What people get wrong about Xanadu

Links are separate from documents. They aren’t IN documents. The author of a document doesn’t have any greater authority over links than anybody else — they can create a bunch, and users can choose to apply them or not. The host that contains a document does not necessarily have any special association with the author, nor does it need to perform any coordination with or acknowledgement of links.

Instead, it’s closer to the ‘Rap Genius’ model — arbitrary users make arbitrary connections, annotations, rearrangements, and re-skins that are applied to the already-downloaded ‘original content’, and the user gets to choose which he or she would like to apply.

Hypertext is not about an author carefully controlling their brand by creating associations. It’s about a bunch of nerds sharing their own connections and footnotes and doing mashups & remixes of existing text.

You know how Thomas Jefferson took a knife to a bible & cut out all the supernatural stuff, then published his truncated version? Part of the point of Xanadu is that you could do that with the newest Hunger Games book, or literally combine the text with Interview with a Vampire, or whatever, and nobody could complain because the authors were getting paid for the passages that were kept from the originals.

It’s social footnoting. That’s the whole point. It’s for fandoms.

Even though I’m not a fan of copyright or micropayments in general, this is why I think they make sense in the Xanadu context.

Rifftrax could exist in a Xanadu type system in a way that encourages movie producers to get riffed, for instance.

People couldn’t use DMCA takedowns to shut down critique, because there’s no grounds, because all of the quotes are getting paid for & the original context is connected to commentary *automatically*. On top of that, any remixing or rearrangment is being done on the client — an argument can be made that derivative works aren’t being distributed so much as instructions on how to modify someone’s personal copy of a piece of media.

The web, by not really putting much thought into how to do data storage or addressing, and by foolishly embedding links into documents, essentially cemented a hierarchical system in which the rich get richer — because control over domain names and access to expensive hardware was directly connected to ability to serve and control data.

It’s stupid to address a document based on a machine, which is why email got rid of routes & fidonet never had them.

On the web, if the little guy’s information gets too popular, he’s punished — his site gets slashdotted and goes down. So, the little guy decides he doesn’t want to run his own web server anymore — now he pays an already-big company to handle it, at a much higher price. If he doesn’t, his information goes away forever (unless somebody else takes it and gives it to the big company first). This problem was created by sticking hostnames in addresses & sticking links in documents.

From the beginning, even in models like xu88 and xu92 where the network was centralized, which machine a document happened to reside on was never part of addressing on Xanadu.

Versions currently in development have lost that, using HTTP for file storage, though big public caches and leaning heavily on archive.org are encouraged. (And there’s no reason they *can’t* be patched to support IPFS. I had experimental IPFS support on one I worked on.)

But, at least links and formatting are totally separate from documents. And, at least, you’ve got first-order mechanisms for rearrangement (EDLs, which are just lists of transclusions — assembly instructions for frankensteining new documents from disparate sources.)

None of these things are *technically* difficult. A full Xanadu-style hypertext system is easier to write than just the HTML parsing component of a modern web browser alone.

Note: I don’t speak for Project Xanadu in any official capacity. This piece is reflective of my understanding of intended use, guiding philosophy, and version history based on about six years working on Xanadu and ten previous years researching it.

Written by

Resident hypertext crank. Author of Big and Small Computing: Trajectories for the Future of Software.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store