Is the syntax for symbolic computation already near optimal or should something be rethought and redone?