What's wrong with HTML that has classnames, or the new data attributes that HTML5 has? They're not dependent on any one implementation and HTML seems like it's the most future-proof thing out there apart from plain text.
In order:
- <div class='expando'>Hide/Show this text.</div>
- <a href="#something">Go somewhere</a>
- <a href="otherdoc.html">Hyperlink</a>
- <video src="hello.mp4" />
- <code>console.log('hello world!');</code>
Almost any JS developer can make these work without any library, from what the browser provides, alone. Give the innerHTML to eval, for example. It seems like the most future-proof way to go, and I'll be happy to be proven wrong and find out about something I don't know about. But Markdown nor any of its derivatives don't seem to be good enough.
Of course, you could also write a Markdown parser that has constructs that will compile to the above too.
> This is about the inadequacy of PDF as a document format for the future.
And I'm guessing, since we're doing this over an HTML document, that that'll be sufficient instead.