I think that'd be useful for commenting but nothing else.

Whenever I see something in code, I wonder, what does this do? If I started seeing bold and italics and it did nothing but draw attention, I would almost certainly feel it a distraction.

> I think that'd be useful for commenting but nothing else.

But isn’t that a good enough use case?

Almost nobody bothers to put proper formatting, diagrams, tables, formulas, references etc in their source code. At best we get ASCII approximations.

To see some examples of what it could be like look at Mathematica, Jupyter, literate programming, and org-babel.

.. all of which require some kind of rendering software that has to actually be installed on your machine.

So do you need for images, photos, spreadsheets, PDF files, websites, schematics, board layouts, databases, mechanical drawings and models, neural network models, music masters, audio tracks, video footage and any other file format used by professionals.

Better yet, all the formats I mentioned are raw text and not even binary. And all have well developed tools for editing.

More than formatting, I am thinking about defining things like function and classes in a formatting-independent way, and dismissing the order of definitions totally.

You should have a list of classes and namespaces and not have to worry about how they are actually stored in flat files.

Yes, that too. There's no point imposing an order on functions and classes; I think that when you're looking at a function, related functions should somehow float nearby. (This doesn't seem to be a very popular view.)

Isn't it more useful if such marks like showing something in bold or a different color are automatically applied to classes of marks like identifiers or strings?