Project idea: Structured prose editor

To do: Turn this into less of an unstructured brain dump.

A prose editor built around highly configurable semantic markup. Think Scrivener for DMark.

Relevant notes:

Technological choices:

Thoughts copied from Mas­to­don:

I find Markdown (and other lightweight markup languages) to be rather heavy on markup. I would like the markup to disappear. Bear does this very nicely — see screenshot:


There are also specific problems with Markdown that make it difficult to use in general:

  • Markdown (and especially its dialects) is overly complex for computers to parse. It is hard to get it right.

  • Markdown is quite limited in what it supports. It lacks markup concepts that I find useful or necessary when doing technical writing — concepts that DocBook (even Simplified Docbook) has.


Also, I specifically DON’T want a WYSIWYG experience. I think that goes against the idea of a semantic markup language.

But I do think an editor that makes the details of the markup go out of the way would be very helpful. Here is my setup where different inline element types have distinct styles (but not WYSIWYG):


There is also the issue that Markdown in general is too coupled to HTML. Historically this makes sense (it was created to write HTML more easily), but it is consistently awkward when generating non-HTML output.


Emphasis does indeed not stand out in Bear, but in my Scrivener setup, the style for emphasis is distinctly purple.

But the purple is purely for authoring purposes in the editor, and the output does not have that purple.

I quite like this setup, because I get all the advantages of a lightweight markup language, without the syntactic noise. And it’s much more easy to spot markup than in ordinary Markdown.

It is comparable to an advanced form of syntax highlighting, I suppose.


It sort of is, but I would very much want to be fully decoupled from HTML. (I’d like to be able to generate TeX and DOCX/ODT files, for instance.)

Note last edited November 2024.