Markdown is the defacto language of the llms. Between this and the python library from Microsoft that converts all of their formats to markdown I think that markdown officially has won the markup language war.
You can make words *bold*, /italic/, _underlined_, =code= and ~verbatim~, and, if you must, +strike-through+.
The official manual states equals is used for =verbatim= quotes: https://orgmode.org/manual/Emphasis-and-Monospace.html
1. Collapsible headlines and headline search.
2. Executable source code blocks (for notebook style work).
3. TODO states.
4. Time tracking and clock tables.
5. Table formulas.
6. Inline LaTeX and document generation in general.
Now 1 is just an editor feature, and some of these others could be, too. But I wish Markdown was more powerful, extensible, or less ubiquitous.
What bothers me most about Org Mode is that support is pretty limited outside Emacs. We use it as a wiki replacement at my company, for that integrations into other editors are kinda good enough. But there's some areas like reporting only Emacs users can realistically work on. GitLab (and Forgejo, which we recently switched to) render Org pretty nicely, so it's easy to consume in a browser. But editing is a different story.
So I guess I wish for either a less complex (and thus easier to support) Org, or a more powerful Markdown.
This of course ends up being editor specific, but if org has the same limitation, by being tied to emacs, is it really any different? I think what emacs has going for it in this case is that it’s been around for decades, and we can assume it will continue to be around for decades into the future. The markdown editor de jour may not be.
Either Markdown++ or Org Lite. I'll take either :) I'm not a fan of coupling formats and tools tightly.
I also like [Markdown's text first](https://hyperlink.com) then link format much more than [[https://hyperlink.com][Orgmode's link first then text]].
But for the life of me, there are so many things I hate about Markdown:
- Alternative syntax for _italic_ *italic* and __bold__ **bold** (why?)
- versus using /italic/ and *bold*. It just feels so right
- You need to use HTML tags <u>to underline</u>, or <s>strikethrough</s>, or just about anything in Markdown (including line breaks! You need to use the <br> tag)
- +strikethrough+ is alright, but _underline_ just makes sense
- Lack of a unified Markdown standard:
- Diminishes portability (e.g. varying approaches for file meta tags like in Obsidian)
- Causes different renderings of the same document (e.g. Obsidian vs GitHub)
- There was a serious standardization effort in 'Standard Markdown', only for the original creator to be a knob about it: https://blog.codinghorror.com/standard-markdown-is-now-common-markdown/
Markdown isn't going away anytime soon.
A ten minute web search would have revealed it was already a solved problem, instead of splintering and inventing a new 'standard'. But that was somehow too much effort.
https://raw.githubusercontent.com/git/git/refs/heads/master/...
When the original is in markdown, it's somehow always noticeable.
It's the new classic example of "worse is better".
Other markups may be more well thought out and more complete, but Markdown beats them hands down on usability.
Consider the comparison between MD and RST, for example...
Thank you for pointing me to the MarkItDown library [1]. I had no idea such a thing existed.
Oh wait, maybe it's 30% of it is repeated...
That keyboard was such a POS that a Wall Street Journal writer did a feature story on it without correcting any of the mistakes the keyboard introduced. It was essentially illegible. Then the Web site had buttons that you could push to remove each kind of error and make it legible: https://www.wsj.com/graphics/apple-still-hasnt-fixed-its-mac...
And if you doubt me on that last point, look at things like callouts in Obsidian, clearly people want to be able to do more but we keep tacking weird formatting extensions onto Markdown.
In some contexts, JSX and markdown are also competing.
I think HTML is the true winner.
Which opens the question, is there a real advantage in using Markdown with AI instead of HTML? Or did nobody ever tried this?
You find readme documents in Markdown in every open-source repo, but why? What are you supposed to read them with?
After much searching I finally ended up with Marked, on the Mac. A paid app. But that's about it for options. And no, I'm not talking about text editors that offer a "preview" pane that you can optionally invoke. I'm talking about a simple reader for Markdown that renders it, so you're not reading a shitty text file with a bunch of formatting codes in it. Why bother putting them in, if nothing parses and acts on them?
Browsers read html. You're supposed to read markdown with a browser by rendering it to html and reading the html.
Or, from the other angle, with Markdown being found in so many places, why aren't there popular viewers for it?
The preview is really only needed if there is advanced formatting or HTML in there. A basic markdown doc should be easy to read in plain text.
And I don't want an extra "preview" pane. That was the specific complaint. Nor do I want these docs in my IDE at all. I want all the available space in my IDE for code.
Again, why have this format all over the place when there are vanishingly few VIEWERS for it? Or vice versa?
If you're on macOS though, the best option for rendered Markdown is a quicklook plugin. There are a few around - I like [1] though.
What is your alternative to Markdown here?
If everyone's just using a plain-text reader to view these files, then why clutter them up with formatting codes?
It's not "clutter" it's meaningful information for the human eye in plain text.
Which is why I'm asking: what's your alternative? Is it no headers, no lists, no standard * or \_ for emphasis, as it has been done long before Markdown existed?
My "alternative" is to have DEDICATED, lightweight Markdown viewers. The major OSes have long come with a simple text-file viewer that can render RTF. Why not Markdown?
In the meantime, I'm mystified as to why the format is so rampant when there is so little support for it.
Maybe it doesn't to you, but it does to most people.
An update to Microsoft Notepad which renders Markdown is currently being rolled out: https://blogs.windows.com/windows-insider/2025/05/30/text-fo...
I'm not sure I agree *something* is intuitive for italic and **something else** is intuitive for bold, or that this is intuitively a block quote:
> #### The quarterly results look great!
>
> - Revenue was off the chart.
> - Profits were higher than ever.
> > Everything is going according to *plan*.
As for - and >, they've been in use in plain text email clients for bulleted lists and quotations for 35 years now [1].
[1]: https://en.wikipedia.org/wiki/Posting_style#:~:text=The%20co...
This is information people often want in a plain text file, and having a common way to signal those things is beneficial when working on teams or sharing with others.
What’s the alternative? Everyone makes up their own bespoke way of formatting their text file that works for them, then has a key to explain it in their doc?
Every source control website I've used will render Markdown as HTML for you.
Ah, so that's the definition of a "loaded question." Thanks for the fallacy!
"Loaded Question: A loaded question contains an assumption that may or may not be true or agreed upon. It's designed to trap the respondent into confirming a premise they might not agree with, regardless of their answer.
Complex Question Fallacy (or plurium interrogationum): This is a logical fallacy where the question presumes the truth of something that may be false or unproven. Answering it directly implies agreement with that presupposition."
The IDE that I use most of the time also renders markdown.
And because of the lack of such readers, everyone just reads the plain text with a bunch of garbage in it. It's just gallingly dumb.
That's what documentation generators are for, to render all the docs and search them if you're just a user.
Your pretension that most programming involves writing Markdown "source code" is absurd. I neither want nor need Markdown in any of my source-code editors, ever. Not once in decades of professional programming have I missed it.
And if I do need to create Markdown, I will want to do so in a simple WYSIWIG editor and save it as MD. Again... not in my IDE.
And how often are you even reading READMEs? For me it's usually once, when I first use a library, because it typically gives an outline of the project and/or instructions on how to install it. So I'm reading it in a browser.
Meanwhile I'm updating my own READMEs often, usually using them as a sort of to-do list and then outlining existing functionality.
Like, are you really using so many libraries as glue that reading READMEs is such a common task that you specifically want it in your IDE? Is that the current state of programming?
Are you being intentionally obtuse? I said that's exactly what I DON'T want.
.gitignore or .ignore file...
text on the left, render on the right pane
example: https://imgur.com/9rjoMa2.png
It's even dumber when you ponder: What is this a "preview" of? How the document will look in... a nonexistent viewer?