5 ms·

Upvoted in the hope someone will explain this to me

canpan 9y ago
I remember that being featured in the anime Serial Experiments Lain. Here is the Wikipedia article: https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu Never really understood what it is exactly for though. Which might explain why it never took off...
pedrocr 9y ago
It's a long dream of a better-than-the-Web hypertext ecosystem that predates the Web by 15 years but has never delivered anything actually useful[1]. A great example of why "worse is better"[2]. Even if you do have a grandiose idea that is better than anything that came before or will come soon there's a good chance you will be beaten by simpler systems that can be delivered much faster. [1] https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu [2] https://en.wikipedia.org/wiki/Worse_is_better https://en.wikipedia.org/wiki/Worse_is_better
firasd 9y ago
It's a visionary hypertext project that doesn't really work yet, but the people behind it are happy to dispense snarky quotes about the how the web we all use is a poor substitute because it's a “bizarre structure created by arbitrary initiatives of varied people and it has a terrible programming language”. https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu Looks like this is their latest demo: https://www.youtube.com/watch?v=72M5kcnAL-4 https://www.youtube.com/watch?v=72M5kcnAL-4 The fact that they seem to be so interested in things like "who pays when you mix and match copyrighted content?" instead of "how do we get people to try out our system?" indicates that it's not going to be in widespread production use anytime soon...
- Animats 9y ago
  Yes. Everything is pay per view in Xanadu. I knew that crowd when Autodesk was funding them. They were fanatical libertarians. The solution to everything is a market. Ted Nelson was focused on text to the exclusion of images. That didn't help. He also had strong ideas about how the internal database should be organized, down to the database implementation level, which didn't help either, because they were 1960s database technology.[1] The closest thing to Xanadu in wide use today is Github. One central place for all public code. Anything can be changed but the history and old versions remain. Full tracking of who changed what. The ability to create a completely new document from an old one while retaining tracking. There's no micropayment system, of course. The real conceptual failure was not realizing that the future was ad-supported. [1] http://dubinko.info/blog/2009/11/22/how-xanadu-works/ http://dubinko.info/blog/2009/11/22/how-xanadu-works/
  - keithpeter 9y ago
    What was your perception of the internal dynamics of the group during the Autodesk phase? Did it strike you as effective/likely to produce something?
    - Animats 9y ago
      They had some really good people, but did seem to be spinning their wheels. It was completely separate from Autodesk. Xanadu had offices on California Avenue in Palo Alto; Autodesk was in Marin County. Autodesk was quite good at getting working software out the door, but that didn't reach Xanadu. Lack of an identified market was a big problem. One market that was considered was financial newsletters. There are lots of expensive newsletters you can buy which have extremely detailed coverage of something important to traders, such as what's happening at the Henry Hub or the Port of Rotterdam or in the Bahamas. These cost $50-$1000 a month. ("Offshore Alert", which covers scams run from tiny island countries, "Platts Oil Letter", which is now a big service, and the Dines Letter, for the gold bugs.) There was no centralized distribution for these, and they had tiny circulations. That didn't go anywhere. All of those are now on the Web as subscription products.
      - keithpeter 9y ago
        Interesting observation - so there was some consideration of a specific information market rather than the 'Kinko's for information' idea that Nelson has described in the Wired article. It was more the development process that I was thinking of - the actual production of software, a working system. I take your point that Autodesk's programmers could certainly produce software. I was wondering if the Xanadu project was seen by Autodesk perhaps as a bit of a 'moon-shot', in the sense of low probability of success coupled with great potential.
        Animats 9y ago
        I was wondering if the Xanadu project was seen by Autodesk perhaps as a bit of a 'moon-shot', in the sense of low probability of success coupled with great potential. Yes, it was. Autodesk has done a few of those. They were into virtual reality very early, before 1990. It turned out that doing CAD with gloves and goggles was like trying to paint wearing oven mitts. Precise positioning is tough. So they gave up on that. Getting into 3D animation worked out better. Autodesk itself did 3D Studio, and gradually acquired other 3D animation companies as they got into trouble. This worked out well, and there was crossover with the CAD business. The really good radioiosity renderer from one of the animation programs was connected to an architectural design program. Now you could place real light fixtures from a library and see the result before building the building and embedding light fixtures in concrete. Carl Bass, as CEO until last February, was into 3D printing. He thought home 3D printing was going to be a much bigger deal than it turned out to be. So Autodesk came out with several low-end CAD programs that were never very profitable and did a lot of work to make 3D printing easier. Not profitable. Bass is gone. Autodesk keeps plugging away at boring but useful stuff. Market cap around $25 billion. (I'm long out of Autodesk, although still a stockholder.)
- leoc 9y ago
  > It's a visionary hypertext project that doesn't really work yet, but the people behind it are happy to dispense snarky quotes about the how the web we all use is a poor substitute because it's a “bizarre structure created by arbitrary initiatives of varied people and it has a terrible programming language”. This is a bit like describing Galileo as "the jackass who wouldn't shut up about heliocentrism despite having no answer to the stellar parallax objection". By no means has he ever been right about everything, and he has been his own worst enemy in several ways, but ...
occultist_throw 9y ago
Long and short of it, it's a form of bidirectional links (so if you update one side, the other side automatically updates). This prevents link-rot. The second is putting a payment processor on EVERY query. You want Internet for the Richies- this is how you get it. Of course, there's hand-wavey garbage that says 'everyone would contribute and nobody would owe anyone anything'. Seriously, IPFS is the best way forward here. Links never change its content (but can still die if noone has the content). And no payment processor is there unless you want to pay other people to pin (read: seed) content. (To see this post, please pay $0.02) Edit: Seriously? multiple -1's in a row? I'd love to hear rebuttals if I'm wrong, or a further discussion. If you've read the Xanadu documents, then you'd know that the basis of this is a "Secure Environment and Secure Identification" of computers. AKA: DRM.
- pavlov 9y ago
  > Internet for the Richies Isn't that what we effectively have, with so much of the Internet's value either locked in Facebook's garden or beholden to Google? I understand where you're coming from with the DRM criticism, but IMO the primary fault in Xanadu's version is that it's an attempt at naïve mom&pop capitalism in a digital world that already is owned by plutocrats wealthy and powerful beyond Rockefellers' dreams.
  - jancsika 9y ago
    > Isn't that what we effectively have, with so much of the Internet's value either locked in Facebook's garden or beholden to Google? No, because you can download the entirety of Wikipedia, including its edit history and discussion. Also, you can download a not insignificant number of primary sources through both sci-hub and various other sources like Project Gutenberg, IMSLP, etc. Add to that all the things that swell to prominence on Bittorrent trackers whenever the Streisand effect is in play. I have a hard time seeing that content disappearing any time soon. But I also have a hard time seeing how that content would have been aggregated in the first place if you had little micropayment barriers between all the things.
    - 9y ago
detaro 9y ago
One of the (arguably the) original hypertext visions (AFAIK Ted Nelson, its creator, coined the term "hypertext" in the early 60s). The main difference to web hypertext as we know it is that it doesn't just link documents together by what we know as "links", but it's bidirectional and with more content inclusion. E.g. if someone takes content from another document and puts it in their article, the text is connected to the source document (instead of them maybe putting a link to the source). And from the source document you can find all documents that used parts of it. Project Xanadu is Nelson's repeated attempts to make this work, while the world choose the WWW model and is happy enough with that (much to the disdain of Xanadu supporters, who believe the WWW to be a vastly inferior and broken approach). Once HTML took off, there wasn't really a chance to catch up, especially not with the limited resources they have. Lot's of bitterness about that, which also means they haven't really tried to make their ideas work with the web and rather aiming for an uncompromising, pure implementation of their ideas.
johnchristopher 9y ago
One of the best wired article ever and about that subject is https://www.wired.com/1995/06/xanadu/ https://www.wired.com/1995/06/xanadu/ and it should entertain you for an evening. Presume I am dead if I ever stop reposting this.
- leoc 9y ago
  If that's the case could you please post Nelson's responses http://web.archive.org/web/20001101230424/http://www2.educ.ksu.edu/Faculty/McGrathD/Fall99/NelsonLtr.htm http://web.archive.org/web/20001101230424/http://www2.educ.k... http://web.archive.org/web/20001003011753/http://xanadu.com.au/ararat http://web.archive.org/web/20001003011753/http://xanadu.com.... along with it.
yantrams 9y ago
Longform article on the history of Xanadu - https://www.wired.com/1995/06/xanadu/ https://www.wired.com/1995/06/xanadu/
- shpx 9y ago
  Discussed on hn https://news.ycombinator.com/item?id=795155 https://news.ycombinator.com/item?id=795155 18 comments Previous discussion of xanadu in general https://news.ycombinator.com/item?id=10642143 https://news.ycombinator.com/item?id=10642143 6 comments https://news.ycombinator.com/item?id=7849389 https://news.ycombinator.com/item?id=7849389 100 comments

I remember that being featured in the anime Serial Experiments Lain. Here is the Wikipedia article: https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu Never really understood what it is exactly for though. Which might explain why it never took off...

pedrocr 9y ago

It's a long dream of a better-than-the-Web hypertext ecosystem that predates the Web by 15 years but has never delivered anything actually useful[1]. A great example of why "worse is better"[2]. Even if you do have a grandiose idea that is better than anything that came before or will come soon there's a good chance you will be beaten by simpler systems that can be delivered much faster. [1] https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu [2] https://en.wikipedia.org/wiki/Worse_is_better https://en.wikipedia.org/wiki/Worse_is_better

firasd 9y ago

It's a visionary hypertext project that doesn't really work yet, but the people behind it are happy to dispense snarky quotes about the how the web we all use is a poor substitute because it's a “bizarre structure created by arbitrary initiatives of varied people and it has a terrible programming language”. https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu Looks like this is their latest demo: https://www.youtube.com/watch?v=72M5kcnAL-4 https://www.youtube.com/watch?v=72M5kcnAL-4 The fact that they seem to be so interested in things like "who pays when you mix and match copyrighted content?" instead of "how do we get people to try out our system?" indicates that it's not going to be in widespread production use anytime soon...

Animats 9y ago
Yes. Everything is pay per view in Xanadu. I knew that crowd when Autodesk was funding them. They were fanatical libertarians. The solution to everything is a market. Ted Nelson was focused on text to the exclusion of images. That didn't help. He also had strong ideas about how the internal database should be organized, down to the database implementation level, which didn't help either, because they were 1960s database technology.[1] The closest thing to Xanadu in wide use today is Github. One central place for all public code. Anything can be changed but the history and old versions remain. Full tracking of who changed what. The ability to create a completely new document from an old one while retaining tracking. There's no micropayment system, of course. The real conceptual failure was not realizing that the future was ad-supported. [1] http://dubinko.info/blog/2009/11/22/how-xanadu-works/ http://dubinko.info/blog/2009/11/22/how-xanadu-works/
- keithpeter 9y ago
  What was your perception of the internal dynamics of the group during the Autodesk phase? Did it strike you as effective/likely to produce something?
  - Animats 9y ago
    They had some really good people, but did seem to be spinning their wheels. It was completely separate from Autodesk. Xanadu had offices on California Avenue in Palo Alto; Autodesk was in Marin County. Autodesk was quite good at getting working software out the door, but that didn't reach Xanadu. Lack of an identified market was a big problem. One market that was considered was financial newsletters. There are lots of expensive newsletters you can buy which have extremely detailed coverage of something important to traders, such as what's happening at the Henry Hub or the Port of Rotterdam or in the Bahamas. These cost $50-$1000 a month. ("Offshore Alert", which covers scams run from tiny island countries, "Platts Oil Letter", which is now a big service, and the Dines Letter, for the gold bugs.) There was no centralized distribution for these, and they had tiny circulations. That didn't go anywhere. All of those are now on the Web as subscription products.
    - keithpeter 9y ago
      Interesting observation - so there was some consideration of a specific information market rather than the 'Kinko's for information' idea that Nelson has described in the Wired article. It was more the development process that I was thinking of - the actual production of software, a working system. I take your point that Autodesk's programmers could certainly produce software. I was wondering if the Xanadu project was seen by Autodesk perhaps as a bit of a 'moon-shot', in the sense of low probability of success coupled with great potential.
      - Animats 9y ago
        I was wondering if the Xanadu project was seen by Autodesk perhaps as a bit of a 'moon-shot', in the sense of low probability of success coupled with great potential. Yes, it was. Autodesk has done a few of those. They were into virtual reality very early, before 1990. It turned out that doing CAD with gloves and goggles was like trying to paint wearing oven mitts. Precise positioning is tough. So they gave up on that. Getting into 3D animation worked out better. Autodesk itself did 3D Studio, and gradually acquired other 3D animation companies as they got into trouble. This worked out well, and there was crossover with the CAD business. The really good radioiosity renderer from one of the animation programs was connected to an architectural design program. Now you could place real light fixtures from a library and see the result before building the building and embedding light fixtures in concrete. Carl Bass, as CEO until last February, was into 3D printing. He thought home 3D printing was going to be a much bigger deal than it turned out to be. So Autodesk came out with several low-end CAD programs that were never very profitable and did a lot of work to make 3D printing easier. Not profitable. Bass is gone. Autodesk keeps plugging away at boring but useful stuff. Market cap around $25 billion. (I'm long out of Autodesk, although still a stockholder.)
leoc 9y ago
> It's a visionary hypertext project that doesn't really work yet, but the people behind it are happy to dispense snarky quotes about the how the web we all use is a poor substitute because it's a “bizarre structure created by arbitrary initiatives of varied people and it has a terrible programming language”. This is a bit like describing Galileo as "the jackass who wouldn't shut up about heliocentrism despite having no answer to the stellar parallax objection". By no means has he ever been right about everything, and he has been his own worst enemy in several ways, but ...

occultist_throw 9y ago

Long and short of it, it's a form of bidirectional links (so if you update one side, the other side automatically updates). This prevents link-rot. The second is putting a payment processor on EVERY query. You want Internet for the Richies- this is how you get it. Of course, there's hand-wavey garbage that says 'everyone would contribute and nobody would owe anyone anything'. Seriously, IPFS is the best way forward here. Links never change its content (but can still die if noone has the content). And no payment processor is there unless you want to pay other people to pin (read: seed) content. (To see this post, please pay $0.02) Edit: Seriously? multiple -1's in a row? I'd love to hear rebuttals if I'm wrong, or a further discussion. If you've read the Xanadu documents, then you'd know that the basis of this is a "Secure Environment and Secure Identification" of computers. AKA: DRM.

pavlov 9y ago
> Internet for the Richies Isn't that what we effectively have, with so much of the Internet's value either locked in Facebook's garden or beholden to Google? I understand where you're coming from with the DRM criticism, but IMO the primary fault in Xanadu's version is that it's an attempt at naïve mom&pop capitalism in a digital world that already is owned by plutocrats wealthy and powerful beyond Rockefellers' dreams.
- jancsika 9y ago
  > Isn't that what we effectively have, with so much of the Internet's value either locked in Facebook's garden or beholden to Google? No, because you can download the entirety of Wikipedia, including its edit history and discussion. Also, you can download a not insignificant number of primary sources through both sci-hub and various other sources like Project Gutenberg, IMSLP, etc. Add to that all the things that swell to prominence on Bittorrent trackers whenever the Streisand effect is in play. I have a hard time seeing that content disappearing any time soon. But I also have a hard time seeing how that content would have been aggregated in the first place if you had little micropayment barriers between all the things.
  - 9y ago

detaro 9y ago

One of the (arguably the) original hypertext visions (AFAIK Ted Nelson, its creator, coined the term "hypertext" in the early 60s). The main difference to web hypertext as we know it is that it doesn't just link documents together by what we know as "links", but it's bidirectional and with more content inclusion. E.g. if someone takes content from another document and puts it in their article, the text is connected to the source document (instead of them maybe putting a link to the source). And from the source document you can find all documents that used parts of it. Project Xanadu is Nelson's repeated attempts to make this work, while the world choose the WWW model and is happy enough with that (much to the disdain of Xanadu supporters, who believe the WWW to be a vastly inferior and broken approach). Once HTML took off, there wasn't really a chance to catch up, especially not with the limited resources they have. Lot's of bitterness about that, which also means they haven't really tried to make their ideas work with the web and rather aiming for an uncompromising, pure implementation of their ideas.

johnchristopher 9y ago

One of the best wired article ever and about that subject is https://www.wired.com/1995/06/xanadu/ https://www.wired.com/1995/06/xanadu/ and it should entertain you for an evening. Presume I am dead if I ever stop reposting this.

leoc 9y ago
If that's the case could you please post Nelson's responses http://web.archive.org/web/20001101230424/http://www2.educ.ksu.edu/Faculty/McGrathD/Fall99/NelsonLtr.htm http://web.archive.org/web/20001101230424/http://www2.educ.k... http://web.archive.org/web/20001003011753/http://xanadu.com.au/ararat http://web.archive.org/web/20001003011753/http://xanadu.com.... along with it.

yantrams 9y ago

Longform article on the history of Xanadu - https://www.wired.com/1995/06/xanadu/ https://www.wired.com/1995/06/xanadu/

shpx 9y ago
Discussed on hn https://news.ycombinator.com/item?id=795155 https://news.ycombinator.com/item?id=795155 18 comments Previous discussion of xanadu in general https://news.ycombinator.com/item?id=10642143 https://news.ycombinator.com/item?id=10642143 6 comments https://news.ycombinator.com/item?id=7849389 https://news.ycombinator.com/item?id=7849389 100 comments