Additional lookups (was Re: [ietf-dkim] Re: 1368 straw-poll)

[Wietse Venema]

Charles Lindsey:
> On Thu, 01 Mar 2007 13:44:21 -0000, Wietse Venema <wietse at porcupine.org>  
> wrote:
> 
> > On a friendly internet with only cooperating parties, this might
> > make sense.  But the world has changed. With today's internet it
> > would be a fundamental mistake to make more distinctions than:
> >
> >     the signature was verified
> >     other
> >
> > If the verifier gives different treatments to different types of
> > "other", then the bad guys will exploit the verifier's behavior.
> 
> And how do you stop verifiers doing that?

There is no cure for stupidity, but I can try to educate.

> Verifiers will do as they think fit (i.e. what their clients will pay  
> for), whatever our standards say. If some likely (though deprecated)  
> verifier behaviour leads to exploits by the Bad Guys, and there is an easy  
> way to counter the exploit (e.g. by clearer information in the SSP), then  
> it would be wise to dopt it.
> 
> "Defence in depth" is the term, I believe.

SSP is not a cure for exploitable verifiers. 

"Wrong solution for the wrong problem" is the term, I believe.

	Wietse