Get under the bonnet of your PC

Ever since Dava Sobel's Longitude became an international bestseller in 1995, there has been a vogue for small-format histories of science and technology aimed at general readers. Neil Barrett's book is the latest of several such histories of the computer in recent years. His is different from most in that it is intended for readers who use computers but have little idea what goes on "under the bonnet". It is not quite computer history for geeks, but readers with a taste for the technical will find it more useful than people who want to know about the cultural, social or commercial impacts of computers.

There are three main themes: computer hardware, software and the internet. The book opens with a fairly traditional history of mathematical instruments and mathematical logic. Most of the usual suspects get a mention - Blaise Pascal, Charles Babbage, Ada Lovelace, Alan Turing, John von Neumann and so on - but there is little recognition of recent scholarship. The subsequent development of electronic computers is told with reference to familiar pioneer machines of the 1950s, such as the Edvac, Edsac and Leo, before rushing on to minicomputers and the personal computer. This style - a rather ponderous account of the 1950s followed by a rapid trot through the next 30 years before landing on contemporary technology - characterises many histories of the computer.

Barrett is clearly more comfortable when he moves on to computer software. Although historically based, this is largely a practitioner's account in which the development of programming technology is interlaced with some historical episodes. There is interesting coverage of key programming languages, object-oriented programming and operating systems - primarily Unix and Windows. The rather technical discussions would be fine for a science sixth-former, a first-year undergraduate or a computer hobbyist, but they are not particularly accessible for general readers.

In its final section, the book describes the internet and its consequences.

There is the familiar story of the evolution of the internet from its military origins in the 1960s until its global explosion in the 1990s. This is followed by the best chapter, "Problems with computers", which describes security issues including computer hackers, viruses, worms and spam e-mail.

Barrett's expertise in forensic computing shines through here.

The book closes with a short chapter on the future of computing. Barrett mainly discusses the prospects for artificial intelligence, which is nicely done with illuminating examples such as the Turing test. No doubt wisely, he has not been drawn into speculations on topics such as pervasive computing, the virtualisation of commerce or the impact of the computer on cultural traditions. He writes well about technology, but like many authors who would emulate Sobel's success, he is short on historical sensitivity and scholarship. There are no citations, and the bibliography lists just two semi-popular books on computer history. For his other sources, he writes: "For many of the topics covered in this book, the reference material was the excellent website www.wikipedia.com, in which there are several essays covering the history and development of different aspects of computer technology." If you believe everything in Wikipedia, I suppose you will believe everything in this book.

Martin Campbell-Kelly is professor of computer science, Warwick University.