We rarely see attack papers published in architecture conferences. What is the benefit of publishing attacks? What makes an attack paper instructive and valuable?