Peter Morrogh

Beirut in September 1971 was the Paris of the Middle East with its smart shops, cafes, corniches and casinos. It was said that you could ski in the morning in the Lebanese mountains and then swim in the Mediterranean in the afternoon. Such pleasures, however, were not options for me. I was down to my last £50 and urgently needed to find a job. The local IBM office put me in touch with EA Juffali & Bros – one of four IBM installations in Saudi Arabia. I took a plane to Riyadh on the next day and joined Juffali as head of systems and programming.

I had graduated from Trinity College Dublin four years earlier and gone to work at a subsidiary of Joseph Lucas – a major UK manufacturer of automotive spare parts. By 1971 I had received training in programming, systems analysis and organisation and methods. That gave me foundation-level systems analysis experience and skills with which to seek employment in the Middle East.

The Juffali computer room in 1975, as seen in a company brochure.
Photographer unknown. Photograph courtesy of Peter Morrogh.

Juffali’s EDP department had 40 employees – all male and mostly from the Middle East and Pakistan. Operations ran efficiently but they were labour-intensive and time-consuming. Around 30 staff operated the computer or verified the data on 80 column punched cards prepared by user departments. At monthly cut-offs everybody was under time pressure and hours were spent finalising the input data.

We ran stock control and management accounting systems for a wide range of trading activities. The company was an agent for European and American partners such as York Air-conditioning, IBM, Heidelberger and Siemens, but Daimler Benz, whose trucks dominated the Saudi roads, was the most important by far.

Our hardware was an IBM System/360 model 20 with magnetic tapes (no disks) and a line printer with a maximum of 132 characters across a page. Card readers were the only method for inputting data. A main storage capacity of 16kb limited the functionality of what could be executed in any program. For example, dates were always formatted as YYMMDD (the source of the Y2K problem which had to be addressed in many organisations when the year 1999 switched over to 2000). All data had to be processed sequentially for updating the magnetic tape files. Designing reports frequently required much ingenuity so that no data exceeded the allocated field size on a 132-character-per line page.

RPG was the main programming language, with memory-intensive programming in Assembler. A single system, such as the one that maintained Juffali’s spare parts records, could consist of more than 50 programs comprising input programs, data verification programs, sorts, updates (with weekly, monthly and yearly summaries being stored in cumulative files) and report production. Hierarchical coding structures enabled products and customers to be categorised and grouped for reporting purposes. As requirements changed, or new product lines were added, coding structures had to be altered, involving modifications to file layouts (typically, by adding new fields to encapsulate the additional categorisations), programs and reports. A “simple” modification could result in extensive recompilation of programs and modifications to live and backup files. Programmers carried out system testing by means of desk reviews and a dry run. Inevitably, bugs got through this rudimentary testing process and programs occasionally crashed midway through execution.

Job sheets provided clear instructions on the sequence of program execution; control totals needed to be checked at several points in updating. As each job was completed, the job sheet was ticked. Any error on a job sheet could have serious consequences such as updating a cumulative file twice. When issues such as this occurred, backup copies of files had to be restored and reruns completed.

There were some things over which the EDP department had no control. These included the computer room environment. Despite having a standby generator and (theoretically) an uninterrupted power supply, a failure in air conditioning could necessitate the shutting down of computer operations. The IBM 360/20 went “thermal” at around 30 to 35 degrees and stopped working, mid-program. There were also periodic hardware failures that could result in days of waiting for IBM to fly in the necessary spare parts or make the necessary repairs. It was said that IBM actually stood for “Inshallah Bukra Munkim” – “God willing, tomorrow, perhaps”.

Other computer installations had similar problems. This led me to make a proposal at an annual computer conference to set up an IBM computer users association (CUA). I was taken aside afterwards by one of the attendees and advised, in a friendly fashion, that the Saudi authorities preferred to be consulted on such matters. Nothing came of my suggestion.

Away from the daily job, life in Riyadh offered many compensations and much generous hospitality. Other expats – mainly English – organised desert expeditions. There were weekend camping trips in wintertime and weekends spent beside a swimming pool in summer. I was lucky to visit many of the major sights in Saudi Arabia, such as the Empty Quarter, an ancient city called Maida’in Saleh built by the Nabateans almost 2000 years ago, the natural lakes outside Leyla, and many small mud-built villages.

‘Life in Riyadh offered many compensations’: Peter in 1974 in Saudi Arabia. Photograph courtesy of Peter Morrogh

After five years in the Middle East I travelled back to Dublin in a Mercedes 200 company car that I bought from Juffali. This journey included detours through Iran, Afghanistan, Pakistan and India.

Back in Ireland in 1977 I found a computer trade that was more diverse than in Saudi Arabia. All the mainframe computer suppliers were represented. There was vigorous competition to gain customers and CUAs were active. A decision to go with any manufacturer typically locked a customer for many years into a relationship with that supplier. Minicomputers were becoming well-established. Digital Equipment Corporation was the most prominent minicomputer supplier in Ireland. Data General, Nixdorf, Singer and others had also arrived. IBM competed against these with its System/3, which ran an operating system that was incompatible with its mainframe software. Consultancy companies such as System Dynamics and Cara worked in parallel with the surge of new suppliers. Some offered packaged software, which meant that user organisations did not always need in-house analysts and programmers, although they still required an operations function.

I received several job opportunities in Dublin. Fortunately, my lack of experience with systems that included disks was not a disadvantage. Photocopier maker Rank Xerox employed me as a project manager. The company ran a sales and distribution facility in Glasnevin and Colm Maum was its IT manager. Rank Xerox billed its customers according to the number of photocopies that they produced and applied volume-based discount rates. It wanted to implement software that Rank Xerox used in other locations on Honeywell systems, but we had to convert it to run on an IBM computer. The transfer sounded straightforward. We would prepare specifications for the enhancements that were needed, amend and recompile the programs, change file layouts and test the software in parallel runs. But the team hit problems that could not have been predicted, such as intricate differences in character representation at the byte level. Converted files still carried Honeywell representation when running on IBM hardware and caused the software to crash. Despite these difficulties we completed the project on target.

My next move was into consultancy with software company System Dynamics, which was led by Tom McGovern and Fred Kennedy. Tom concentrated on IT solutions for Irish businesses whilst Fred focused on consultancy services for the EEC. Tom had established an excellent working relationship with the Irish Management Institute where System Dynamics staff were expected to give talks on technical subjects. The company had previously attracted people like Mike Purser and Dennis Jennings who would subsequently distinguish themselves in their IT careers.

My work was largely consultancy with a good dollop of systems analysis training. I assisted Tom with troubleshooting on customer projects and Fred with his work for the European Commission, including a study of future “blue sky” applications of technology. System Dynamics also enabled me to build up good IT contacts in Ireland. I met people like Dudley Dolan, Owen Dalton and John Blunden, who were all active members of the Irish Computer Society (ICS). I joined and offered my services to produce a periodic newsletter – a low-cost means for ICS members to keep in touch.

System Dynamics had a relationship with Beecom for the delivery of a structured systems analysis and design course. Paul McWilliams and Peter Smith had set up this company in Belfast in the mid-1970s. In 1980, they decided to open an office in Dublin under Dudley Dolan and John Blunden. Later that year I joined Beecom as a senior consultant and worked on several multi-year contracts over the next nine years.

User organisations were now adding more functionality to systems designed and built in the 1970s, seeking increased flexibility and cost savings without wanting to throw away their core applications on mid-range or mainframe computers. Most struggled to find the necessary skills to maintain legacy programs, operating systems, database technologies and hardware. The 1980s brought an exotic variety of programming languages, database technologies, hardware suppliers, software suppliers and software packages. When microcomputers achieved respectability in business environments, they replaced VDUs and linked core applications with satellite systems that supported application-specific software or office tools such as word processing. The benefits included greater reporting flexibility, the transfer of data entry responsibilities to end users and remote data collection from the point of sale and other locations.

Consultants like me tried to see a safe way through this jungle.

One of my Beecom projects was the selection of a corporate banking system for ACC. I prepared the request for proposals, assessed how ACC’s requirements might or might not be met by each proposal and recommended a software package called Bankmaster. It came from a little-known company called MA Systems. Tony Kilduff was its managing director and Cian Kinsella was technical director. They had previously supplied Bankmaster to Ansbacher Bank in Dublin and its functionality was largely Ansbacher-specific. MA Systems undertook extensive redesign and redevelopment to meet ACC’s needs. This work turned the application into a genuine package with international sales potential. MA Systems, later known as Kindle, became one of the most successful Irish software companies.

Following three years in ACC, I developed PC-based software to manage loan maturities and roll-overs in Allied Irish Finance. Gerry O’Hara was the IT director there.

Peter Morrogh in 1989
Photographer unknown. Source: Irish Computer.

My last major Beecom assignment was with Guinness Group Sales Ireland (GGSI). Most of my work there was consultancy, advising how technology within the IT framework could either be harnessed or modified to meet a specific business need. For example, data feeds from mainframe systems could be used to develop bespoke solutions.

I was based in Guinness for around three years. My manager there was David Judge, who was not only recognised as a fair and incisive IT manager, but was also a very distinguished hockey player, with his own entry in the Guinness Book of Records. One of my projects was a system that tracked the presence of non-Guinness taps (such as those supplied by Heineken) in pubs. This was key information for Guinness to measure market penetration and changes in its market share. Another project concerned “waste beer” – mapping how beer left the manufacturing process without ending up in a barrel, bottle or can. Beer could go to waste in many ways: every brew had to tasted and sampled and some was used for research purposes.

The ICS, meanwhile, enabled me to find out what was going on in the world of computing beyond the specifics of the particular project I was working on. There was now an important trend towards standardisation in hardware, database technologies, programming languages and operating systems. IBM no longer had a monopoly on its sites. For example, Bank of Ireland famously replaced IBM disks with more powerful Amdahl disks at no risk to data processing.

I attended most of the society’s evening lectures and seminars where people discussed their particular hobby horses. Tom Gilb, for example, gave seminars on software quality improvement and ICL held a lecture on its content addressable file storage. The ICS ran its activities through the efforts of one part-time paid secretary, some dedicated volunteers and the goodwill of many people and organisations. Income from membership fees was essential for its survival, but professional membership rarely went above 1,000 people. The ICS council reckoned that this number should be double at least.

In 1988 I attended the inaugural meeting of CEPIS as the ICS representative. I chaired the ICS itself from 1989 to 1997 and became honorary secretary and president of CEPIS as well.

My work with Beecom was varied and interesting, but it was not long-term. In 1989 I received an offer to join Lifetime Assurance, the life assurance arm of Bank of Ireland, and took it up. I managed Lifetime’s client-server computers, the point-of-sale systems used by its advisers, the software that supported its quotations process and many reporting systems.

Lifetime merged with New Ireland Assurance in 1997. The combination gave me a portfolio of applications to develop, enhance and maintain until the Bank of Ireland Group promoted me to audit partner in 2004.

Last edit: May 2020

© Peter Morrogh 2020