Hacker News Reader

Comment by AmazingTurtle | Hacker News Reader

Markdown, Quardown, Typst, this and that. This looks to LaTeX. All of this became so confusing and un-standardized - I went back to HTML+CSS.

View on HN

You know what, while you're at it, I might as well bring it up. How about XML?

I haven't really tried writing large pieces of text in it but I am already seriously considering. All other alternatives are too complicated and have a learning curve that gets in the way of writing itself. With XML, I'd be able to define my own tags and run them by a parser later on to auto-generate indexable footnotes, and create my own ways of structuring text besides the usual ones (chapters, sections, etcetera). Has anyone tried this approach?

Basically DocBook?

Exactly. Seems like we already invented that but it didn’t really catch fire.

Yeah last time I touched it was during an internship 15 years ago, the few memories I recall were not enjoyment at all (however I was using LaTeX everywhere at the time).

Writing documentation as XML is powerful but not enjoyable at all I guess

Professional tech writer here: We use GitHub and a tool called OxygenXML to write docs-as-code in an XML DTD called "DITA". It's a hefty IBM invention from the early aughts, but it covers every use case I've thrown at it, from small documentation sets to multi-thousand-page monsters. Supports PDF, HTML, Word, and many other output types.

DITA's so great at everything _except contributions from non-tech writers_ that half of my career has involved migrating tech writing stacks that use it to Markdown/SSGs in git repos.

DITA's benefits require a certain scale that most tech companies never achieve. And the Open Toolkit is a nightmare piece of software.

Agreed, and we use Markdown where we can. But inevitably some product manager comes along and demands tables inside tables or embedded reuse of content... and it's back to DITA.

OxygenXML makes the OT much more manageable. I haven't had to touch an OT XSL transform in a few years now. Worth every penny.

There's also a standard to turn XML into HTML via XSLT.

I have been thinking about this seriously myself. Not with a specific existing schema like DocBook, but with a custom schema (defined by me) that I then compile to standard schemas, like DocBook or HTML.

This seems extensible to the degree that I want (i.e. semantically rich enough that you can conceivably hang any application from it). But I just can't bring myself to write in XML syntax, especially for maths.

Well basic markdown is super useful to do stuff quickly.

Problem is when people started to build systems upon systems upon something that should not be used for more complicated cases.

Maybe not a problem because I don't care but I just see how loads of things that someone created to be "just that" someone takes without understanding and builds on top instead of understanding limitations and scope of the initial idea or system.

It is "oh it is missing a feature" - where it is "no it wasn't built to do that".

I have seen notepad in windows shipping some formatting features, I don't see that as an improvement. Notepad was notepad for purpose.

I think a good balance for a markdown syntax is something you can write nearly naturally on the fly (very low syntax needed for basic use) but can continue to mark up in more detail when you want to turn something into a complex use case (very high syntax ceiling for the complex cases). HTML+CSS almost gets there except the tags are a bit awkward for simple things like just <strong>bolding</strong> a word.

Regarding Notepad, keep in mind they dumped an entire separate app (WordPad) by merging the ability to format into Notepad. This kind of mirrors what I was talking about above: one app with a very low floor but a flexibly high ceiling when you want to go there with it rather than two separate apps you need to pick from up front.

There’s also the mature and reliable Org-mode, if you don’t mind Emacs as an editor.

Does org-mode have a spec? From what I've seen, org-mode is defined as whatever Emacs does. Without a spec I can't see how it's a viable format for widespread usage.

Org-mode has all three of:

1. Spec, written after the fact, not by a core dev.

2. Live help in 'org-info'.

3. Elisp string extraction via 'M-x describe-function' and the like.

Difficulties arise from subtle inconsistencies between the three, and inability of any one to comprehend the kitchen-tool-bin jumble of low- and high-level functionality that has been added over the years.

OTOH, content display is economical of eye and finger effort, and some of the Emacs interaction features startingly powerful.

I wrote my own, use-case specific converter to expand org-format to HTML. Runs on the CL, written in Go.

There is a Neovim "fork" of org mode that defined a spec for their format (which differs somewhat from org mode though)

Same. Replacing elements in with DOM in webdev is surprisingly fun with websockets too.

Having to know and learn 300 clunky frameworks, 97 different syntaxes it gets old.

HTML. CSS. Javascript.

Ask the AI to give me a markdown to html converter, good2go

> Ask the AI to give me a markdown to html converter, good2go

Why would you do that? There are dozens of libraries which do exactly that. Instead of AI you can ask npm or pip, heck you can even ask CRAN[1].

1: https://cran.r-project.org/web/packages/markdown/index.html

Exactly. It tells me the same things. It'll often give me the require() for javascript on packages I should use.

Personally, I don’t see the point. Searching for markdown libraries is not hard, a good library (the same that will show up first) will show you usage instructions at the top. You can copy-paste those and be done.

Also modern JavaScript uses import, require() is very 2015. If the AI is recommending you use require() I would suggest a better one.

If 2005 text editors autocomplete made it easy to balance and indent html/xml tags and syntax highlighting like today, would JSON, Yaml, Markdown have taken off? In The Art of Unix Programming from 2003, the author states editing xml by hand is torture, hence we must invent unique text formats and parsers for the same.

I have not been able to configure an editor that makes it less painful. I think it's not just the verbosity (although that is painful) - the prefix notation makes some things inherently hard to read. MathML is almost impossible.

Ah yes, writing my notes down in css and html. My favorite!

Seriously though, every time some new hotness comes along, you don’t really have to use it or even waste your time looking at it. Markdown will likely be here after all of its derivatives are long gone.

Gruber and Schwarz knew that in simplicity is a beauty.

Basic markdown format is good enough.

For specific cases, it depends what is the outcome - html, pdf…

But hell are binary formats that are used for page layout - vendor lock for centuries. Backward compatibility? Forget about it.

From this point of view people should stick to text based formats. Markdown, LaTex, Typst… you name it. Just use it :)

For me the only addition to markdown I usually enjoy is mathematical typesetting like on GitHub using the inline $$ syntax or the code blocks with the language set to `math`

Reply on news.ycombinator.com