> In short, no matter how large you try to span your test coverage and surface, > there will always be a residual chance for something slipping through. "the narrower the window, the more likely you are to hit it." :)