Archive of UserLand's first discussion group, started October 5, 1998.

Xanadu--some initial reactions

Author:Dennis Peterson
Posted:8/26/1999; 10:15:53 AM
Topic:Xanadu--some initial reactions
Msg #:9959
Prev/Next:9958 / 9960

1) The parallel documents idea is interesting, but the idea of putting markup parallel (outside of) the document misses the point of XML. From one of Ted's papers: "The currently popular model of SGML (and its derivatives HTML, XML) scrambles together two intrinsically different kinds of data which are intrinsically parallel. "

The idea here seems to be that the text is one kind of data, and the markup is another. This is valid for HTML. What he doesn't seem to get is that the data contained within XML tags is lots of different kinds of data, with the data type defined by the tags. It would be nonsensical to separate out these tags into a parallel stream as he describes, so that we can apply different tags to the bytestream.

It makes sense if you're thinking of text-related tags like "paragraph" or "subhead." But if you have tags like "price" or "creation date," tags that denote meaning instead of just organization, it's better to have those tags firmly fixed in place. Swapping tags around would be like doing random typecasts.

2) Transpublishing is the idea that you can include part of someone else's document in yours, and you don't have to worry about copyright because the bytes actually come from their server. This has already been done to a limited extent with frames, and I remember at least one lawsuit that resulted. Not to mention the recent spat about deep linking. These problems are motivated by the fact that advertising is circumvented.

3) Of course, the Xanadu idea is that you'll pay for those bytes directly. But there seems to be an assumption that everyone will play nice, no one will make free copies. There doesn't seem to be any real enforcement mechanism, which isn't surprising because real enforcement is probably impossible--sooner or later you have to generate plaintext for the user to see, and if there's plaintext there's a crack. There has to be a different payment model.

There certainly are a lot of interesting ideas in the system, though. They may not revolutionize the web just yet, but their code could certainly make for some nifty editors and groupware. Collaborative editing with every version efficiently preserved, where you can recombine versions in separate forks, however you stuff.

There are responses to this message:

This page was archived on 6/13/2001; 4:52:08 PM.

© Copyright 1998-2001 UserLand Software, Inc.