Ted Nelson (one of the original prophets of the Information age) had recently started publishing Xanadu basics video series:
#1a, #1b, #1c, #1d, #2, #3, #4
Founded in 1960 by Ted Nelson, Project Xanadu was the first hypertext project, and is possibly the longest-running vaporware project in the history of computing.
The basic concept of Project Xanadu are visible connections between documents on the screen.
Person ostensibly at a computer screen (mock-up, because there were no computer screens in offices at that time) in an office, 1972.
World Wide Web was Nelson's idea back in 1960s, and for more than 25 years he thought he would be the one to create worldwide hypertext (and hypermedia), but apparently Tim Berners-Lee in 1989 created his own version (which Ted called a narrowed fork, because Xanadu was much more ambitious), which left out aformentioned visible connections.
Ted had imagined such system before individuals owned computers, and even before computer screens were available he could make it up in his own mind.
He argues that concept of computing was not supposed to be digital paper:
Today's popular software simulates paper.
The World Wide Web (another imitation of paper) trivializes our original hypertext model with one-way ever-breaking links and no management of version or contents.
Emphasis was stepping away from a classic page view (à la single paper sheet) to more dynamic and linkable kind of object.
We should go beyond the paper with our computers.
Interactive screen implementation of transpointing windows, 1999.
Xanadu vs the web
Essentially, Ted's vision was to create two-way (or actually multi-way) links, instead of one-way hyperlinks that we have now (which he called jumplinks before the web). Two-way links allow to connect every quotation in the document with it's original source. Every quotation can be connected to its source.
In system like that, there is no cut/copy-pasting, there is transclusion which in this case means pulling the quote from the original document, and keeping the path (connection) to the quote of the original document. Links do not break as document versions change, so it is possible to see the origins of every quotation.
Any changes made by the authors do not break any existing link. Links are connections between different contents, and transclusions are connections between same contents.
With ever-breaking one-way links and no version menagment some projects like web.archive.org and archive.li emerged to preserve timestamped version of requested document. Of course, not fully reliable as these can go off at any time and not every document is archived, but archiving the content in multiple 'archives' is the best form of decentralization that we can get now.
Edit Decision List (EDL) or Xanadoc File is a list of content to bring in, and links to apply to that content.
For pulling the external content, we would use EDL instead of HTML, which is very different.
EDL would look something like this: http://xanadu.com/SampleEDL.txt
A virtual document is composed of the items in the list. That virtual document is displayed to the user.
Some (JavaScript) examples of connected parallel pages (note that it's just one of the many possible interfaces):
http://xanadu.com/xanademos/MoeJusteOrigins.html
Connected documents are both for hypertext and hypermedia. This means that proposed system is not limited to text files (e.g. books, papers or blog posts).
One of the examples of "quoting" non-text files is having "studio version" and "director's cut" videos, but both would be played out of the same stream-file instead of different video lump-files which is currently the case.
We can conclude that Xanadu is highly different from the Web.
Transcopyright - a valid alternative copyright system
Xanadu makes content available with certain permissions. Content is distributed in form of a list (EDL) which is fullfilled with rights and possibilities, and browser then obtains the content separately. It allows mixing of both free and paid content. A notion is that anyone may use any amount of rightsholders' content in their EDL. A path back must be provided, and the path back allows user to access the content's original context.
Thus anyone can quote at will, and without negotations.
Transpublishing - an alternative approach to copyright
It allows individuals to publish online documents that include other copyrighted content, without negotiations. An author may legally (without violating the copyrights) use this approach to quote from other documents, inside a new document, without even contacting the original owner. This is allowed because the contect effectively comes from the original document, so the original publisher is the one supplying the portion (quote) of that document. Since there is an immediately available connection to the original , quotes are not out of context. This also means that if the published content is paid, users have to pay for each downloaded portion.
Micropayments
Since it's possible to combine both free and priced content, Xanadu required a micropayment system for transpublishing. It should be possible to set a price per character, so for example 1000 characters could cost 1 cent (presumably, authors could set arbitrary price). Because of the relatively low cost of the content, system would automatically purchase independent parts (fractions) of a composed document so that it doesn't annoy users or interrupt their experience. Of course, user could also set arbitrary treshold which would alert him that transaction might be too expensive for an automatic purchase. This seemed to be suitable idea for selling small amounts of content. Users should have complete anonymity.
One of the working prototypes of such micropayment system was HyperCoin, implemented 1996 by Andrew Pam, but likely it didn't have all the necessary features.
Why Xanadu never took off
In 60s, Ted was not a hacker, engineer nor a computer scientist.
He was a philosopher, artist and a visionary, and was going to be a show-business guy.
However, his ideas inspired many computer scientists. Perhaps if he was a hacker we might've gotten his version of the web sooner, because his original visions were brilliant.
Status quo
Because of Web having one-way links, instead of two-way links which Ted envisioned, someone (e.g. Google) had to constantly scrape the whole Internet and dig every backlink that might have been there and keep track where the most of them were in order to sell that information as a service (for example to advertisers). Thus, incomplete system design centralized a gigantic portion of income that shouldn't even exist in the first place. Had Ted's vision (Web with two-way links) became a reality, all someone would have to do was count up all those things and find where most links are and that would be "Google", except all that information would be public instead of private.
So ironically, filling the gaps between Xanadu and our current Web, created one of the biggest fortunes in the history.
Ted's speech at Internet Archive Christmas lunch, 2017:
I only believe in one thing now... and that's the Internet Archive!
We need something which will remedy this problem of one-way ever-breaking links and no management of version or contents.
But is the Internet Archive really a dead end? Or a new high-level protocol for the Internet can emerge?