[thangalin]

Can you quantify "much better"? I compiled a set of over 70 features offered by a variety of plain text formats:

https://keenwrite.com/blog/2025/09/08/feature-matrix/

Are many features missing from the list? From what I can tell, objectively, plain text formats offer largely equivalent functionality.

[aragilar]

There are those that have well defined extension points (e.g. TeX, rst), and those that are ad-hoc, of which the best example is markdown. TeX, via packages and wrappers, can do practically anything. rst has directives (blocks) and interpreted text (inline) which can also effectively do anything (along with substitution references which are more macro-like). Specific "interpreters" (for lack of a better term) which you link to naturally have specific features by default (and some are more extensible than others e.g. pandoc which when writing out LaTeX lets you embed LaTeX in the markdown, so "markdown" in this case is turing complete).

I think if you define "better" as having well-defined extensibility to enable multiple implementations (i.e. not ad-hoc things pandoc lets you do), then rst (which can be transformed into XML as per https://docutils.sourceforge.io/docs/ref/doctree.html) would be "better" than markdown.

[pseudohadamard]

Does anything need all those features though? If you need more than headings, bold, italic, quoted text, and code, you should probably be using HTML or Latex or something. So a better goodness-of-fit function would be is it easy to remember/use, and is it non-intrusive inline with the text?

[Linux-Fan]

I don't think anybody needs all of the features at once but people have different preferences. E.g. I typically do well without bold formatting (I only need one level of emphasis which is served well by italic) but I want tables, links and lists very often.

Also I like the WYSIWYG feature of Markdown where it has an advantage over the traditional Markup languages like HTML, LaTeX, groff etc. of being easier to read in the text file. Dedicated syntax highlighting can go a long way to make markup easier to read, though.

[Linux-Fan]

The original Markdown has fewer features than listed for the more advanced formats in the table. Hence if someone uses reStructuredText it is more precise than just saying “Markdown” because Markown could refer to anything from the original minimalist featureset to the vastly extended format supported by pandoc if given the appropriate CLI arguments.

Some text-based formats have more options for tables e.g. alignment of columns (it may help with numbers to right-align them) or multirow/multicolumn options. Some formats support definition lists (corresponds to <dl> in HTML) - a feature which I often find valuable but was not included in the original Markdown IIRC.

One advantage of using a text-based format is that it can be exported to either LaTeX or HTML and Markdown seems to prefer the HTML output by explicitly supporting inline HTML as an escape hatch for more complex constructs (e.g. tables with rowspan/colspan). In addition to often not being supported for a non-HTML export-type it also hurts the WYSIWYG experience when reading the file like plain text.