John Gray

I felt no major pull towards any particular career path when I completed my second level education in 1963. After a bit of thought I decided to study engineering at Trinity College, which was where most of my school friends were heading. These included William Burgess, who was MD of IBM Ireland from 1991 to 2002. Faced with having to choose one of the four available engineering options, a quick glance at the newspapers’ jobs pages suggested electronics (and the emerging computer area) paid best, so that decision was easily made! I graduated from Trinity with a good degree, did the legendary ‘milk round’ of interviews, and ended up taking a job with English Electric in a town called Kidsgrove, outside Stoke-on-Trent.

Suffice to say that neither town was in any way like Dublin, or London, a city I was also quite familiar with. The social scene was dramatically different and the job itself wasn’t what I had envisaged – writing small blocks of machine code for an antiquated processor to feed discrete amounts of coal into a steel mill furnace. One evening I returned to my rented accommodation to find a letter from Trinity…

‘What’s this about?’ I wondered as I tore it open. I’d completely forgotten that months earlier I’d applied for the MSc course in computing that John Byrne was developing – and the letter confirmed I’d been awarded a place. I resigned my position the next morning and was soon back in Dublin readying myself for another year of study. By now, thanks to having actually worked, albeit for a short time, in a real-life environment, I also a better idea of what I wanted – and didn’t want – career-wise. Experience is often the best teacher.

The MSc course was interesting and quite broad. Matt Crotty, who went on to found AMS/Insight, was a class colleague. We absorbed knowledge from John Byrne, Neville Harris, and John Moriarty, wrote our programs on an IBM 1130, mainly in Fortran, and marvelled at the wonders of modern computer power.

At the end of the course in the summer of 1968 I took a job with Univac in London. London as a location was ideal and, indeed, the office location was rather interesting in that it was just a two minute walk away from the Old Bailey.

The local pub usually had a mix of Univac employees, barristers in full regalia, and some expensively dressed men of serious demeanour and large build! Shortly after I joined Univac the notorious Kray Brothers went on trial at the Old Bailey, and I often enjoyed a drink in that bar while listening to their supporters, buoyed up by a few pints, loudly offering their views as to the ‘obvious’ innocence of the defendants and the ‘lies’ of the witnesses for the prosecution. One of the most vehement critics of these ‘lies’ was a famous ex-boxer who had started his career at a very light weight but by now was seriously out of condition and approaching heavyweight proportions. I found it rather ironic that the judge later described the boxer’s evidence as being ‘barely credible’.

But of course there was also a job to attend to. I was a bit frustrated that most of the work assigned to me was in Assembler, admittedly a big step up from machine code, but I was inherently a higher level language guy from my time at Trinity. An opportunity arose to join the Univac training group – and as this offered the chance to extend my knowledge of programming languages, and the wider aspects of computing, I was very pleased to secure the move.

One week after attending a course in RPG I was teaching it. And a while later, again just a week or so after doing a COBOL course, I had to teach a class of twenty students how to use it to write programs. ‘Learning on the job’ had a certain meaning in that environment!

After two years in London I was sent a copy of an advertisement from The Irish Times. Honeywell was looking for COBOL programmers in Dublin. I suppose I was a bit homesick, or at least I thought I was. I applied and met Honeywell’s Irish manager Pat Rafter for an interview at Dublin Airport. A month or so later I was in their office in Leeson Street. The company didn’t lack for talent at that time – Denis Hanrahan went on to top positions with Cara, Aer Lingus, and PJ Carroll and later chaired the IDA. Pat O’Sullivan kick-started Digital in Ireland. And Pearse Mee founded Memory Ireland and AMT Sybex. He received a Tech Excellence lifetime achievement award in 2016 for his contribution to the Irish computer industry.

Honeywell had offices in Dublin and Belfast and was a relatively small organisation in 1970 – three people in sales, perhaps ten in software support and, at a guess, twenty between field service and administration. Our focus was purely on the local market – the era of software for export was still some years away – and the name of the game was hardware sales. Application software was seen as something the user could write for himself, pay someone to write, or buy as a package from a third party.

Support was often thrown in for free as a ‘sweetener’ or to overcome the fears of novice users. It generally included writing programs and helping with systems design and development. The single stream machines of those days did just one job at a time, and this usually meant that we had to use the night shift for development work.

It was not uncommon to launch a compilation from a big card deck, head off for a meal or a drink, and return an hour or so later to study the results, only to find that one of the cards had ‘jammed’ and the time had been completely wasted. After two or three such frustrations you learned that watching the cards go through while dining on a sandwich was the better approach!

When I joined Honeywell, the 200 Series of mid-range systems was their main line. While these were perfectly good machines the company clearly needed to modernise its product set to meet the challenges of the systems that Honeywell’s main competitors – IBM and ICL – had brought to market. Honeywell acquired GE’s computer division in 1970 and from that acquisition emerged the Series 60 – a reasonably ‘state of the art’ range that could address small to very large users.

I worked for Honeywell till 1977. I did programming, systems design and development, pre-sales support, project management and delivery, and conversions from IBM systems. The senior software people were encouraged to ‘sell’ to the user – nudge them to upgrade and keep an eye out for opportunities. Through this I began to get a feel for selling, without being exposed to the risks and the strains, of course.

Honeywell’s major customer in Ireland was The Revenue Commissioners, but I did very little work on those large machines. The company enjoyed some good wins – Clover Meats, Irish Ropes, Aer Rianta, Semperit, and Irish Press come to mind. We suffered a bitter loss when Roadstone (Honeywell users) and Cement (IBM) merged to become CRH. Months of late night work and weeks away from home on benchmarks – I spent most of the summer of 1973 in London, working at nights – failed to win us the combined company’s business. On the upside I did get to know Cement’s DP manager Fionan O’Sullivan. When I worked at Digital in later years, I renewed contact with Fionan at CRH, so I guess that summer of ’73 wasn’t entirely wasted!

The toughest loss of all occurred in 1976. What seemed to be a major win at Youghal Carpets was cancelled for technical reasons. Honeywell lost to some small, ‘toy-computer’ outfit called Digital, led by my former colleague Pat O’Sullivan.

Of course Digital was no ‘toy-computer’ outfit, as the market was soon to discover. By 1976 the company was attracting a mounting level of interest in the local market. They’d enjoyed success in education and science, but still weren’t regarded as really serious competitors in commercial systems, where most of the mid-range market action was. The Youghal Carpets win changed that. It became clear that their PDP-11 range had power, versatility, and was loved with near fanaticism by its users. It was also obvious that they had people who knew what they were doing, and who were very skilled technically.

To the outsider, Digital had a strange corporate structure. Everyone had two managers – one (usually in America or Geneva) who was responsible for the overall business of a corporate product line, and another who was responsible for local sales. The Ireland of those days didn’t fit easily and exactly into that model, but it worked, or at least, it was made to work. The company also had several operating systems that appeared to suit specific application areas although they sometimes overlapped – the most popular were RSTE/E for time-sharing, RSX for industrial or process control work, RT for small real-time applications, and TOPS20 for their DECSystem range. And, unlike most other companies, they didn’t pay commission – it was ‘salary only’ for their sales people.

Regardless of these apparent complexities they were clearly doing something right. From what I saw of them from afar they were a ‘fun outfit’, with a strong team approach, and they seemed to have tremendous confidence in themselves. And by pure chance I was about to become part of all that…

I was in London on a course in 1977. It finished ahead of time, and I went to Heathrow and got myself onto an earlier flight. I was reading a newspaper in the lounge when Pat O’Sullivan, who’d missed his flight and was now a fellow passenger, took the seat beside me. A brief chat in the lounge was interrupted by the call to board – only for it to resume when we discovered we were seated together on the airplane. Talk about chance meetings that change your life!

Within a few weeks I’d been interviewed for a job in Digital’s Irish software organisation, told I should really be a salesman, and despite worries (mortgage, young family…what if I fail?) about my ability to succeed in such a role, been persuaded to take the plunge. I joined Digital in Park House in Dublin in August 1977, and in the years that followed it was the greatest company I have ever worked for. Digital gave me confidence, opportunities to advance, and freedom to act on my own initiative without fear of being ‘punished’ if things didn’t quite work out.

At that time the main product offering was the PDP-11, a versatile range of 16-bit minicomputers. We also had the larger 36-bit DECSystem-20 which was popular in the educational and scientific market. There was talk of a new system that was ‘coming soon’ and would be much more powerful – this of course turned out to be the 32-bit VAX family which had such a major impact on both the company and the marketplace. But that talk of new machines didn’t in any way deter people from continuing to buy PDP-11s – indeed the PDP range lived on, mainly in embedded applications, for decades.

Two major differences between Digital and the traditional system vendors were that we had multiple product business streams and multiple channels to market. We sold complete systems at one end of the scale and modules at the other. We sold direct to end-users and indirect via OEMs, VARs and resellers. Sometimes we sold just the base system and the user did his own applications. Other opportunities dictated that we work with a software house with specific industry know-how – companies such as GC McKeown and Mentec. The resellers bought from us in volume, added software that was suitable for smaller users, and resold the total package as a turnkey solution.

ICS, led by Tom Winter, was our largest partner when I joined, operating an online bureau service from Dublin and Belfast and also selling turnkey solutions based on PDP-11s. As the market for computers expanded, other partners came on board – Online Computing, founded by Denis McMahon, ‘Big Frank’ Waters and Ted Murphy, was a significant addition, and I was pleased to be their account manager at the start of the relationship. Matt Crotty’s AMS also worked with us, though their main business was with IBM systems. Gerry McKeown and Paul Murphy ran Datasystems, Richard Stevens (ex ICL) started Timon Computing, and Mike Peirce grew Mentec from a small start-up focused on point solutions for manufacturing into a major international organisation.

I joined Digital as a sales engineer, or as one of my friends unkindly referred to it ‘as a rep’. Pat O’Sullivan was an inspirational man, the classic start-up guy, and I learned a huge amount from him. The company was growing quickly, and it needed a larger and structured management team, so Derek MacHugh took over as country manager and Pat continued to do what he loved best – sell, sell some more, guide, encourage, and sell even more. Derek was a great leader, and a brilliant salesman. He believed in people, and let you get on with the job.

John Gray photographed during his time as Digital sales manager.

Two years into my time with the company I was made commercial sales manager, which involved growing the company’s direct and indirect sales to everyone other than government, scientific, and educational customers. I later became sales manager for the whole operation in the Republic of Ireland and migrated to Digital UK in the mid-1980s.

The company at that stage was a formidable force. Product-wise our VAX systems and clustering technology were in a class of their own, and our people – sales, service, and administration – were dedicated and extremely capable. If we reckoned an opportunity was winnable (once it was carefully qualified), we went after it with determination and we expected to win – not from any sense of arrogance, I should add, but simply because of our belief in our products, the strength in depth in our services teams, and the ability and creativity of our partners when they were involved. We didn’t always succeed, but the win percentage was huge.

We did have setbacks, of course. In some cases we had a shortfall in the power of the system, or didn’t have the right product at that stage. One time, before we had the VAX, a prospect asked us to quote for a PDP-11 and to run a program just to verify how long it would take. He was very happy when it ran in just over thirty minutes, but our expectations of the order were dashed a few days later when he phoned to tell me that ‘hot box’ supplier Perkin-Elmer had run it in under two minutes!

It was also difficult to dislodge an established supplier. If a user had enjoyed working with IBM, ICL or Honeywell, then they tended to stay there, though we often sold minicomputers as additional systems for such accounts – Bank of Ireland and Cahill May Roberts (then part of major IBM user PJ Carroll) come to mind.

Cahill May Roberts was a significant win for me. The company was a pharmaceutical wholesaler, and had a specific requirement involving alphabetic look-up for its products. Our Digital Standard MUMPS operating system allowed for this in quite an optimised way. We located a proven solution from a UK based partner and between us managed to win the business with several distributed PDP-11 systems after a long and hard campaign.

Armed with the ever popular PDP-11, and the enthusiastically received and powerful VAX range, the company went from strength to strength. It was a great feeling to host customer presentations knowing that whatever question was raised the answer would almost invariably be ‘Yes, we can do that!’ What could possibly go wrong? Well, that’s a very different story…

Looking back, and of course with the benefit of hindsight, it’s ironic that the mistakes made by Digital’s competitors in those early days – under-estimating their products and their people – were much the same as those made by Digital some years later, when they failed to grasp and counter the threat from PCs, workstations, Unix and networked systems. But it was great while it lasted!

Last edit: August 2017

© John Gray 2017