Comment by slx26

I tend to agree with your line of reasoning and wanted to write a blogpost about it eventually.

I implemented a "markdown-like" parser, and I noticed the biggest problem is that the edge cases are a mess, the notation is a mess, and doing efficient single-pass parsing is unnecessarily complicated (urls are also an unnecessarily complicated element that hits a parser of this kind, but that's a story for another day).

I generally agree that making special unicode characters to represent these things could go a long way into making everything easier to understand, parse, disambiguate.

To be fair, this could be tested without insane effort: unicode has a range of characters reserved for private use. Adapt a font and make a simple online editor where holding some key will make javascript generate the appropriate unicode character. We should just try it and see if it's really better.

If the concept works out well, same as we have modifier keys for alt, shift, ctrl, etc., we could use that and integrate in keyboards without much trouble. Of course, it's a big change, but I think it's being proved that there's a common set of markup needs that most platforms should handle.

And many might say: well, it's not like everyone needs to implement markdown, we can have a single implementation and re-use it. It's not such a big deal. Well, I'd argue that the right approach is make things as simple as they can be. This kind of markup is necessary, but we can make it much simpler (now someone might say that pushing more things into unicode is not a good idea, as unicode is far from as simple as possible and perfect, but that shall be discussed another day too).