Leo: The First Business Computer
By Peter J Bird
Hasler Publishing; 272pp;
Review by Jane Bird
A teashop is hardly where you would expect to find the latest in computer technology. Nevertheless, in the 1950s J Lyons pioneered the world's first commercial mainframe. (Lyons, since subsumed into Allied-Lyons, was in those pre-merger days famed for its Corner Houses, where the weary shopper might be refreshed by a genteel cup and entertained by a real violin.) Peter J Bird - no relation of this reviewer - is well placed to tell the story of 'Lyons Electronic Office' (better known by its acronym Leo) from its inception, through its growth into a separate business, and on to its eventual sell-off. As a Lyons employee from 1964 to 1991, he was an eyewitness to many of the events he writes about.
Bird is in addition a diligent researcher. He has unearthed a great deal about what happened on Lyons' computer side before his own day. For computer professionals his chronicle is full of revelations: about the obstacles and the countless technical breakthroughs that were made, about the introduction of test programs and flow-charts, about the development of programming and the origins of the world's first business application - payroll.
For the general reader this could be somewhat heavy going, although the author makes a valiant attempt to explain the intricacies of early computer technology in layman's English. Even the totally computer illiterate, however, will appreciate the immensity of the project - in every sense. The original Leo had 6,000 thermionic valves and occupied 5,000 square feet of office space. Its processing power was of course rather less than that of a modern desktop computer.
Budgets and costs also show how times have changed. When Leo Computers began hiring out its resources and know-how, it charged £50 an hour for machine usage and £2 an hour for programmers. Today those figures would be transposed.
Leo: The First Business Computer will be an invaluable reference for students of the history of computing, meticulously recording events, dates, precisely who was present on that occasion. Appendices, biographies and source notes occupy one-third of the volume. But it is in the anecdotes that the book comes to life. It appears that error-checking was introduced after a teashop manager accidentally ordered 10,000 rolls instead of 100. Various ingenious ploys were used to ensure that visitors never saw the machine while it was suffering one of its many breakdowns. Banquets were laid on to boost the morale of the overworked research team.
They can scarcely have been necessary. The reader can sense the obsessive determination of those involved to see the project through, often at the expense of friends and family. As one participant said later: 'It is easy to remember much of what we did, but somehow easier still to remember the mutual understanding, the comradeship, and the high good humour with which we all tackled it.'
Bird was present himself for some of the book's best moments. The factory in which the second Leo was built was, he reveals, more than a little primitive. One day in 1966 when he was visiting the plant for the last day of trials, the heavens opened and rain poured through the roof of the elderly building. Most of the production area was flooded, including machines under test. Happily, no serious damage resulted. The water was driven off by a selection of industrial heaters and hair driers kept ready for just such a crisis. Bird recalls that 'It was always difficult to justify to visiting customers why they needed to invest in special, clean, air-conditioned premises when the production facilities were less than satisfactory.'
Leo did not have the UK market for commercial computers to itself for long. Like many companies in those days, it entirely failed to predict the ferocity of the transatlantic onslaught of IBM, Univac and Honeywell. In any case, Leo Computers was hardly the most businesslike of businesses. During the first six years of its life it did not even have an accountant it could call its own. Through lack of financial acumen it effectively provided customers with free consultancy, by not charging enough for time spent on developing their applications.
Lyons' managers knew that the answer lay in combining their Leo subsidiary with another computer manufacturer, but they were slow to move. At last, in 1963, the company was merged with English Electric in the first of a series of amalgamations which eventually led to the creation of ICL. Seven years later, Lyons threw out its Leo computer in favour of a machine by the unassailable IBM. How times continue to change.
Jane Bird is an IT journalist.