Written by David Tebbutt, MacUser Aug 1988 (guess) - scanned

One of my heroes is a man called Ted Nelson. About 10 years ago, when I was editor of Personal Computer World, his writing inspired me to write an ideas processor. This program, which contains today's hot concept, HyperText links, was duly published and has been bought by several thousand people, so I have good reason to feel grateful to Mr Nelson.

Nelson was the man who first invented the term 'Hypertext' and, for that matter, 'HyperMedia'. He modestly claims that the actual ideas go back to work done by Vannevar Bush in the 1940s, but it was Nelson who popularised the concepts through his Computer Lib/Dream Machines and Literary Machines books, published in 1974 and 1981 respectively. It's interesting that the concepts behind the Macintosh user interface also date back to 1974.

A few weeks ago I jumped at the chance to meet Nelson in Paris. He was over to promote a new HyperText program for PCs called HyperDoc. This is of little interest to you. What may be of interest, though, is an insight into Nelson's dreams and expectations.

To give an idea of the man's style, here's a quote from the middle of a long talk he gave in Paris: "I can't understand why people spend more on their cars than on their computers. A car only carries the body around from place to place. A computer carries your mind to the end of the universe."

To describe the man as an enthusiast would be a gross understatement. He has visions and he sweeps his audience along with him. Well, most of them. Several people at my table thought he was bonkers. They were mentally shackled by their long exposure to traditional mainframe computing.

While he came as a bit of a shock, I didn't think he was mad. Far from it. He just happens to live about 20 or 30 years in the future all the time. He first cooked up the terms HyperText and HyperMedia 29 years ago. He published them five years after that. These are the ideas which are considered 'state of the art' in today's computer world. And it wasn't simply a case of cooking up the terms, he knew exactly what they meant. That's no mean feat when you consider that in those days an advanced computer boasted punched cards and teletype terminals.

For years now, Nelson has been working on a project which he calls Xanadu. His vision is of a worldwide network of connected computers through which everyone has access to all electronically published material. Billions of users will have access to billions of documents and every access will result in a royalty credit being totted up for the original author. The dream ignores technological and political barriers but, in theory, it can be made to work. Whether it will be allowed to work is another matter.

The Xanadu team has worked out how these individual documents may be accessed swiftly, no matter where they are stored. Those interested in the details could do no better than read Nelson's Literary Machines. The most recent version of the book (87.1) is, in fact, a system specification for the Xanadu project as it stood in 1987.

Let's say you wanted to publish something in Xanadu about Ted Nelson. You might decide to cite a paragraph out of the Xanadu copy of this article. You'd be able to do this by incorporating a pointer in your document directly to this document. As people read your work, my paragraph would appear on their screen, just as if it were an integral part of your document. At that moment, the bytes being read from my document would cause my royalties to increase instead of yours.

Every piece of literature need only be stored once. All documents can refer to others and a complete audit trail of the development of a document could be online for all to see. Nelson sees Xanadu as "the manifest destiny of literature". All these crosslinks mean that the Xanadu 'Docuverse' is, in fact, a huge HyperText system.

Last year another visionary, John Walker, decided to buy into Nelson's Xanadu project. Walker is the head of Autodesk, the hugely wealthy AutoCAD company, and he knew that Nelson needed funding if Xanadu was ever going to get anywhere. He bought up Nelson's project and has a team working to a 1989 release of the first version of the program. Nelson, who is probably unemployable in the conventional sense, is the dreamer and the inspiration on call to the project team. The team, based in Palo Alto, has its collective head down for a winter launch of a single-user Xanadu product.

At the announcement of his takeover of Xanadu, Walker said, "We believe that Xanadu will be as important to everyone who thinks, as television is to everyone who doesn't think."

Doesn't this all remind you of the early days of the Macintosh? The days of the 'insanely great' little computer. The chances are there's something really big brewing and you, dear readers, are due to have an early involvement. The Macintosh is high on Xanadu's list of implementation platforms.

The global version of Xanadu poses special problems. The distributed nature of the project means that central control is impossible. The team had to find a consistent way of creating, remembering and using unique addresses for all documents and part documents wherever they may lie in the Docuverse. Mark Miller, one of the team members, invented the 'Tumbler' system of addressing items. This borrows a little from the Dewey decimal system used in libraries, a little from the tumblers in locking mechanisms, and much from Miller's powerful imagination; (See Literary Machines for an explanation of the Tumbler.)

Nelson believes that if the system is cheap enough to access, then people may be tempted to keep personal information in closed sections of Xanadu. At least, he says, this would safeguard it from damp, rodents and other conventional threats. He talks of the 1000 boxes of papers that he has to deal with every time he moves house. Now he lives on a boat, his need for Xanadu must be becoming quite desperate.

The planet may have a serious need of Xanadu too. The population keeps growing. Ever greater numbers of trees are torn down to produce the paper on which carbon is impressed. Huge amounts of petrol are gobbled up to deliver the printed word. This can all be done electronically. Access to any document can be swift and cheap. Researchers into urgent ecological topics could access to each others' findings within minutes of publication.

In recognition of the importance of the Macintosh, Autodesk has chosen it as the second Xanadu platform, behind Sun and ahead of IBM.