When tim Berners-Lee invented the world wide web in 1999, his proposal – presented the year before – was that “a global hypertext space be created in which any network-accessible information could be referred to by a single ‘universal Document Identifier’”.
The idea was beautifully simple. On the server side, there were webpages written in a hypertext markup language (HTML) that followed simple conventions and rules. On the client side, there was a browser that was able to translate the HTML code into a readable format. The web of browsable pages was knitted together by hypertext links, which became known as URLs.
Berners-Lee had an unfashionable vision of “the Web’s potential to foster a global village, not its potential to earn him a villa and a fleet of cars”, but he was not the first to have that vision. He credits his inspiration for the Web to Professor Ted Nelson, the man who coined the term ‘hyperlink’ back in the Sixties and described the whole messy concept in Dream Machines, published in 1974.
Nelson is one of those inspired thinkers whose concepts don’t always make it into the practical world in the form that he would wish. For nearly six decades he has been working on Xanadu, a mysterious computer system to beat all systems, and a complex vision of a future reality that exists in a parallel universe to our own – the imminent release of which has been promised at regular intervals for years.
At last, a part of this vision has been realised with the release of OpenXanadu. “OpenXanadu means you can see all the parts. It is not yet open source,” although it is intended to be. OpenXanadu is a subset of the original vision and hopes to redefine the way we see documents. “We foresaw in 1960 that all document work would migrate to the interactive computer screen, so we could write in new ways,” wrote Nelson. “We screwed up in the Eighties, and missed our chance to be world wide hypertext (the Web got that niche). However, we can still compete with PDF, which simulates paper, by showing text connections”.
About contemporary computer systems Nelson has written: “Today’s popular software simulates paper. The World Wide Web (another imitation of paper) trivialises our original hypertext model with one-way ever- breaking links and no management of version or contents”. Nelson’s claim is that Xanadu solves these problems and more.
During the Eighties vast amounts of money were pumped into Xanadu by Autodesk and others, with the objective of coming up with something akin to the Web before the Web was invented, although notably more complex. The virtue of Berners-Lee’s solution was threefold: it was beautifully simple, accessible, and it worked. Xanadu was obscure and difficult and, despite numerous attempts, never successfully made the leap from concept to realisation.
In 1960, Nelson enrolled on a computer course for the humanities, and “was struck by a vision of what could be”. For his term project, written in Assembler on a mainframe, he attempted to devise a text-handling system that would allow writers to revise, compare and undo their work, years before the concept of word processors had been invented.
Some five years later he came up with the term ‘hypertext’. Since that date, Nelson has been working on his life’s project, a software framework which would be a “magic place of literary memory,” named after Coleridge’s Kubla Khan, and described at length in his 1981 book, Literary Machines. Xanadu is and was as illusory as Coleridge’s poem, which was famously “composed, in a sort of Reverie brought on by two grains of Opium taken to check a dysentery, at a Farm House between Porlock and Linton, a quarter of a mile from Culbone Church, in the fall of the year, 1797”, and was never completed because of the arrival of “a person on business from Porlock”, after which the memory of the dream escaped him.
In Dream Machines, published in 1974, Nelson promised that Xanadu would be released by 1976. His system, a welcome throwback to the ideals of the Sixties, would be used to “foster free expression and empower people where earlier forms of communication, TV, radio and print, were largely one-way, hierarchic, and disempowering.” In the 1987 edition of Literary Machines, the release date was revised to 1988. In January 1988, Nelson wrote in Byte magazine that Xanadu would be completed by 1991.
Amazingly, an initial version was released in August 1999 – apparently to prove prior art on patents that had been claimed on ideas that he had proposed years before – but has yet to see any kind of acceptance, its raison d’etre long since superseded by the Web we know today. A later manifestation of aspects of Xanadu was known as Zigzag. Nelson wrote of Zigzag: “We believe the computer world can be simplified and unified. Today, ordinary people must deal with an appalling variety of programs and mechanisms to maintain their information. We have discovered a new simplification based on one simple concept: a new, liberated form of data that shows itself in wild new ways…”.
Sometimes we need the man with the mad idea to show the way to others so that the mad idea may become reality – and maybe the Web owes more to Nelson than he is usually given credit for. Although he despises the mechanisms of the Web and regards XML and other web technologies as suboptimal, the Web and Wikipedia and Creative Commons have rendered much of Nelson’s vision redundant, although he did recognise the problem of the lack of renumeration for musicians, writers and artists, which has yet to be resolved by the real World Wide Web – although Nelson’s solution to this problem, a kind of net tax, was suboptimal for any number of reasons.
It is perhaps all this that makes Nelson quote Oscar Wilde: “Everybody’s waiting for me to die so they can say how much they appreciated my work. But nobody will back me.”