Yeah, unicode characters in Nim code are supported. However, if by `x²` you'd want to square an identifier `x`, that won't work. The Nim lexer parses `x²` as a single identifier.
We do have infix unicode operators though. So `x ÷ y` (spacing required though!) could be implemented easily. I use this for `±` in Measuremancer [0] (which btw also supports Unchained units, for error propagation on unitful measurements).
Required spacing is a blocker/friction, that's one of the benefits of DSLs, I guess, being able to eschew the conservative syntax of the general-purpose languages (though Unicode support is a good step forward)
