Some folks say their lives are affected by the moon. In Barry de Ferranti’s case it’s the sun: every 11 years a rash of sunspots hurls our close cosmos into chaos, and coincidentally, the course of his career.

Over five of these phases, his career has parallelled the course of computing itself, and shows no signs of slowing.

Its cycles run like this: 1948-58, developing hardware and establishing electronic data processing (EDP) in Australia and NZ; 1959-69, international marketing of large-scale systems; 1970-80, independent consulting; 1981-91, education, training and end-user support, 1992-2003, promoting professionalism.

These days, he is deeply involved in ethics and certification issues.

He sees a growing predilection for litigation as a looming threat for ICT professionals and says “it’s a matter of time before there are significant claims against individuals” and that within the ACS there are members significantly affected by the unethical behaviour of others.

“The most likely source of unethical behaviour impacting members is from outside the membership.

Other ICT professionals, practitioners and even students could be at fault, he says. “Beyond ICT, others such as engineers, lawyers, accountants – anyone using computer-related systems – may be guilty of grossly unethical behaviour and members can be drawn into litigation by default.”

So far, breaches of the ACS Code have been largely the subject of anecdotal evidence, and the disciplinary committee (which he chairs) set up within the Society’s Membership Board has yet to take specific action against more than a handful of violations.

But there is a growing tendency for breaches to become public as the “don’t dob” mindset dwindles in the face of corporate financial scandals and academic research methods are shown wanting when scrutinised after complaint.

“We need to be proactive against a rapid deterioration in ethics across the whole professional spectrum because ICT is the common link between the professions, and particularly in responding quickly and effectively to members’ needs.”

As a member of the Council of Professions Australia and with strong affiliations with various local and international bodies, the ACS has moved to widen cooperation with other professional bodies to establish a combined front against failing ethical standards.

de Ferranti cites professional standards legislation adopted in NSW, WA, and listed in other states primarily to cap professional liability, as useful in raising the level of ethical standards through education but stresses that more needs to be done.

(Federal legislation was introduced by Treasury late last year to adopt the standards pursued by states by amending the Corporations, Australian Security and Investment Commission and Trade Practices Acts.)

“The point that Prof Don Gotterbarn makes in his comments on the Code of Ethics for Software Engineering and Professional Practice (see preceding story, “XXXX”) about there being a need for guidance for specialists to avoid some high profile disaster is well made and is the crux of what we are trying to do.

“Professional certification combined with demonstrated compliance with a comprehensive code of ethics can provide effective insurance against false accusation which can lead to expensive legal action however unfounded the claim.”

With other members of the ACS ethics committee (Oliver Burmeister (chair), John Weckert, Mike Bowern, Craig McDonald and Andrew Meyenn) he will continue to review the tenets of the ACS Code of Ethics, particularly in establishing priorities.

“The Software Code is right in giving primacy to the public good with other considerations given secondary priorities. It makes it a more applicable document than our code as it stands at present, and reinforces the need for codes to be constantly reviewed to remain relevant to the environments in which they exist.”

Solar cycles and the gaining of wisdom

de Ferranti’s career spans the gamut of computing in Australia, and an encyclopaedic knowledge not only of its evolving technologies but the pioneers who created and applied it, sets him apart.

In previous “heritage” stories run in IA, much of the anecdotal information about contemporary luminaries like Harry Messel, John Bennett, Peter Jones and Phil Singleton have been drawn from his vast archive of memories and notes.

They studied and worked together, established the foundations of our industry – and have remained a closely knit group continuing to offer energy and dedication to mentoring to those of later generations.

He shares with them all a passion not only for the technology, but the human aspects of their work which persists for half a century or more, in his case particularly, as he wrestles with the social issues which beset ICT today.

At the time of the ACS’ 25th anniversary, he submitted a personal account of his career for inclusion in “Computing in Australia” and edited excerpts from it follow. His narrative ends in 1992 with some prescient forecasts of the state of Australian ICT in the decade since:

University of Sydney

My earliest enthusiasms were for problem solving, trying to fix things and wondering at number patterns. One teacher, Dick Gillings, enthused me with aspects of mathematics, such as magic squares and the algebra of integers. Another advised that engineering offered more career possibilities, so, in 1946 I joined the throng of undergraduates who inundated the Peter Nicol Russell School of Engineering at Sydney University in the post-war rush.

Among the hours of lectures and computation, calculating experiment results manually with the aid of slide rules and graphs, one lecture series, ”numerical methods”, caused many to shake heads, or dismiss it as impossible. So new was it that it did not make much sense, particularly since the lecturer seemed always to be incomprehensible and in a muddle: he was none other than Trevor Pearcey (who was designing and building CSIR Mark I, later to become CSIRAC). Those heady days of Distinction Maths and Physics with classmates from Science and Arts, who included Alistair Mackerras and Margaret Oates, were followed by the pragmatics of Mechanical and Electrical Engineering, to make it all work.

It was my good fortune, as an honours student, to join the team working with Professor David Myers who combined all aspects of mathematics, mechanical and electrical engineering in pioneering work on automatic computation. My research was on '”The Automatic Control of the Synchronisation of a Magnetic Drum Memory with an Electronic Digital Computer”.
Other research was already well advanced on the, non-digital, electromechanical differential analyser. This was most impressive: a room full of distributed computing units, each in themselves mechanical but connected electrically and synchronised by means of stepping motors which drove shafts, a fraction of a revolution at a time, to perform precisely the various integration, arithmetic, input and output functions. There were many dramatic moments in 1950 when the first problems were set up for solution.

One memorable episode stemmed from modelling likely flows of water in Snowy Scheme tunnels. Suddenly, the pen on the graph of Tooma-Tumut went into what looked like wild hunting gyrations. Immediately all suspected an error in the program or set-up, or maybe it was those motors. But it was a real phenomenon: we were watching evidence of a huge case of rampant water-hammer, in very much faster than real time. This led to that great phallic symbol above the Jindabyne reservoir - the surge diverter.

But that was analogue technology. For my research topic I needed digital control. I designed and had made a 15kg brass drum to spin at 3000rpm driven by a velodyne, - controlled, in speed and angular position, by a feedback path of high-speed digital pulse circuitry in sync with a quartz clock-timed computer. Interestingly, the electronics seemed easy; it was the mechanical bug that bit. The bearings were inadequate. Soon they wore and the whole device, most of it rotating at frightening speed, marched alarmingly across the bench while my fellow Honours students demanded to see “drum pulses”'.

Thesis completed, I was stunned to find that in the UK others had reached similar conclusions on the synchronisation of a drum with a computer, particularly the need for special bearings.
I then left for graduate apprenticeship with GEC in England, sponsored by British General Electric, Sydney, and Telephone Electric Industries. It led to work on the first electronic telephone exchange developments. Applying pulse technology, devising delay lines and new switching units, I was granted British Patents for features of my tiny devices, “ferrestors” -- low cost, low consumption magnetic core components, flip-flopped by ferroresonance.

A shift register for exchanges or computers could be made and run for a fraction of the cost of an equivalent valve device. The lab's excitement was short-lived, however, as soon the transistor eclipsed all that. Nevertheless, ferrestors were used at the Signals Research Development Establishment.

Harry Messel and SILLIAC

After touring Britain with a paper that had won an IEE Premium, I attended a course on computing at Imperial College, and visited the BESK computer in Stockholm. I then received confirmation that I was to join Harry Messel's team back in Sydney, building SILLIAC, and observe the USA's ILLIAC en route.

Fabrication of the various chasses (forerunners of the electronic packages of subsequent machines) had been subcontracted to Standard Telephones and Cables.

My role was to test these, assemble them into the mainframe for further testing as a register, or in the case of the 40 electrostatic storage (Williams) tube units as the whole memory. At the same time we designed input/output circuits and commissioned regulated power supplies, and the air cooling and conditioning plant -- important, not just to extract the 40 kilowatts of heat but to provide air of the correct temperature at the inlet, where access was coincidentally exactly the right shape and size for two long-neck bottles of beer. Many nights we worked past midnight and then quenched our thirst, courtesy of SILLIAC.

SILLIAC was ready for live testing in June 1956, despite unexpected gremlins. There was much competition but John Blatt's three-body problem program was the first to run completely

By September we had introduced many to programming. Classes were open, so work for SILLIAC covered a surprising range: Hardy Cross analysis for gas pipe network leaks, unofficial postmasters' payroll, demonstration of how banks' branch managers might use computers for decisions allowing overdrafts, chain-store inventory, and many others.

IBM and America

In December 1956 I joined IBM. There were few there who understood computers, so internal as well as external education was needed. My task was to set up and introduce the company's first business computer, the 650, with a data processing centre to support such customers as AMP, MLC and AGL. This would demonstrate and sell to prospects and help build an Applied Science capability for IBM Australia. Difficult.

Customers could not conceive limitations on their hopes for the 650 (we did not know much more either). For months we laboured with AMP on a system for insurance-agents, for me to test in Endicott (USA). What had been expected to run at near full speed (tens, if not hundreds, of printed output lines per minute) eventually worked at one line every 10 seconds. We had yet to achieve programs' optimised use of limited resources.

Numerically-intensive computing was yet unnamed, but news of technical applications on other machines, like the 704, persuaded 650 users to apply digital methods to subjects which earlier were only explored through analogue models or special-purpose machines.

Great disparity in the scope of user-organisations was apparent. Some were so large that 650s were considered local departmental processors, for few applications, so job scheduling and economic justification were low priorities.

As with SILLIAC, I concluded that we must educate markets and provide expert support to bind customers. That conviction has been the thread running through my whole approach to computing, and has stimulated many projects. It amounts to a missionary zeal, shared by many others, to develop more, brighter users by providing them with better understanding and more informed support as they explore their use of the technology.

Establishing EDP

By 1958 EDP was being seriously applied or planned in Australia, particularly in government departments -- although not all the work was well informed. It was also enthusiastically pursued in New Zealand, which I visited on my return from the States. My experience in Endicott had demonstrated the 650's limitations, so I preached moderation to the New Zealand Treasury who were ready to sign for one in the largest configuration possible. This did not endear me to sales management but it probably avoided embarrassment and encouraged subsequent progress.

It was important to ensure effective support for users. This was achieved by hiring good programmers like Margaret Oates, whose mathematical and statistical background proved valuable. Her patience, attention to detail and ability to explain were also vitally important. To meet our aim of transferring technology to our users we taught many to plan and write their own programs using the 650 “internal translator” to convert programs written in FORTRAN, a language successful on the large machines overseas. It was a struggle, but well worthwhile.
Long hours on the machine were a permanent feature of those early days: On his return from the United States, Hugh Philp, then Reader in Education at the University of Sydney, requested processing of examiners' marks for the Leaving Certificate. We programmed to run the 12,000 students' English results through a battery of processes for validity, arithmetic and other aspects to create a histogram by which the breakpoints could be determined, over one weekend. I stood at the machine all Friday night and most of Saturday and Sunday, to ensure continuity. There was one terrible moment when the system indicated one excessive mark, but emergency calls disclosed that bonuses were occasionally awarded.

Then, as ever, systems analysis had failed to uncover all possibilities because none of the computer people knew enough about the subject area.

To cope with the healthy interest in the application of available software we recruited well-qualified staff, including some who later achieved greatness in other fields: Keith Burrows, Bill Skyvington and Michael Arbib.

Competing for big plums

In 1959 the first really large Australian government tendering opportunity, for Defence, attracted a number of new competitors. One of these was Ferranti. I left IBM to join that company, then a leading defence and electrical engineering firm in Britain. The team I assembled, to establish the Melbourne Computer Centre in Stanhill, Queens Road, in 1960, helped promote technical computing in Victoria. Before long Melbourne-based government departments, manufacturers, universities and utilities were loading our little Sirius beyond expectation; even a motor journal ran a survey, with extraordinary response, causing us to work shifts to cope.

In what was probably a first, we processed linear programs for steel-making at BHP Newcastle, using telex as a remote terminal. This was followed by an even longer link, set up for the conference and exhibition at Melbourne University, which connected a telex there to the supercomputer, Atlas, in Manchester, England. With only a small stand, but a huge sign inviting attendees to “Get Your Atlas Computing Here”, we confronted the larger glossy stands of the competitors.

The most important plum we sought with this activity was the sale to CSIRO of Atlas and the Ferranti-Packard FP6000 from Canada. Soon that became an even larger prospect with Treasury. My preferred strategy, knowing we were opposed by the formidable Control Data and others, was to offer the systems at an attractive price, knowing that the flow-on business could be very lucrative. It was a bad time for someone in the antipodes to be asking the parent in England to sell, as they saw it, unprofitably. My logic was rejected and our chance was lost. It could have been a turning point for the industry: Atlas and FP6000 were superb designs, well ahead of their time and well suited to this country's needs.

ACS: the beginnings of a professional society

In the early 60s new organisations such as the Victorian Computer Society brought computing professionals together. I am glad to have been involved so early and to have contributed, in their 1963 Thesaurus. Soon after, Ferranti Computers were absorbed into ICT and I returned to Sydney. There the NSW Computer Society was undergoing change -- as in other states, conflicts between computer-science and commercial emphases, professionals and practitioners, were evident and there seemed to be growing tension.

Following amalgamation into the ACS in 1966, harmony should have ensued. Although I was only tangentially involved, I sensed unfortunate difficulties flowing from the strongly differing views on the proper roles of the Society. Ideally, ACS could and should be both a forum and a force for the whole community of computer-users, to develop their skills and voices in a world increasingly dependent on effective use of information technology.

ICT merger and people The conjunction of different cultures when companies merge is traumatic for all involved. ICT had already met this in the differing philosophies of its two main precursors, Powers and Hollerith. The difference was not only in the physical punched-card design context (staff were categorised as round-hole or square-hole) but more particularly in the approach to solving customers' system requirements: one had a turn-key approach, the other tended to teach the user and let him do it. Added to this were the more technical approaches of Elliott and GEC (whose computer divisions had been drawn into ICT) and now the even more advanced group from Ferranti with their time-sharing technology.

In Australia I was leading a small but very capable team whose members felt they were about to be swallowed by the ponderous Powers/Hollerith bureaucracy.

Some fled, sooner or later, mainly to former rival Control Data or university, but others stayed and proved their worth. For me personally the challenge was to take on ICT NSW management and move it into the new computer-market. Soon we established a separate branch office and identity in William Street.

Again education and training became a major focus. Internally the processes of organisation development and culture change demanded that we develop the staff personally as well as technically, and also in marketing and sales techniques. The investment in people was significant. Externally, customer and prospect training was no less important for realising the potential of the new 1900 range of computers. It proved a long uphill battle against inertia.

At the same time there was growing interest in the potential careers in EDP, which might attract university graduates, and the University of Sydney Appointments Board established an industry-wide body which I chaired.

Another battle was fought on an international front. In the effort to sell the new technology to the private sector, decisions previously taken locally were now referred to head offices overseas, and since our company was unknown there, we were losing sales to American competitors. In 1967 I established marketing/technical liaison in New York, not just for ICT
Australia but for the new global company, ICL.

Founding that first ICL office was challenging. High above Madison Avenue, in fact right above IBM's uptown New York branch, a brave band of marketers and specialists from UK, USA, NZ and Australia, we set our sights on influencing sales to ICL prospects worldwide, and were successful in tens of millions of dollars' worth of contracts.

Part of the technical liaison involved making contact with the leading industry consultant and information source of the day, Isaac Auerbach. The doyen of the International Federation for Information Processing, Isaac became a good friend and later business associate of benefit back here in Australia.

Consulting

In 1970 for family reasons I left ICL and returned to Australia, establishing my own consulting practice. Dramatic changes were about to affect all industry: microcomputers shifted the balance of power from the central high priesthood of mainframe computing to desktop users, who then needed good advice and independent information.

So began the second main, service phase of my career in which the objective of adding value to the technology, gradually predominated. The sagas of the many projects that were undertaken, the continuous process of self-development to keep up with the escalating complexity of the technology, and the difficulty of moving from corporate life to effectively a one-person business, is enough for several vignettes. Instead, here is a view of present and future challenges.

The challenge of self-help IT

The computer industry is now, in 1991/92, suffering its first major downturn. What does that mean? What should anyone do to avert the potential dangers?

There are several warnings being sounded in this time of change, which those dependent on computers for their livelihood must heed. No longer is the computing professional the only one qualified to control the processing of data in the workplace. There are now more practitioners, all highly skilled in their own vocational specialties capable of using the powerful desktop tools loaded with remarkable software tuned to their needs, perhaps in the ratio of ten to one, than professionals with computer science degrees.

The downturn is among providers of the technology rather than its users, who are daily becoming more capable and self-sufficient. As the interconnectivity provided by networks enables sharing of computing and communication resources, emphases shift from buying traditional vendors' complete solutions to users building capability themselves in a self-help IT world. Such computer users are less dependent on vendors but more likely to seek specialised support from others who may be familiar with their applications. These advanced users and those they turn to for support need access to sound advice, plus adaptable and relevant education and training.