The one that got away

Dylan Evans on A Computer Called LEO, a surprising combination of teacakes and computers by Georgina Ferry

Review of A Computer Called LEO by Georgina Ferry

The Guardian , Saturday May 24, 2003

With Microsoft, IBM and other US companies dominating the computer industry today, it may be hard to believe there was a time when British computer manufacturers competed on an equal footing with their American counterparts. Yet four decades ago, in the late 1950s and early 60s, the British firm Leo Computers was making and selling office computers that were cheaper and faster than those produced by IBM. In her latest book, the science writer Georgina Ferry tells the engaging story of the company that embodied the rise and fall of British computing.

Leo Computers was set up in 1954 as a subsidiary of Lyons, a company whose fame rested principally on its chain of teashops. As Ferry wryly remarks: "A background in catering is not normally seen as an obvious qualification for hi-tech startup companies." Indeed, the association of Leo Computers with cakes was eventually exploited by rival computer manufacturers to portray Leo as light and fluffy, unlike their own "hard science". Yet, as Ferry points out, in the context of 50s Britain, in which Lyons was a prestigious company, "the juxtaposition of a the hi-tech electronic machine with the homely comfort of a cup of tea and a penny bun" was more intriguing than off-putting.

The first seeds of Leo Computers were sown in the 20s, when a young manager at Lyons called John Simmons dreamt of automating the mind-numbing task of calculating and checking sales figures. In those days, Lyons employed hundreds of clerks, each tapping away on clumsy adding machines that worked entirely by means of cogs and wheels. Two decades later, Simmons had risen through the ranks at Lyons, so when a couple of younger managers told him about the "electronic brain" that had recently been unveiled at the Moore School of Engineering in Pennsylvania, he was able to send them to the US on a fact-finding mission. Eniac, as this early computer was officially known, had been built to calculate shell trajectories for the US army's Ballistics Research Laboratory, but the men at Lyons quickly spotted its commercial potential. Simmons set up a team of talented and farsighted mathematicians, engineers and managers at Lyons to develop a machine that would turn his prophetic dream into reality.

The first computer built at Lyons was dubbed the Lyons Electronic Office, or Leo for short, and entered service at Lyons in 1951, calculating bakery valuations. But Simmons was not content to let things stop there. By then, his team had acquired so much expertise in manufacturing and programming computers that he was able to persuade the board to set up a subsidiary to make and sell computers to other firms. Leo Computers was sold by Lyons in 1964, but went on making computers for another few years before eventually discontinuing the Leo line in 1969.

What went wrong? Why did the British computing industry lose its early lead? Ferry invokes a number of factors, from the small size of British computer companies relative to their American counterparts, to the inability of British companies to combine effective salesmanship with innovation. But she puts most emphasis on the role of government investment. In the 50s alone, IBM won contracts worth almost $400m from the US government, while the British government spent only a tiny fraction of that on business computing. What little money the British government did spend on computing all went on military projects. This is all the more paradoxical in light of the fact that it was a British company, Lyons, that first saw the commercial potential of computers, as opposed to their scientific and military applications.

Another difference between British and American computer manufacturers in the 60s lay in their different approach to their clients. While Leo Computers offered specialised computing solutions, tailor-made to fit the needs of each client, IBM moved towards standardised, off-the-shelf packages that could supposedly be made to do whatever the client wanted. It is ironic that the pendulum has now begun to swing back again, with many companies increasingly seeking specialised software solutions, as they realise that the standard packages are not always as "user-friendly" as they appear.

Ferry tells the tale of Leo with her usual crisp style. Refreshingly, she does not attempt to bestow a spurious importance on her subject by making exaggerated claims for its role in the history of computing. Rather, she allows the facts largely to speak for themselves, only now and again adding unobtrusive commentary and never bullying the reader.

The book weaves a very human tale with strands of scientific and social history. And it is, above all, an endearingly British story - a tale of tea, cake, and a bold vision ruined by government short-sightedness.