I don't always use LaTeX, but when I do, I compile to HTML (2013) (opens in new tab)

Sphinx and reStructuredText are, IMHO, underrated power houses of document building. With extensions, you can hook them up to Zotero (or whatever)-managed bibtex files. You can render to beautiful HTML files, and you get latex PDFs and epubs for free. First class latex-math support, plenty of integrations with things like mermaid, graphviz, and the ability to build super-powerful custom directives to do basically anything. And way simpler/easier than pure LaTeX.

Heck you can even integrate a full-on requirements management system in them using sphinx-needs https://sphinx-needs.readthedocs.io/en/latest/

I feel ambivalent to LaTeX.

I don't like the language, the ecosystem is too big, complicated and breaks, but the end result is hard to do any other way.

This applies both the equations part, and the text reflow part (I think them as separate things, but they usually go together).

It should be possible to write text in HTML or markdown, and write the equations in latex or asciimath, and turn it into a beautiful/article style pdf, but sadly it is not.

Although CSS (colored and rounded boxes and such) + MathJax-SVG also can look nice.

So, I originally posted this last year. When I posted it, I was using tectonic as my LaTeX compiler, and since it didn't support HTML output yet, I didn't actually try the article's suggestion.

Today, when I saw that I got an invitation to repost this article from the mods, I thought I'd take the time to try it out.

The two commands that the article suggests can be combined into one:

    latexmlpost --dest=mydoc.html --format=html5 <(latexml mydoc.tex)

I did a comparison[1] of pdflatex and latexml using some old assignments, and it looks like compiling to HTML isn't fully there yet: the spacing was off in some places, and manual line breaks didn't work. But, I remain hopeful. If this gets polished, viewing LaTeX documents on phones would be much nicer.

[1]: https://imgur.com/a/yyyXWL8

For me, the main problem with most tools that render to HTML was that they don't support all math typesetting libraries that latex supports. I used to work with category theory, where it's common to use the tikz-cd library to typeset commutative diagrams. tikz-cd is based on tikz, which is usually not supported for HTML output.

But apart from math typesetting, my latex documents were usually very simple: They just used sections, paragraphs, some theorem environments and references to those, perhaps similar to what the stack project uses [3]. Simple latex such as this corresponds relatively directly to HTML (except for the math formulas of course). But many latex to html tools try to implement a full tex engine, which I believe means that they lower the high-level constructs to something more low level (or that's at least my understanding). This results in very complicated HTML documents from even simple latex input documents.

So what would've been needed for me was a tool that can (1) render all math that pdflatex can render, but that apart from math only needs to (2) support a very limited set of other latex features. In a hacky way, (1) can be accomplished by simply using pdflatex to render each formula of a latex document in isolation to a separate pdf, then converting this pdf to svg, and then incuding this svg in the output HTML in the appropriate position. And (2) is simply a matter of parsing this limited subset of latex. I've prototyped a tool like that here [1]. An example output can be found here [2].

Of course, SVGs are not exactly great for accessibility. But my understanding is that many blind mathematicians are very good at reading latex source code, so perhaps an SVG with alt text set to the latex source for that image is already pretty good.

[1] https://github.com/mbid/latex-to-html

[2] https://www.mbid.me/lcc-model/

[3] https://stacks.math.columbia.edu/

Tangently related, does anyone have experience with AsciiDoc? I've used reStructuredText before, but AsciiDoc is tempting, it looks cleaner.

The recommendation to use Markdown+MathJAX fall short when you want to write longer documents with numbered section, subsection, and theorem/definition/figure etc tracking and referencing.

I'm sure with Sphinx and reStructuredText you can get that large-scale document tracking stuff, but with LaTeX it just works for the most part and you don't need to juggle a bunch of different side-projects and extensions. Plus you get things like automatic index generation (for a physical book).

This is from 2013, so the bet that "nobody will want to read [PDFs] in 5 years" can be considered failed. If anything, PDF has become the lingua franca of the academic web, crowding out even DjVU at the thing that DjVU was made for and PDF was not.

I have not been following the development of mathjax, pandoc, etc. carefully, so I'm wondering: Have the main issues been solved? By these I mean

(1) support for most popular packages,

(2) automatically breaking long outputs into small pages that don't overheat my laptop or crash my browser and yet reference each other properly,

(3) printability (without lines broken in half, senseless overflows and the likes) or cross-compilability with a regular PDF compiler?

I know the ar5iv project is getting closer and closer to (1) and (3), but is that available to regular users?

I write markdown, use pandoc to make LaTeX and from that a PDF for a printed thing and just supply markdown for non-printed stuff.

I spent a few weeks last year doing the opposite, HTML to LaTex in order to print and nicely typeset top HN articles, so I'd have a nicely printed booklet each morning. I think creating hard copies of web content for offline reading holds a lot of promise, but the internet is a beast.

Another LaTeX-to-HTML tool is lwarp (https://github.com/bdtc/lwarp) which starts from the idea that there only exists one program that can parse LaTeX: the LaTeX compiler itself. Implementing a new parser is almost futile. So instead, the lwarp package redefines all the macros to output HTML. Something like \renewcommand[1]{\textbf}{<b>#1</b>} This way, compiling LaTeX gives you a PDF whose text is HTML code, so now you can extract the plain text from it and you have an HTML file. The advantage is that it can easily deal with custom macros etc., because these are natively resolved by the LaTeX compiler.

I use lwarp to make https://tikz.dev/, an HTML version of the TikZ manual, which is probably one of the most complicated LaTeX documents in existence.

This article really doesn't get what LaTeX does. Of course it is overkill to have 5 lines of text rendered with LaTeX into a PDF. But the point of LaTeX is exactly to set the typesetting of an output document in stone. PDF is meant to do that and HTML cannot do that. A PDF conserves everything and that is precisely the point to have a set layout for printing or displaying on different devices.

Yes, there should be easy ways to display math on the web. No, this doesn't mean that LaTeX is obsolete.

Besides, what about references, both external and internal? Probably needs more "modern" tooling.

Note that one can convert PDF to HTML using tools like:

https://pdf2htmlex.github.io/pdf2htmlEX/

Example of a paper with equations:

https://pdf2htmlex.github.io/pdf2htmlEX/demo/demo.html

If I'm using LaTeX, I'm writing scientific articles. I expect scientific articles to be read by people on computers with normal screen sizes or printed off. Therefore there's no reason to bother with anything other than PDF. PDF works great.

I honestly don’t see the point of using LaTeX if you’re generating HTML. The great strength of LaTeX, in my view, is the precise control it provides over typography and formatting. As such, it works best with an output format which can faithfully render these documents — such as PDF. For an output format like HTML, which encourages reflowability over faithful rendering, I’d much prefer to use an ‘easier’ document format like Markdown or reStructuredText.

When I use LaTeX, it's because I want a way to store book manuscripts and their layout as code in version control. I never use any of the math layout. I get the impression that my use case is rather in the minority.

I would use CSS+HTML for layout, but what do I do about automatically generating tables of contents and indexes?

I guess I could write my own tool for that. Hmm.

Correct me if I’m wrong, but there isn’t a way to do a compile that incorporates Biblatex.

Instead of MathJax, maybe also consider KaTeX: https://katex.org/

It's faster than MathJax and also can be pre-rendered on the server (or in your SSG!).

I like LaTeX for the quality of it's pdf output, I use in for docs that need to be "printed" (non necessarily on paper, but still 'fixed typographical form for potentially long term archiving) not for anything else and yes I DO HATE pdfs because of their design, but PostScript is not much common these days and while a bit better for certain aspect is not much better in general, dvi is even worse.

For my notes, for anything that need to be "live" I use org-mode because:

- it's a far more natural markup than anything else

- it's rendered INLINE, no need to jump between a source form and a rendered one, a thing MD lovers fails to understand

- it's an outlining tool, another thing most other tools fails miserably to understand

- it easily incorporate live things in other languages (org-babel) a thing no modern REPL-alike DocUI like Jupyter can't do

Long story short I prefer the best tool depending on the job. HTML might be the least common denominator tool, making it the worst in essentially all cases. XML for machine usage, SGML in general, are good for machine usage, but they are very impractical in current usage, just see the actual crappy state of things for e-invoicing with XML/XADES docs + XSL to render them in the end as pdf for the human. They are a good too in some case, but again not the best for any specific case.

> don't just produce PDFs that nobody can read on small screens

I was thinking about this recently. If you get pedantic enough* about it, the typesetting quality you can get from a LaTeX+PDF is strictly better than what can be achieved using (sane) HTML.

I wanted to blog in LaTeX, and to solve the screen-size issue I thought I'd pre-bake to a wide range of page geometries, and then serve up an appropriate one to the client using pdf.js.

Fortunately for everyone, I decided against it in the end and continued blogging in markdown+html (with mathml support)

*well beyond what most readers would possibly care about

LatexML has come a long way. Even arXiv uses LatexML internally to offer HTML5 versions as of late 2023. It does have limitations in not supporting all packages, or producing a high-quality translation in all cases.

If you don't need to convert entire LaTeX documents, MathJaX and KaTeX are really good at rendering a subset of LaTeX as MathML/SVG. I run MathJaX + an xypic extension for commutative diagrams with server-side rendering on my website, and it works great in practice.

A lot has happened since 2013. Have a look at https://quarto.org/, if you plan to publish in HTML. Quarto has already support for Typst: https://quarto.org/docs/output-formats/typst.html

LaTeML [1] is presumably the latex to html tool that arXiv is testing right now. What are peoples thoughts about it compared to other such tools?

[1] https://github.com/brucemiller/LaTeXML

One solution is to embed alternatives within PDF itself. LibreOffice can embed inside a PDF the original editabble source in ODF format. You could also embed ePub. That would mean you would have a single file that could be processed in many useful ways.

Although I use markdown (and similar) for memos, I turn to latex for longer and more complex material.

A lot of this is just because latex has been a standard for publishers in my field since I started (approximately a thousand years ago).

When writing for journals, latex saves a lot of work. Publishers provide latex templates that ensure that articles have a prescribed format and scope of content. Being able to see a good facsimile of the final published form is quite handy for authors. Oh, this paragraph is going on for over a column -- I'll break it up. That sort of thing.

This still applies when writing for longer things, such as textbooks and course notes, but another factor (for me, the larger one) is that latex (more properly, the tex upon which latex sits) is a programming language. Macros can be written to do lots of things that would be a pain if done manually, and once a macro is written, altering an entire text is easy. I did this in a book I wrote a while back, writing macros to colourize text that would be indexed, add margin notes for things I wanted to return to, categorize paragraphs by function, and so on. I could turn all those macros on and off by uncommenting a line. This is really quite helpful in writing something that takes months to years to complete. Frankly, I use this macro approach even in memos written in markdown. Inside almost all of my markdown documents, there are latex commands.

As for reading things on a small screen, which I guess is really the topic here, I must admit that this is something I rarely do within my own field. Sure, I do it if reading one of those 10-km overview articles in Science or Nature. But when it comes to my own field, things are technical and demand long periods of study. I don't try to read this stuff on the bus or in a coffee queue. I need time (hours or days) and I need to be able to take notes.

Another reason I prefer PDF is that it is fixed. My brain puts information into a sort of spatial framework. Somehow, if I look at a paper I first read 40 years ago, I still know what information is on which page, which of the diagrams summarizes the whole thing, and which of the citations is key. This may be a flaw in my brain functioning, but I just don't find these sorts of memories forming when I read content that has a plastic format, with paragraph breaks changing if I adjust my view. But maybe this is just my age talking, I suppose.

I learned LaTeX in grad school in 2013, starting with LyX. Yesterday, I compiled an Rmarkdown document into an APA6-conformant PDF with just a bit of YAML, with a tex file as an intermediate output.

We're almost there for skipping LaTeX entirely, but in my experience, Google Docs and Overleaf still offer vastly superior collaborating tools. Now if we could just edit {.md; .rmd; .ipynb} files directly on Overleaf, with comments and track changes, we'd be in business...

I love the author’s “if you want to leave a comment email me”. I saw this somewhere else and it motivated me to make an automated system that works like that: https://r3ply.com

I just moved "up" from gfm markdown to asxiidoc and oh do I miss LaTeX.

html rendering of LaTeX is a godsend. and imnsho asciidoc a work around to not fully having that.

At work all reports are html. If you want pdf, cmd-P

Would be nice if this article included some equations!

LyX is the way to LaTeX

I dont always use latex but when I do I always hate it.