Re: Logotypes in certificates

["David P. Kemp" <[e-mail]>]

"Manger, James H" wrote:
> 
> An entity does not have to have a unique logo.  To be useful,
> it is sufficient that the logo unambiguously identifies
> an entity (within the context of use).
> 
   ... and
> 
> Images (e.g. logos) are a widely used form of identification (and for good
> reason, as they are very human-friendly (ease to use)).  I suggest PKIX
> should change to support logos, rather than trying to changes our five
> senses.



A correlary of the first statement is: If the logo does not
unambiguously identify an entity within the context of use, then
it is not useful.  I predict that Stefan will find it impossible
to propose a criterion against which it can be determined that
a CA is operating correctly, i.e. that the CA will certify a logo
that identifies the entity, and that it will not certify a logo
that mis-identifies the entity.

If no one can propose a method by which a CA should decide not to
certify a logo "similar to" the Chevrolet bowtie or the AT&T deathstar
for someone other than Chevrolet or AT&T respectively, then it is
obvious that changing PKIX to support logos has *NO* value to the
user.

I submit that it has negative value; a misleading logo
in a certificate is worse than none at all.


(I concede that a logo capability has marketing value to CAs and
entertainment value to us geeks.  But PKIX shouldn't be about
satisfying those urges at the expense of the user.)