Here's my hot take: if you're designing a new programming language in the modern era, even a systems language, ignore the precedent of C and don't use &|^~ as bitwise operators. You can still have infix bitwise operators, but spell them out as bitand/bitor/bitxor/bitnot. Then you can just use &| for logical and/or, which are 1000x more common than bitwise and/or, and you can reclaim ^ for exponentiation as well. And don't forget to fix C's broken bitwise operator precedence while you're at it.
Interesting. I'd go the other way. Continue using single-character bitwise operators, but spell out "and" and "or" for the logical operators.
Rationale: in normal usage, short-circuiting logical operators are, in effect, a special kind of control flow statement, and control flow statements are typically spelled out. Bitwise operators are more unequivocally meant for calculation, and therefore perhaps more deserving of similar syntax to the arithmetic operators, despite their less frequent usage.
I think that this way of drawing the distinction might be particularly relevant in a language that disallows conditional branching - and, by extension, short-circuiting logical operations - on certain kinds of data the way Rune does.
Hard disagree on single characters for logic and exponentiation. ASCII has too few special characters and you'd want to use those elsewhere, without creating ambiguity. Elixir got it right.
