IBM 650 computer. It was described to me as a cabinet full of little glass bottles with candles inside. It was built mainly on thermionic valves and used a rotating drum storage. Along with a large room full of tabulating machines, it managed the billing for all ESB customers.
The engineer responsible for keeping the candles lighting was Barry O’Connell. Other engineers I remember were Tommy Dwyer, Peter Rice and Harold O’Hare. My memory of those days is that most IBM customers used punched cards and tabulating machines. The only exceptions were the ESB and the Sugar Company who had those early valve-based computers.
Read George Morgan’s testimony
© Newsmail 2020
23 February 1909: A proposed analytical machine
By day he was employed as a clerk in a corn merchant’s office. In the evenings, working alone at his home in Drumcondra for about six years, Percy Ludgate designed ‘machinery capable of performing calculations, however intricate or laborious, without the immediate guidance of the human intellect’. This research was an early attempt to define the elements and functions of a programmable calculator. It is now recognised as a landmark in computer history.
In 1908, at the age of 25, Ludgate submitted a summary of his proposals for an ‘analytical machine’ to the science committee of the Royal Dublin Society. On 23 February 1909 its members heard a formal presentation of the report. The RDS published this in April as part of its scientific proceedings. Ludgate’s paper envisaged a compact desktop device that would use shuttles with protruding metal rods to calculate and store decimal numbers. A sheet or roll of perforated paper would control its operations.
The paper acknowledged that the analytical machine concept bore some resemblance to the ‘analytical engine’ that Charles Babbage had invented about 80 years earlier. Percy Ludgate noted, though, that he had completed the first draft of his machine before he became aware of his predecessor’s work. The fundamental process in the Babbage design was addition. The focus for his machine was its multiplication function. According to the Ludgate paper, moreover, ‘the contrast between the proposed structure of the two projected machines could scarcely be more marked’.
No one ever tried to build the analytical machine and, in the era of mechanical engineering without electronics, its performance would have been rather slow.
In the decade after publishing his proposals Percy Ludgate qualified as an accountant and served on a First World War committee that secured a supply of oats for cavalry horses. He died in 1922, aged just 39, from an infectious disease. His achievements were soon forgotten. Apart from a single diagram that appeared in an engineering magazine, his drawings of the analytical machine were lost.
The rediscovery of Percy Ludgate started in the 1970s. Brian Randell at Newcastle University chanced upon a reference to the RDS paper, then tried to track down people who had known him. More recently Brian Coghlan at Trinity College Dublin has led another round of Ludgate studies, demonstrating the originality of his concepts, exploring how his machine could have been constructed and making contact with members of his extended family.
Today, 100 years after his death, this new plaque on the wall of 30 Dargle Road, Drumcondra commemorates Percy Ludgate as a true computing pioneer.
For more about Percy Ludgate, visit the John Gabriel Byrne Computer Science Collection in TCD
© TechArchives 2023
27 July 1866: ‘It is a great work a glory to our age and nation’
The largest ship in the world, the Great Eastern, steamed out of Bantry Bay on Thursday 12 July 1866. In the early hours of the following morning it reached a buoy that marked the location of a newly laid cable from Valentia Island. This line was hauled onto the ship and spliced with an insulated cable that was long enough to span the Atlantic. The massive vessel then set course for Newfoundland.
Electric telegraphy was the advanced information technology of the mid-19th century. Telegraphers were the coders of their age. Cable designers were leaders in materials science.
Starting in the 1840s, a new communications infrastructure carried text messages in Morse code across the countryside, enabling humans to exchange information at unprecedented speeds. In the 1850s the telegraphy network expanded with underwater links from one piece of land to another. The first seabed communications cable between Ireland and Britain was laid in 1853.
Cable materials, insulation, instrumentation and handling equipment improved steadily. In 1857 a first attempt was made to connect Europe with North America. The eastern end of the line came on shore at Valentia Harbour. This mission ended when the cable snapped at sea. A second expedition in the following year achieved a transatlantic connection, but its success was short-lived. The high voltages that were used on this cable damaged its insulation. Messaging became increasingly difficult and the connection stopped functioning after a few weeks.
There was a seven year gap – not to mention a civil war in the US – before the next effort. By 1865 low-voltage signalling had been refined and new cable designs had proved effective in the Mediterranean and the Red Sea. But a mid-ocean cable snap halted the third attempt to bridge the Atlantic.
The Great Eastern’s journey in July 1866 was the fourth bid to establish a connection from the southwest corner of Ireland to the eastern edge of Canada. While the ship was crossing the ocean, its cable specialists conducted frequent electrical tests and maintained regular communications with Valentia. This link enabled the voyagers to produce a daily shipboard newspaper, passing on the latest news from Europe.
On 27 July the steamship reached its destination in the Newfoundland fishing village of Heart’s Content. The transatlantic wire was joined with a local line and the connection was finally complete. Historians refer to this as the ‘permanent cable’.
While the Great Eastern waited to enter the harbour at Heart’s Content, a Morse code message arrived from the other side of the ocean. It contained a tribute to the cabling mission from that morning’s issue of the London Times, together with a breaking news headline: ‘It is a great work a glory to our age and nation, & the men who have achieved it deserve to be honoured among the benefactors of their race. Treaty of peace signed between Prussia & Austria.’
For a comprehensive guide to the transatlantic cabling projects, including profiles of the investors, scientists and engineers who made them happen, visit http://atlantic-cable.com.
The Cable Station
From October 1868 onwards this building was the primary communications hub for transatlantic telegraphy on this side of the ocean. The restoration plan includes a museum on the ground floor and an innovation centre upstairs. The museum exhibits will include items collected by James Graves, the first superintendent at Valentia, who oversaw the station’s operations for 44 years.
The Telegraph Field
This site was the departure point for the cable laying expeditions in 1865 and 1866. A timber building stood here during those years. The structure in the photograph replaced it, but was used as an operations centre for only one year before the permanent cable station opened.
The Slate Yard
The first-ever transatlantic message was received on 13 August 1858 at a telegraph room inside this now-derelict building. It was transmitted over the short-lived cable that was laid across the ocean floor on the second attempt. The site was later used for cutting slate from a local quarry. The restoration scheme proposes that the structure should be rebuilt to show how it looked on the day when intercontinental telecommunications began.
Valentia Island Today
Valentia became an international communications hub. A new cable station opened on the island in October 1868 and operated until 1966.
The cable laying expeditions were widely publicised while they proceeded and extensive records still exist. The subsequent experiences of the cable station and its telegraphers, who came from a variety of backgrounds and integrated with the local community, are less thoroughly documented. There are plans, however, to establish a museum in the building and to present its technological and social histories together.
This month Valentia Island is commemorating the 150th anniversary of the Great Eastern’s successful voyage with a series of events.
The Valentia Transatlantic Cable Foundation Company, meanwhile, is seeking funds to restore three locations of historic significance. It needs to raise at least €6 million and expects that equal sums will come from government and private sources. The foundation also hopes to secure Unesco recognition for the island as a world heritage site. This process, which involves the submission of academic studies that prove the authenticity and global importance of a site, is now at the penultimate stage.
(Photographs by Michael Lyne, courtesy of Valentia Transatlantic Cable Foundation Company)
© Newsmail 2016