7 ms·

As someone who has written a few of LaTeX documents over the years, I'd like to point out that rendering TeX math does not constitute a TeX document at all. Wh

As someone who has written a few of LaTeX documents over the years, I'd like to point out that rendering TeX math does not constitute a TeX document at all. What LaTeX truly shines at, is typography. Yet a bunch of people seem to think that it's only a way to write math. This is an interesting project. It just bums me out because I thought it was a TeX distribution for the web. I'd remove "TeX" from the name to avoid confusion.

nabla9 8y ago
It would be nice to have a option that uses LaTex to render html text in browsers. p.legibility { text-rendering: LaTeX; }
- majewsky 8y ago
  Rendering text is only a small part of typography. A significant part, sure, but not all of it by far. For your consideration, this is the manual for KOMA-Script: https://mirror.las.iastate.edu/tex-archive/macros/latex/contrib/koma-script/doc/scrguien.pdf https://mirror.las.iastate.edu/tex-archive/macros/latex/cont... That's 552 pages worth of typography that only considers page layouting. It doesn't deal with fonts or kerning at all. It's actually a fun read even if just for the author's occasional frustration with how often laymen confidently ignore centuries-old typography rules that exists for a reason.
  - M_Bakhtiari 8y ago
    Does it say anywhere in that manual why the right margin is supposed to be so narrow? I find it very uncomfortable to read texts with very narrow margins.
  - jacobush 8y ago
    I am now more tempted to and also more scared of LaTex than ever before.
  - nabla9 8y ago
    The context for the discussion is browser rendering for variable sized screens and user interaction in mind. Centuries-old typography rules can't be applied directly for screens. For paper documents you can already do Markup -> LatTeX -> pdf for good quality documents.
    - HerrMonnezza 8y ago
      > Centuries-old typography rules can't be applied directly for screens. Some of it, certainly not. A (larger IMHO) part of typography wisdom has really been distilled from experience making text easier and prettier to read over generations, and should not be thrown away lightheartedly because "screen is different from paper". Yes it is, but human eyes and brain are the same ;-)
      - nabla9 8y ago
        Human eyes and brains don't necessarily treat screen the same way they treat read paper.
  - enriquto 8y ago
    wow! the KOMA-script documentation must be the most unreadable LaTeX document I've ever seen - no margin on the binding side - silly colors everywhere - mixes normal and reverse text for no reason - uses german typographical conventions but the text is in english - extreme abuse of monospace italics - mixed serif/sans/monospace/boldface on every page - nonstandard paper size (18x21 cm) All of this could be expected, and not too worrying, for a text about any subject, except a fucking document about page layout!
    - blattimwind 8y ago
      The online and the book versions are different w.r.t. margins, colors etc. The online version highlights everything you can click and has tiny margins to avoid on-screen whitespace.
      - outadoc 8y ago
        Tiny margins... but only on one side. Because it still has weird version labels in the left margin, so it all looks asymetrical. Whyyy!
      - enriquto 8y ago
        > The online version highlights everything you can click and has tiny margins to avoid on-screen whitespace. This is ridiculously condescending... everybody is able to zoom-in with their pdf viewer.
        Eldt 8y ago
        What about this is condescending?
        etatoby 8y ago
        "Here, good boy, I know you can't be expected to know how to zoom your PDF file, so I'll just prepare a version for you without the margins and with some colors or something. Here, take a cookie. Good boy."
  - herpes 8y ago
    Is there supposed to be no margin on the right side, or is my phone being an ass?
- drb91 8y ago
  A) LaTeX is a library for laying out papers and common text memes. It’s not actually necessary at all. You would want to use TeX, which refers to the actual rendering engine. B) converting layout with css to Tex sounds like an absolute nightmare compared to just implementing proper word, line, paragraph, and page flowing/breaking in the browser. In fact, HTML is pretty opposed to page breaking at all, which is arguably a forte of TeX.
- lixtra 8y ago
  It’s opposite from what i feel. On one hand hand I suffer from seeing all the PDFs or papers with poor MS Office type set. But when it comes to the web I feel like people spend too much time (and waste my resources) to arrive at a perfect layout. There I would prefer a simple robust layout that renders well on all devices and resolutions.
- emarsden 8y ago
  Funnily enough, the Closure web browser, implemented in Common Lisp, had (it's no longer actively developed) a TeX-like paragraph formatting algorithm. The results were great. src/renderer/texpara.lisp in https://github.com/dym/closure https://github.com/dym/closure
minademian 8y ago
I agree. It is indeed a novel project, but has little to actually do with TeX.
- susam 8y ago
  I agree too. I write a lot of LaTeX documents and use the TeX Live distribution to typeset them, so I understand why the name TeXMe could cause some confusion. This project only cares about protecting the LaTeX content (the MathJax supported LaTeX, not the "real" LaTeX) by hiding it from the Markdown processor, so that the Markdown processor cannot mangle the LaTeX code before it is fed to MathJax. I could have named it MathMe, JaxMe, or something similar that would have eliminated this confusion between MathJax supported TeX/LaTeX[1] and the "real" TeX/LaTeX. However, unfortunately I did not spend sufficient time thinking about a good name for this project, thus the name TeXMe. [1]: http://docs.mathjax.org/en/latest/tex.html http://docs.mathjax.org/en/latest/tex.html
bfirsh 8y ago
If you’re interested in an actual TeX distribution for the web, I’m working on something like that: https://github.com/arxiv-vanity/engrafo https://github.com/arxiv-vanity/engrafo It’s based on LaTeXML, which does the heavy lifting of converting LaTeX to HTML. I’m more interested in the CSS applied on top of that. I want to make web output of the same calibre as the PDF output. (Or, near to it at least. :)
- danShumway 8y ago
  > I want to make web output of the same calibre as the PDF output. This to me is the unsolved problem. Interpreting a LaTeX file on the web is probably not a crazy problem -- heck, worst case scenario recompile the entire LaTeX compiler into WASM or something. But, typesetting on the web is really hard and browser capabilities that I know of are just not up to snuff, to the point that conventional wisdom is you pretty much never justify text on the web, ever, because it won't ever look good. So generic solutions to that problem are something I'd be interested in. The secondary problem past that is to make justified text look good while still being responsive :) But even just for fixed column widths it's something that I'd love to see. Props, it looks like a very cool project.
paraditedc 8y ago
> What LaTeX truly shines at, is typography. I respectfully disagree. I think people use LaTeX to write scientific documents because there is no better alternatives, not because LaTeX is good. I find the typesetting system in LaTeX is extremely verbose and over-complicated, as compared to something like HTML+CSS or Markdown. If someone comes up with a good enough typesetting system with some inspiration from Markdown, HTML and CSS, LaTeX will lose its popularity. HTML and CSS are more expressive than LaTeX, so it should be theoretically possible. One possible explanation for the lack of better alternatives is that the people who really need the alternatives and people who are capable of coming up with better alternatives are somewhat mutually exclusive. Former is more research-heavily and latter is more coding and user experience heavy.
- nabla9 8y ago
  pandoc -s markdown.txt -o latex-output.tex
- kderbyma 8y ago
  Latex is a fully Turing complete language as well as a typesetter. Only with JavaScript can you compare them.
  - danShumway 8y ago
    I mean, CSS/HTML is technically a Turing complete language given enough caveats[0]. ;) But I'm obviously just being a pedantic jerk when I say that. I get what you're getting at and you're not wrong. The real question to ask is, "is a programming language a good place to put a typesetter?" Making things more powerful doesn't always make them better, if that extra power forces you to introduce extra complexity. [0]: https://my-codeworks.com/blog/2015/css3-proven-to-be-turing-complete https://my-codeworks.com/blog/2015/css3-proven-to-be-turing-...
    - taeric 8y ago
      Why is that the real question to ask? The dogmatic separation of view and control is probably among the chief reasons of code bloat and bad decisions I've ever been close to.
wuliwong 8y ago
If a bunch of people use LaTeX only to write math and this contains that functionality, why wouldn't the author want to advertise it? Is there a specific short-hand name for the LaTeX math functionality?

nabla9 8y ago

It would be nice to have a option that uses LaTex to render html text in browsers. p.legibility { text-rendering: LaTeX; }

majewsky 8y ago
Rendering text is only a small part of typography. A significant part, sure, but not all of it by far. For your consideration, this is the manual for KOMA-Script: https://mirror.las.iastate.edu/tex-archive/macros/latex/contrib/koma-script/doc/scrguien.pdf https://mirror.las.iastate.edu/tex-archive/macros/latex/cont... That's 552 pages worth of typography that only considers page layouting. It doesn't deal with fonts or kerning at all. It's actually a fun read even if just for the author's occasional frustration with how often laymen confidently ignore centuries-old typography rules that exists for a reason.
- M_Bakhtiari 8y ago
  Does it say anywhere in that manual why the right margin is supposed to be so narrow? I find it very uncomfortable to read texts with very narrow margins.
- jacobush 8y ago
  I am now more tempted to and also more scared of LaTex than ever before.
- nabla9 8y ago
  The context for the discussion is browser rendering for variable sized screens and user interaction in mind. Centuries-old typography rules can't be applied directly for screens. For paper documents you can already do Markup -> LatTeX -> pdf for good quality documents.
  - HerrMonnezza 8y ago
    > Centuries-old typography rules can't be applied directly for screens. Some of it, certainly not. A (larger IMHO) part of typography wisdom has really been distilled from experience making text easier and prettier to read over generations, and should not be thrown away lightheartedly because "screen is different from paper". Yes it is, but human eyes and brain are the same ;-)
    - nabla9 8y ago
      Human eyes and brains don't necessarily treat screen the same way they treat read paper.
- enriquto 8y ago
  wow! the KOMA-script documentation must be the most unreadable LaTeX document I've ever seen - no margin on the binding side - silly colors everywhere - mixes normal and reverse text for no reason - uses german typographical conventions but the text is in english - extreme abuse of monospace italics - mixed serif/sans/monospace/boldface on every page - nonstandard paper size (18x21 cm) All of this could be expected, and not too worrying, for a text about any subject, except a fucking document about page layout!
  - blattimwind 8y ago
    The online and the book versions are different w.r.t. margins, colors etc. The online version highlights everything you can click and has tiny margins to avoid on-screen whitespace.
    - outadoc 8y ago
      Tiny margins... but only on one side. Because it still has weird version labels in the left margin, so it all looks asymetrical. Whyyy!
    - enriquto 8y ago
      > The online version highlights everything you can click and has tiny margins to avoid on-screen whitespace. This is ridiculously condescending... everybody is able to zoom-in with their pdf viewer.
      - Eldt 8y ago
        What about this is condescending?
        etatoby 8y ago
        "Here, good boy, I know you can't be expected to know how to zoom your PDF file, so I'll just prepare a version for you without the margins and with some colors or something. Here, take a cookie. Good boy."
- herpes 8y ago
  Is there supposed to be no margin on the right side, or is my phone being an ass?
drb91 8y ago
A) LaTeX is a library for laying out papers and common text memes. It’s not actually necessary at all. You would want to use TeX, which refers to the actual rendering engine. B) converting layout with css to Tex sounds like an absolute nightmare compared to just implementing proper word, line, paragraph, and page flowing/breaking in the browser. In fact, HTML is pretty opposed to page breaking at all, which is arguably a forte of TeX.
lixtra 8y ago
It’s opposite from what i feel. On one hand hand I suffer from seeing all the PDFs or papers with poor MS Office type set. But when it comes to the web I feel like people spend too much time (and waste my resources) to arrive at a perfect layout. There I would prefer a simple robust layout that renders well on all devices and resolutions.
emarsden 8y ago
Funnily enough, the Closure web browser, implemented in Common Lisp, had (it's no longer actively developed) a TeX-like paragraph formatting algorithm. The results were great. src/renderer/texpara.lisp in https://github.com/dym/closure https://github.com/dym/closure

minademian 8y ago

I agree. It is indeed a novel project, but has little to actually do with TeX.

susam 8y ago
I agree too. I write a lot of LaTeX documents and use the TeX Live distribution to typeset them, so I understand why the name TeXMe could cause some confusion. This project only cares about protecting the LaTeX content (the MathJax supported LaTeX, not the "real" LaTeX) by hiding it from the Markdown processor, so that the Markdown processor cannot mangle the LaTeX code before it is fed to MathJax. I could have named it MathMe, JaxMe, or something similar that would have eliminated this confusion between MathJax supported TeX/LaTeX[1] and the "real" TeX/LaTeX. However, unfortunately I did not spend sufficient time thinking about a good name for this project, thus the name TeXMe. [1]: http://docs.mathjax.org/en/latest/tex.html http://docs.mathjax.org/en/latest/tex.html

bfirsh 8y ago

If you’re interested in an actual TeX distribution for the web, I’m working on something like that: https://github.com/arxiv-vanity/engrafo https://github.com/arxiv-vanity/engrafo It’s based on LaTeXML, which does the heavy lifting of converting LaTeX to HTML. I’m more interested in the CSS applied on top of that. I want to make web output of the same calibre as the PDF output. (Or, near to it at least. :)

danShumway 8y ago
> I want to make web output of the same calibre as the PDF output. This to me is the unsolved problem. Interpreting a LaTeX file on the web is probably not a crazy problem -- heck, worst case scenario recompile the entire LaTeX compiler into WASM or something. But, typesetting on the web is really hard and browser capabilities that I know of are just not up to snuff, to the point that conventional wisdom is you pretty much never justify text on the web, ever, because it won't ever look good. So generic solutions to that problem are something I'd be interested in. The secondary problem past that is to make justified text look good while still being responsive :) But even just for fixed column widths it's something that I'd love to see. Props, it looks like a very cool project.

paraditedc 8y ago

> What LaTeX truly shines at, is typography. I respectfully disagree. I think people use LaTeX to write scientific documents because there is no better alternatives, not because LaTeX is good. I find the typesetting system in LaTeX is extremely verbose and over-complicated, as compared to something like HTML+CSS or Markdown. If someone comes up with a good enough typesetting system with some inspiration from Markdown, HTML and CSS, LaTeX will lose its popularity. HTML and CSS are more expressive than LaTeX, so it should be theoretically possible. One possible explanation for the lack of better alternatives is that the people who really need the alternatives and people who are capable of coming up with better alternatives are somewhat mutually exclusive. Former is more research-heavily and latter is more coding and user experience heavy.

nabla9 8y ago
pandoc -s markdown.txt -o latex-output.tex
kderbyma 8y ago
Latex is a fully Turing complete language as well as a typesetter. Only with JavaScript can you compare them.
- danShumway 8y ago
  I mean, CSS/HTML is technically a Turing complete language given enough caveats[0]. ;) But I'm obviously just being a pedantic jerk when I say that. I get what you're getting at and you're not wrong. The real question to ask is, "is a programming language a good place to put a typesetter?" Making things more powerful doesn't always make them better, if that extra power forces you to introduce extra complexity. [0]: https://my-codeworks.com/blog/2015/css3-proven-to-be-turing-complete https://my-codeworks.com/blog/2015/css3-proven-to-be-turing-...
  - taeric 8y ago
    Why is that the real question to ask? The dogmatic separation of view and control is probably among the chief reasons of code bloat and bad decisions I've ever been close to.

wuliwong 8y ago

If a bunch of people use LaTeX only to write math and this contains that functionality, why wouldn't the author want to advertise it? Is there a specific short-hand name for the LaTeX math functionality?