Project Xanadu

Before having read through The New Media Reader recently, I had never heard of Project Xanadu. And yet, reading Ted Nelson’s descriptions of it, starting in the 1960s, felt like reading someone in the past copying the ideas right out of my head—sometimes word-for-word. You may ask, what has become of Xanadu, given that it’s supposedly been in development for decades?

The short story is that decades after first proposing the idea, in 1998, Ted finally released some of the source code, in hopes of invalidating several patents that ignored his prior art. Then it sat again for over a decade, until in 2014 a demonstration version was released. Having read through some of his statements, as well as some notes by individual that were also involved with the project, I’ve taken away several key warnings:

• Xanadu was initially proposed in the 60s, before even the original computer intranets were widely available. Although the ideas were later refined, they were still beholden to the idea of some sort of central repository. As computer networks, and the internet, arose and gradually moved further and further away from that model, it became harder to graft that idea onto the new world.
• One of Ted’s primary concerns with Xanadu, and transclusions (basically “windows” of text from a source document into a new document, without being cut-and-pasted) was proper attribution and credit. This is admirable, and possibly something that will still need to be refined. However, years were spent on developing a credit system for authors, based on how much text was transcluded, putting the cart well before the horse.
• There were literal decades between the first proposal and the first line of code; it wasn’t until the mid-80s that Xanadu left the drawing board, and with the financial backing of Autodesk, started to be turned into code. However, eventually there were disagreements that led to Ted leaving Autodesk. In the 90s and 2000s he again had a few coders moving the project forward, but, from what I can tell, each individual never worked on it for long.
• It seems that Ted rigidly held on to his view of what Xanadu should be, based on his ideas from the 60s and 70s, which made it difficult to recruit developers, that were using languages and technologies that had developed in completely different eras.
• It also made it difficult to implement, as computers and their user interfaces, according to Ted, were “wrong”. How can you develop an application, when the tools and interface paradigms are so far from the idea in your head? One cannot, easily, use a screw driver to hammer nails.

What I took away from reading the history of Xanadu is that if Lexiator, and the larger, currently-unnamed project, are to ever see the light of day, these mistakes should not be repeated.

• More developer involvement. Since I cannot hire developers, Open Source, a model used successfully on thousands of projects, is the way to go. I cannot force anyone to join, but hopefully others will be interested enough to contribute.
• Use the tools that exist now. Even if my hope is to bring to light many of Ted’s ideas, we have to use the tools and paradigms that currently exist. And not esoteric ones, but tools that are widely familiar and easy to pick up.
• Perfect is the enemy of good. There are many details, such as attribution, that will have to be figured out along the way. But knowing those bridges will have to be crossed, doesn’t mean development should wait to figure out perfect solutions.