On Saturday 21 May 2005 17:33, Ram A Moskovitz wrote:
> The fallacy in your argument (there is no contradiction between statements)
> is that it requires there to be no value to VeriSign's reputation today
> which is a false statement.

Well, I would have put it more softly than that.

First, it is not a statement that there is no value
in VeriSign's brand.  There are three important
limitations:  the brand of import is that with the
retail certs brand, and not the rest of the market
where Verisign plys a trade.  The certificates is
only a small part of Verisign's operations, one
part of one division.  So Verisign mayy have a
good brand elsewhere, but what is important is
the brand that the buyers of certs see.

Secondly, it is not that the argument only works
if there is no value.  The relative equation is
whether there is *sufficient* brand recognition to
have the effect.  This is a complicated thing to
measure, so I put that forward as a conjecture
(although see earlier posts for some cites).

Thirdly, bear in mind this is not an attack on Verisign.
If one were to substitute Verisign for Comodo, then
there would be no argument that their brand is
relatively small and unheard of out in the retail
cert buying market place.  Even a relatively
large player like Godaddy might not have the
associated brand as people might not realise
they sell certs.

If anything it is an argument that Verisign also
should be included in the list as not having
valuable brand as a brake on behaviour.

So in a marketplace where arguably only one
CA might have a plausibly strong enough brand
to effect their behaviour, that still leaves us with
the conclusion that there isn't much stopping
the original Duane-posted threat:  whether a
CA would decide to fight the feds in issuing an
unauthorised substitute cert.

Drawing from Frank's policy, we now know
that MoFo does not consider such possibilities
directly in the choice of the root list, just like
it doesn't rule out control-of-domain certs, and
all the problems that arise from that issue.

Which leaves a software based response.  The
only way to protect the user against a substitute
cert attack is in the browser.  Which was what
Duane was asking for.

iang

> On 5/21/05, Ian G <[EMAIL PROTECTED]> wrote:
> > On Saturday 21 May 2005 02:22, Ram A Moskovitz wrote:
> > > You have repeatedly argued that the value of brand and reputation
> > > plays into a CA's behavior. Here you are saying that a CA would toss
> > > its reputation to keep one of it's small (revenue size) customers
> > > happy.
> >
> > Correct on both counts. Now, you are implying that
> > there is a contradiction in these two statements.

-- 
Advances in Financial Cryptography:
   https://www.financialcryptography.com/mt/archives/000458.html
_______________________________________________
mozilla-crypto mailing list
[email protected]
http://mail.mozilla.org/listinfo/mozilla-crypto

Reply via email to