Maverick star of stage and screen

Earlier this year an email was distributed around the UK computing community. Richard Stallman was due to be visiting London and had agreed to give a talk at Queen Mary, University of London, on "Copyright vs community in the age of computer networks". To anyone with knowledge of Stallman, the opportunity was too good to pass up. The venue was moved from a lecture room to "The People's Palace", previously a theatre and now part of QMUL. The location was apt since Stallman has a theatrical streak despite his computing and mathematical background. The venue was packed and there was a religious fervour about the event, even though Stallman is of distinctly atheist leanings.

Stallman was an undergraduate at Harvard and demonstrated high mathematical aptitude from an early age. However, his fear of failure prevented him from attempting the Putnam exam, a prestigious mathematics prize in North America. In any case, Harvard was probably too stifling for such a renegade and maverick. Fortunately, the Massachusetts Institute of Technology was close by and Stallman naturally gravitated towards the more relaxed "hacker" culture available there. This was in the days when the term hacker still meant someone writing software altruistically for fun and interest rather than the more sinister overtones of security violation that the term invokes today.

At MIT, Stallman's experiences convinced him that software should be "free", in the sense of "freedom" rather than in the financial sense - hence the title of the book. This confusion with "free software" as a term has led to the alternative "open source", although Stallman is less than keen on this term. Because of the proprietary nature of parts of the Unix operating system, Stallman decided to rewrite it from scratch, starting with utilities such as the Emacs editor (originally written by Stallman for another operating system), a C programming language compiler called "gcc" and so on. He dubbed the system GNU (GNU's Not Unix), a typically jocular self-referential and recursive acronym. Stallman's programming productivity and expertise meant that he was largely successful in his attempt. He founded the Free Software Foundation as an umbrella organisation for his activities.

Unfortunately for Stallman, the last major and ultimately unsuccessful piece of the GNU puzzle was the underlying kernel, Hurd. Because of delays in its production, Linus Torvalds from Finland was able to produce a much lighter-weight kernel, Linux, for use on PCs. With the help of other altruistic programmers, and the convenient communication medium afforded by the internet, he was able to muster and manage a large-enough body of expertise at the right time to produce what has become the kernel for the rest of GNU. As a result, many have heard of Linux, but a much smaller number have heard of GNU. Understandably, Stallman prefers the term GNU/Linux, but this is little used by anyone else.

The book approaches Stallman from a journalistic rather than a technical point of view. A number of interviews were arranged and these are quoted in detail, providing a more personal if somewhat snapshot view of his personality. Stallman is the first to agree that he has a somewhat extreme character. He has described himself as almost autistic, but he has enough charisma and flair to draw a huge following, as demonstrated when he spoke in London. He is also a difficult enough person to have generated enemies, not the least of them being Microsoft.

As well as the interviews, the book does present Stallman's life story and technical contributions in enough detail to make it reasonably well rounded. It is not the last word on Stallman, but it is a useful contribution to the debate he has generated.

An interesting dilemma in the production of the book was the availability of an electronic online version. The original publisher wished to produce an "e-book", a technology designed to limit the usage by the purchaser even more than printed books. This would have been totally anti the spirit of the subject matter of the book, and Stallman was opposed to it. Fortunately, the final result is apt. O'Reilly is a leading publisher of Unix and other software technology books, many covering open source (or "free") software, so it understands the issues well.

As a result, the book is also available under the GNU Free Documentation Licence (GFDL) online in HTML format (see www.faifzilla.org ). This means that many of the errors in the printed version have already been corrected in the online version. It also means that anyone is able to copy and augment and improve the book, provided it is made available to others under the same terms. GFDL has been adapted from Stallman's General Public Licence agreement that provides a similar function for software. It ensures that once a piece of software or a document is "free" in Stallman's sense, it and any publicly available derivatives must remain free under the same agreement. This unique style of copyright agreement seems to be working remarkably well and may prove to be Stallman's most lasting legacy.

This book will be of interest to anyone who already knows Stallman's work. Whether it will produce further converts, I am not so sure; but if the idea of free software is intriguing to readers in any sense, whether technical, political, philosophical or ethical, this book could be of interest. Certainly it is very approachable for non-technical readers. Stallman's ideas could be applied in a much wider arena than just software, but Stallman himself has always been reluctant to do this. Perhaps this book will inspire others in unrelated areas to do so.

Jonathan Bowen is professor of computing, South Bank University.