4 ms·

A lot of these scripts that focus on rendering just wrap around mathjax/katex and md processors that do the actual heavy lifting. To me the real innovation need

A lot of these scripts that focus on rendering just wrap around mathjax/katex and md processors that do the actual heavy lifting. To me the real innovation needed to advance math authoring lies in processing cross references. Browser tech has yet to be fully utilized to present arguments in a manner that best facilitates reading instead of the linear fashion that comes from the constraint of printed documents. To be sure there are some work done in this direction in online open textbooks, but the tools and formats are too heavy for wide adoption.

utopcell 6y ago
Wondering what you have in mind ? Deep linking to other papers for example ?
- fspeech 6y ago
  For example like an IDE one should be able to easily see definitions defined somewhere else. Right now it is doable only if an author spends a lot of manual effort.
  - cxr 6y ago
    http://perma.pub/alpha/edward/openZRe7/fulfil http://perma.pub/alpha/edward/openZRe7/fulfil https://news.ycombinator.com/item?id=23491587 https://news.ycombinator.com/item?id=23491587 (Make sure you're viewing in a fullscreen browser window.)
  - mnky9800n 6y ago
    You mean like how Wikipedia renders the beginning of a page if you hover over a link in a wiki page?
    - fspeech 6y ago
      Sure floating tooltips or "show and hide" buttons are possible ways to let one preview cross reference within current context instead of leaving the page. This would reduce friction and memory load. This book https://mphitchman.com/geometry/ https://mphitchman.com/geometry/ extensively uses this and other (inline exercises, e.g.) through the pretext markup scheme and tools. As a first step it would be nice to have some of these features working in the more popular md schemes. But I also happen to think that it is possible to go beyond. Math could use a better semantic markup that enables better presentation of underlying structures in addition to pretty tex formulaes.
      - mnky9800n 6y ago
        I would like to see a math equation traceback. Like for a stochastic model of animal population distribution you might have something like p_ih = p_0exp[(ad_ih)^c + (b*s_star)^e] and it would be really nice if you could just see the traceback on everything, like you click on s_star and it takes you to the definition s_star = (s_max - s_h)/(s_max - s_min) and you can keep going until you are done. I guess I'm just tlaking about a wiki at this point but the focus would be on integrating the definitions into the equation itself so that when you read it in a textbook or whatever you can see where the ultimate comes from.

utopcell 6y ago

Wondering what you have in mind ? Deep linking to other papers for example ?

fspeech 6y ago
For example like an IDE one should be able to easily see definitions defined somewhere else. Right now it is doable only if an author spends a lot of manual effort.
- cxr 6y ago
  http://perma.pub/alpha/edward/openZRe7/fulfil http://perma.pub/alpha/edward/openZRe7/fulfil https://news.ycombinator.com/item?id=23491587 https://news.ycombinator.com/item?id=23491587 (Make sure you're viewing in a fullscreen browser window.)
- mnky9800n 6y ago
  You mean like how Wikipedia renders the beginning of a page if you hover over a link in a wiki page?
  - fspeech 6y ago
    Sure floating tooltips or "show and hide" buttons are possible ways to let one preview cross reference within current context instead of leaving the page. This would reduce friction and memory load. This book https://mphitchman.com/geometry/ https://mphitchman.com/geometry/ extensively uses this and other (inline exercises, e.g.) through the pretext markup scheme and tools. As a first step it would be nice to have some of these features working in the more popular md schemes. But I also happen to think that it is possible to go beyond. Math could use a better semantic markup that enables better presentation of underlying structures in addition to pretty tex formulaes.
    - mnky9800n 6y ago
      I would like to see a math equation traceback. Like for a stochastic model of animal population distribution you might have something like p_ih = p_0exp[(ad_ih)^c + (b*s_star)^e] and it would be really nice if you could just see the traceback on everything, like you click on s_star and it takes you to the definition s_star = (s_max - s_h)/(s_max - s_min) and you can keep going until you are done. I guess I'm just tlaking about a wiki at this point but the focus would be on integrating the definitions into the equation itself so that when you read it in a textbook or whatever you can see where the ultimate comes from.