Signature semantics

[reynt0]

On Tue, 4 Nov 2008, Robert J. Hansen wrote:
  . . .
> Idiot-proofing is a very bad idea.  Systems cannot be made idiot-proof, since 
> we're constantly developing higher and better grades of idiots.  Systems can 
> be made user-friendly; they cannot be made idiot-proof.
  . . .

Hmmmm, so the problem is ..... It's hard to be smart enough to
understand idiots well enough to predict them?  *<;*}