Unicode and encoding notesfmt treats char-based strings as UTF-8. That's the right default for nearly every modern codebase,