4 ms·

Ted Nelson has some novel ideas. But I can't help to think that the failure of the mass adoption of his ideas stems from him being too greedy. If he had open so

Ted Nelson has some novel ideas. But I can't help to think that the failure of the mass adoption of his ideas stems from him being too greedy. If he had open sourced the code he wrote for Xanadu, if he had even published specs of a simple Xanadu system free of copyright he could've brought us a better hypertext. The idea that data should consist of small addressable chunks, with each chunk having relations with other data actually makes a lot of sense. We are thinking of data linearly, you start from byte 1 to end. Even if you copy the file or make minor changes the OS considers it another file. You make a reference to a movie, a novel etc., if you do not explicitly add the reference no one knows. Then we end up in piles and piles of data we cannot process. We have to rely on proprietary search engines, we have to rely on our browser history and our memory. And with the latest shiniest web apps that fully render in JS we can't even get addressable future-proof links. Try to access a news page from 10 years ago, most likely the link won't work. Archive.org helps in that regard. But things could've been easier. We have all these resources at hand, we could implement the idea of transclusion even if it is only limited to our domain. Information-heavy sites such as Wikipedia, C2-wiki etc. would greatly benefit from it. It's true that we need a new way to look at data. It is not pen and paper anymore, we need to be able to go beyond searching. Even now, most of the world uses plain-old PDF, which is an image of a printable paper. We can use animations, videos, sound, interactivity in our content, but mostly we don't. Ted Nelson thinks this is caused by the invention of the GUI (or PUI -Parc user interface- as he calls it) in Xerox Parc. They needed to show their managers something they could make sense of. So they created the desktop. Here is your trash-bin, here is your notepad, here are your documents (e.g Postscript/PDF). It is just like your physical desk. But it is not your physical desk, what you have in front of you is a computer. We teach hours and hours of hand-writing to hour children, but only very few countries teach touch-typing. The old generation either doesn't understand or doesn't want kids to utilize their full potential with computers. Fortunately kids learn on their own, build their own online communities, so there is hope. Any kid that can get his/her hands on Computer Lib/Dream Machines won't be able to believe the idea that computers are only good for showing us an enhanced simulation of a work-desk.

tudorw 9y ago
Open Source is a contemporary invention introduce by Eric S. Raymond in 1997, so any failure of Ted to utilise it is a bit unfair as he was living in a different era. For me the key take away from Ted is the democratisation of computer use, without Ted there is a very real chance we never would have had fully functioning stand alone PC's, it was Teds discussions with Bill Lower at IBM that persuaded them there would be a personal computer market, and that as a creative media tool, it would find a home in every house, thanks Ted :) Imagine a world like now, except we only ever got dumb terminals to connect into Facebook, etc, sure 98% of the world might not notice the difference, but the flexible computing power in the hands of the 2% has massive potential, as the open source movement has adequately demonstrated!
- e12e 9y ago
  Eh, bsd is a bit older than that. Raymond certainly caught the buzzword wave with "the Cathedral and the Bazaar" - but he wrote it after Gnu was started, which enabled Torvalds to write/release Linux etc. Raymond offered an explanation of how Free software could be subsumed by business. If just some of Nelson's books had seen wider distribution, that might have helped a lot. It's easy to link to Fielding's thesis, because it's online: http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm And I actually think some of Nelson's ideas form a fine counter to Fielding's "sales pitch" for REST. REST is all about achieving high performance alongside correctness in distributed document/media systems. But as Fielding details in his introduction: there are other architectures that make more sense for what is effectively distributed computing: when you need features beyond what REST can efficiently support. I believe Nelson had some compelling arguments for why we might want more. And if the contemporary Web is any indication, he was right. There's a place for REST-style hypertext, and there's a place for richer hypermedia. I get as angry as the next grumpy old person when a simple Web page is turned into a single-page application for no gain and considerable pain; but I also see that there are things we need beyond simple hypertext (which we've tried to fill with: excel macros with floppies as the transport medium, java applets, active-x, flash...). And there's the question of naming/cataloging. It's doubtful DNS is the best we can do.

tudorw 9y ago

Open Source is a contemporary invention introduce by Eric S. Raymond in 1997, so any failure of Ted to utilise it is a bit unfair as he was living in a different era. For me the key take away from Ted is the democratisation of computer use, without Ted there is a very real chance we never would have had fully functioning stand alone PC's, it was Teds discussions with Bill Lower at IBM that persuaded them there would be a personal computer market, and that as a creative media tool, it would find a home in every house, thanks Ted :) Imagine a world like now, except we only ever got dumb terminals to connect into Facebook, etc, sure 98% of the world might not notice the difference, but the flexible computing power in the hands of the 2% has massive potential, as the open source movement has adequately demonstrated!

e12e 9y ago
Eh, bsd is a bit older than that. Raymond certainly caught the buzzword wave with "the Cathedral and the Bazaar" - but he wrote it after Gnu was started, which enabled Torvalds to write/release Linux etc. Raymond offered an explanation of how Free software could be subsumed by business. If just some of Nelson's books had seen wider distribution, that might have helped a lot. It's easy to link to Fielding's thesis, because it's online: http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm http://www.ics.uci.edu/~fielding/pubs/dissertation/top.htm And I actually think some of Nelson's ideas form a fine counter to Fielding's "sales pitch" for REST. REST is all about achieving high performance alongside correctness in distributed document/media systems. But as Fielding details in his introduction: there are other architectures that make more sense for what is effectively distributed computing: when you need features beyond what REST can efficiently support. I believe Nelson had some compelling arguments for why we might want more. And if the contemporary Web is any indication, he was right. There's a place for REST-style hypertext, and there's a place for richer hypermedia. I get as angry as the next grumpy old person when a simple Web page is turned into a single-page application for no gain and considerable pain; but I also see that there are things we need beyond simple hypertext (which we've tried to fill with: excel macros with floppies as the transport medium, java applets, active-x, flash...). And there's the question of naming/cataloging. It's doubtful DNS is the best we can do.