It was always interesting to me growing up in the days before LaTeX how, although #math and computer science were clearly joined at the hip— it was only my math homework that was impossible to type. And not just due to missing symbols— there were whole missing models of expression. This has changed in many ways— but the person who can type calculus notes in real time remains rare. Probably, handwriting recognition will fill this need before fast and natural #LaTeX typing becomes common.
For a time typesetting math was about presentation— increasingly math documents are a kind of encoding. To write the equation is to establish that relation with variables. I’m excited to see interactivity increasingly a part of our documents. This is a proof, but it’s also a demonstration— and you can mess with it!
Still the world of word processing, social media, and presentations lag behind— some of the most exciting new presentation software fails to do anything useful with equations at all!
@futurebird The fact that LaTeX is more about presentation than meaning is a barrier to making mathematics more accessible to the visually impaired. There's Content MathML but in practice it's so annoying that no one uses it. It would be great to have something with the (relative) simplicity of LaTeX but which can be converted to sensible audio as well a visual presentation.
@futurebird It is not inconceivable to me that "traditional math notation" might be replaced, at least in some sections of the population, by a more computer-style notation. When I want to text my kid (studying math three time zones away) some snippet of math, I will send "sqrt(x)", or at the extreme "\sqrt(x)", not "√𝑥".
When kids grow up knee deep in Python (or whatever will be fashionable in the future) why force them to learn how to re-re-represent the same thought in some weird, hard-to-type fashion just because mathematicians used that a hundred years ago? Especially when the "code style" math is completely unambiguous, while the "traditional math notation" is full of self-contradictions and ambiguities?
@futurebird I always thought it was strange that even though the World Wide Web came out of CERN, HTML completely punted on rendering mathematics.
MathML was an attempt to deal with that--and it was also forked into a semantic language and a presentational language. But because it was an XML application, it was very complex and not designed to be human-writeable to even the degree that LaTeX or Mathematica is. The understanding was that you'd have some kind of editor generating it. General-purpose browser software never supported it. And I don't get the sense that it's really caught on--web pages that do support math markup are more likely to use some TeX-based presentational thing like MathJax.
@futurebird (The separation into Presentational MathML and Content MathML bothered me too-typical W3C behavior, probably they couldn't decide which they wanted it to be so they just cooked up two whole separate MathMLs.)
@futurebird ...Which is not to say I can't see the justification. Traditional mathematical notation intended for consumption by humans is actually quite structurally loose and ambiguous, but people have strong preferences about it, as you can tell by jumping into any Facebook argument about PEMDAS.
Whereas a content encoding for mathematics needs to be completely unambiguous, like a computer language, but is going to struggle to encompass all of mathematics, in the way that you can by inventing offhand a new presentational notation for your paper--and to present that, you probably want to just have some entirely typographical freedom.
Add comment