That's really interesting, thank you for explaining that! Somehow I've never thought to implement interferences to describe types that already exist.