4 ms·

If you want a general term for this sort of thing: https://en.wikipedia.org/wiki/StretchText https://en.wikipedia.org/wiki/StretchText. (The original note descr

If you want a general term for this sort of thing: https://en.wikipedia.org/wiki/StretchText https://en.wikipedia.org/wiki/StretchText. (The original note describing the concept: http://i.imgur.com/jOCOQGI.png http://i.imgur.com/jOCOQGI.png)

chime 10y ago
Here's a long-term AI/ML goal - interpolate between Simple Wikipedia and regular Wikipedia article on the same topic and use StretchText to allow the reader to decide their level of depth.
- themodelplumber 10y ago
  Oh how I wish every math article on Wikipedia had this feature.
- ge96 10y ago
  I want this for learning faster, summarizing a page, but it's hard to teach something (to me). Like have it pull all the text in plain text, you read the page and pull the information that is new/important to you. Which this is subjective/different per person, but if the "AI" is unique to you then I think that would make sense. But for this to work, you'd have to have a history/scope of your current knowledge.
WhitneyLand 10y ago
That image is an absolutely fascinating historical artifact. He's describing (in a related paper) the essence of web sites in detail in 1965. It's difficult to even imagine how few people were thinking of such things when they were so far removed everyday life and culture. Tim Berners-Lee I thought was much more well known for starting a revolution with the world wide web. I hate to say but the web seems a bit less of an insight when you know that Hypertext had already been so well established for almost 25 years. https://wikipedia.org/wiki/Ted_Nelson https://wikipedia.org/wiki/Ted_Nelson
- dboreham 10y ago
  I'm surprised that anyone could believe it required insight. Having lived through that period, nothing about it seemed insightful or surprising at the time. The only time I remember being surprised was when I saw a UPS truck with their URL painted on the side, in 1993. I remember thinking it was a couple of years ahead of time.
  - bernardlunn 10y ago
    I recall somebody showing me the Internet in 1992 and I dismissed him as nuts - oops!
- 0x0 10y ago
  Interesting wikipedia article, immediately brings to mind "worse is better" or "perfect is the enemy of good"
- smacktoward 10y ago
  He's actually describing the essence of something much more ambitious than the World Wide Web, if you can believe that. Ted Nelson's idea was/is called "Xanadu": https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu The Xanadu vision is centered on a hypertext system, like the Web. But unlike the Web, it also includes such things as: - A mechanism for embedding live documents within other documents ("transclusion": https://en.wikipedia.org/wiki/Transclusion https://en.wikipedia.org/wiki/Transclusion) - Bidirectional links, that updated themselves automatically when the address of the linked document changed (see http://dubinko.info/blog/2009/11/22/how-xanadu-works/ http://dubinko.info/blog/2009/11/22/how-xanadu-works/) - An addressing system that goes beyond just documents to let you link directly to any range of content you wish ("tumblers": https://en.wikipedia.org/wiki/Tumbler_(Project_Xanadu) https://en.wikipedia.org/wiki/Tumbler_(Project_Xanadu)) - Automatic, redundant cross-network storage of all documents (a built-in Wayback Machine) - A micropayments system for content creators to sell access to what they create - Strong identity built in (a necessity if you want to also have payments built in) ... and much more. Xanadu is (in theory anyway) a much more sophisticated system than the Web is even today, never mind the Web as it was when TBL first designed it. Back then it was so much more crude than Xanadu as to barely merit comparison. But in a way that was TBL's genius, because, while the Web was much more crude than Xanadu architecturally, it was something that could actually be implemented with the early-'90s technology he had on hand. Xanadu, a much grander vision, was so much grander that it has defied fifty years' worth of efforts to implement it. (Back when Wired was worth reading, they ran a good piece on the state of those efforts circa 1995: https://www.wired.com/1995/06/xanadu/ https://www.wired.com/1995/06/xanadu/ It's still defying implementation today, more than twenty years later.) So we ended up with the Web we have, which kinda sorta works except for all the ways that it really doesn't. (Paying content creators, for instance.) Ted Nelson saw all those problems coming, all the way back in 1960, and he tried to come up with a system that would head them all off. He hasn't succeeded (not yet, anyway!), but you have to admire both the vision and the attempt.
StavrosK 10y ago
I worked on a related implementation of this: https://skorokithakis.github.io/expounder/ https://skorokithakis.github.io/expounder/
- ryanmarsh 10y ago
  Wow this is really great. Thanks for sharing.
  - StavrosK 10y ago
    Thank you, I'm glad you like it!
- danieltillett 10y ago
  Great idea. My only suggestion is to have SEQme sort of symbol attached to the link that indicates that it expands the text and does not link to other material.
  - StavrosK 10y ago
    Yes, that's indicated by the dashed underline (although you can style it differently, including with an icon, if you're already using that style).
- dyates 10y ago
  This sort of expanding text is used heavily in hypertext-based interactive fiction,[1][2] but I haven't seen it used outside of that before. Nice work! [1]: https://twinery.org/ https://twinery.org/ [2]: https://texturewriter.com/ https://texturewriter.com/
aheilbut 10y ago
I've always wanted a version of this for math.
acjohnson55 10y ago
I hadn't heard of that phase, but that's great touch point. It reminds me of http://tomasp.net/coeffects/ http://tomasp.net/coeffects/.
bernardlunn 10y ago
I like the idea of stretch text to created distributed bodies of work. Off topic but why send text to Imgur?
fractallyte 10y ago
'Telescopic text' featured a few years ago on HN: https://news.ycombinator.com/item?id=2551120 https://news.ycombinator.com/item?id=2551120 A nice example: http://www.telescopictext.com/ http://www.telescopictext.com/
greendestiny_re 10y ago
Why isn't Wikipedia made like this?

chime 10y ago

Here's a long-term AI/ML goal - interpolate between Simple Wikipedia and regular Wikipedia article on the same topic and use StretchText to allow the reader to decide their level of depth.

themodelplumber 10y ago
Oh how I wish every math article on Wikipedia had this feature.
ge96 10y ago
I want this for learning faster, summarizing a page, but it's hard to teach something (to me). Like have it pull all the text in plain text, you read the page and pull the information that is new/important to you. Which this is subjective/different per person, but if the "AI" is unique to you then I think that would make sense. But for this to work, you'd have to have a history/scope of your current knowledge.

WhitneyLand 10y ago

That image is an absolutely fascinating historical artifact. He's describing (in a related paper) the essence of web sites in detail in 1965. It's difficult to even imagine how few people were thinking of such things when they were so far removed everyday life and culture. Tim Berners-Lee I thought was much more well known for starting a revolution with the world wide web. I hate to say but the web seems a bit less of an insight when you know that Hypertext had already been so well established for almost 25 years. https://wikipedia.org/wiki/Ted_Nelson https://wikipedia.org/wiki/Ted_Nelson

dboreham 10y ago
I'm surprised that anyone could believe it required insight. Having lived through that period, nothing about it seemed insightful or surprising at the time. The only time I remember being surprised was when I saw a UPS truck with their URL painted on the side, in 1993. I remember thinking it was a couple of years ahead of time.
- bernardlunn 10y ago
  I recall somebody showing me the Internet in 1992 and I dismissed him as nuts - oops!
0x0 10y ago
Interesting wikipedia article, immediately brings to mind "worse is better" or "perfect is the enemy of good"
smacktoward 10y ago
He's actually describing the essence of something much more ambitious than the World Wide Web, if you can believe that. Ted Nelson's idea was/is called "Xanadu": https://en.wikipedia.org/wiki/Project_Xanadu https://en.wikipedia.org/wiki/Project_Xanadu The Xanadu vision is centered on a hypertext system, like the Web. But unlike the Web, it also includes such things as: - A mechanism for embedding live documents within other documents ("transclusion": https://en.wikipedia.org/wiki/Transclusion https://en.wikipedia.org/wiki/Transclusion) - Bidirectional links, that updated themselves automatically when the address of the linked document changed (see http://dubinko.info/blog/2009/11/22/how-xanadu-works/ http://dubinko.info/blog/2009/11/22/how-xanadu-works/) - An addressing system that goes beyond just documents to let you link directly to any range of content you wish ("tumblers": https://en.wikipedia.org/wiki/Tumbler_(Project_Xanadu) https://en.wikipedia.org/wiki/Tumbler_(Project_Xanadu)) - Automatic, redundant cross-network storage of all documents (a built-in Wayback Machine) - A micropayments system for content creators to sell access to what they create - Strong identity built in (a necessity if you want to also have payments built in) ... and much more. Xanadu is (in theory anyway) a much more sophisticated system than the Web is even today, never mind the Web as it was when TBL first designed it. Back then it was so much more crude than Xanadu as to barely merit comparison. But in a way that was TBL's genius, because, while the Web was much more crude than Xanadu architecturally, it was something that could actually be implemented with the early-'90s technology he had on hand. Xanadu, a much grander vision, was so much grander that it has defied fifty years' worth of efforts to implement it. (Back when Wired was worth reading, they ran a good piece on the state of those efforts circa 1995: https://www.wired.com/1995/06/xanadu/ https://www.wired.com/1995/06/xanadu/ It's still defying implementation today, more than twenty years later.) So we ended up with the Web we have, which kinda sorta works except for all the ways that it really doesn't. (Paying content creators, for instance.) Ted Nelson saw all those problems coming, all the way back in 1960, and he tried to come up with a system that would head them all off. He hasn't succeeded (not yet, anyway!), but you have to admire both the vision and the attempt.

StavrosK 10y ago

I worked on a related implementation of this: https://skorokithakis.github.io/expounder/ https://skorokithakis.github.io/expounder/

ryanmarsh 10y ago
Wow this is really great. Thanks for sharing.
- StavrosK 10y ago
  Thank you, I'm glad you like it!
danieltillett 10y ago
Great idea. My only suggestion is to have SEQme sort of symbol attached to the link that indicates that it expands the text and does not link to other material.
- StavrosK 10y ago
  Yes, that's indicated by the dashed underline (although you can style it differently, including with an icon, if you're already using that style).
dyates 10y ago
This sort of expanding text is used heavily in hypertext-based interactive fiction,[1][2] but I haven't seen it used outside of that before. Nice work! [1]: https://twinery.org/ https://twinery.org/ [2]: https://texturewriter.com/ https://texturewriter.com/

aheilbut 10y ago

I've always wanted a version of this for math.

acjohnson55 10y ago

I hadn't heard of that phase, but that's great touch point. It reminds me of http://tomasp.net/coeffects/ http://tomasp.net/coeffects/.

bernardlunn 10y ago

I like the idea of stretch text to created distributed bodies of work. Off topic but why send text to Imgur?

fractallyte 10y ago

'Telescopic text' featured a few years ago on HN: https://news.ycombinator.com/item?id=2551120 https://news.ycombinator.com/item?id=2551120 A nice example: http://www.telescopictext.com/ http://www.telescopictext.com/

greendestiny_re 10y ago

Why isn't Wikipedia made like this?