In the wake of the Post Office affair it is hardly surprising that Professor Thimbleby (Letters, January 22) and others call for regulation of the IT sector — or perhaps we should call it the digital sector. Such calls are hardly new. But consider the consequences. Regulation would impose barriers to innovation and competition. When the Horizon project started it was normal for technology projects to be “big upfront”. However, such detailed requirements and designs, despite best efforts, did not foresee the future, as Mike Bracken, the UK government’s former chief digital officer, notes (“Big IT is a failed 90s model that has ruined too many lives”, Opinion, January 20). Indeed, this approach created more problems than it solved. (I know. I worked on Railtrack systems at about the same time!)
Today a “little and often”, or an agile approach is the norm: understand a little, deliver a little, understand a little more and repeat. Yet 20 years ago this was considered radical and unsound by many. If regulation had existed when Horizon started, it would have prevented this new model emerging. Regulation would have entrenched the position of large suppliers and prevented new competitors using new approaches. Indeed, what the bespoke digital sector needs is more competition — compare this to retail digital where Google, Microsoft and Apple drive each other on.
Buyers need genuine competition between suppliers offering different approaches and different technology. More importantly, suppliers need to be judged on past performance — past projects and the last month. That Fujitsu, and other large technology suppliers, are not judged on their past performance is a monumental folly. In software engineering, unlike in financial markets past performance is a pretty good indicator of future performance.
Allan Kelly
London W3, UK
