Death of mixed methods? Or the rebirth of research as a craft