I’ve been fascinated with Ted Nelson’s work for years and the Xanadu project has always intrigued me. It’s great that Ted is still commited to sharing his ideas and making the concepts easy to understand for people. A lot of critics say that he caused his own downfall with Xanadu by being difficult to work with, but I feel like we see this a lot from those who don’t want to sacrifice their dreams to spawn inferior versions of a given product.
It was impressive seeing some of these concepts come together somewhat in the Xanadu deliverable that was released a few years back, but a small part of me thinks that wikis could be a good way to express Xanadu concepts in a more accessible way.
Wiki software has incorporated some Xanadu ideas, but largely in loose ways that miss key attributes. For instance, mediawiki has transclusions but doesn’t use them for versioning or make the connection to the original context clear to users, and while it has a bidirectional link tree internally, it doesn’t expose that information to the user. Nothing, aside from Xanadu demos, has visible connections (aka transpointing windows) in normal document viewing and editing (though some things, including mediawiki, use something that resmbles transpointing windows in diff displays).
Having worked for Ted, I can verify that he’s “hard to work with”, but not in the way people usually use that term. He’s polite and cordial. Mostly, he gets easily off-track during meetings & has a hard time finding work he’s already done (and so he often ends up re-doing it). He’s got all sorts of tech & filing systems to help him with organization these days, so it must have been more difficult for him to act as a manager in the 70s. (Of course, during the Autodesk era, Ted had nothing to do with the day to day development of Xanadu. They had their own management difficulties, resulting in schisms and abandoned work, some of which is now open source.)
Xanadu has changed a lot over the years, and made a lot of concessions. The past couple demos to have public release actually were in the browser & used regular HTTP for fetching documents – a huge concession, since HTTP doesn’t have any guarantees about content stability. The previous demo (XanaduSpace) also used HTTP. The stuff that Ted is bullish about is mostly stuff that works together in subtle but important ways – ex., transclusion-based versioning makes transcopyright straightforward & visible connections solve the droit moral problem of remixing somebody’s work without their explicit permission (since it acts as a very precise automatic attribution), & all this depends upon permanent addresses with unchanging content (or, in some cases, append-only content).
I’ve been fascinated with Ted Nelson’s work for years and the Xanadu project has always intrigued me. It’s great that Ted is still commited to sharing his ideas and making the concepts easy to understand for people. A lot of critics say that he caused his own downfall with Xanadu by being difficult to work with, but I feel like we see this a lot from those who don’t want to sacrifice their dreams to spawn inferior versions of a given product.
It was impressive seeing some of these concepts come together somewhat in the Xanadu deliverable that was released a few years back, but a small part of me thinks that wikis could be a good way to express Xanadu concepts in a more accessible way.
Wiki software has incorporated some Xanadu ideas, but largely in loose ways that miss key attributes. For instance, mediawiki has transclusions but doesn’t use them for versioning or make the connection to the original context clear to users, and while it has a bidirectional link tree internally, it doesn’t expose that information to the user. Nothing, aside from Xanadu demos, has visible connections (aka transpointing windows) in normal document viewing and editing (though some things, including mediawiki, use something that resmbles transpointing windows in diff displays).
Having worked for Ted, I can verify that he’s “hard to work with”, but not in the way people usually use that term. He’s polite and cordial. Mostly, he gets easily off-track during meetings & has a hard time finding work he’s already done (and so he often ends up re-doing it). He’s got all sorts of tech & filing systems to help him with organization these days, so it must have been more difficult for him to act as a manager in the 70s. (Of course, during the Autodesk era, Ted had nothing to do with the day to day development of Xanadu. They had their own management difficulties, resulting in schisms and abandoned work, some of which is now open source.)
Xanadu has changed a lot over the years, and made a lot of concessions. The past couple demos to have public release actually were in the browser & used regular HTTP for fetching documents – a huge concession, since HTTP doesn’t have any guarantees about content stability. The previous demo (XanaduSpace) also used HTTP. The stuff that Ted is bullish about is mostly stuff that works together in subtle but important ways – ex., transclusion-based versioning makes transcopyright straightforward & visible connections solve the droit moral problem of remixing somebody’s work without their explicit permission (since it acts as a very precise automatic attribution), & all this depends upon permanent addresses with unchanging content (or, in some cases, append-only content).