Re: [Trans] DNSSEC also needs CT

["Osterweil, Eric" <[e-mail]>]

On May 22, 2014, at 1:38 PM, Nico Williams <[e-mail]>
 wrote:

> On Thu, May 22, 2014 at 12:21 PM, Stephen Kent <[e-mail]> wrote:
>>> DNSSEC is a PKI [of sorts; please, no need to pick nits about that].
>> 
>> agreed.
>> 
>>> It stands to reason that DNSSEC should have similar trust problems as
>>> PKIX.  I believe it does indeed.
>> 
>> PKIX, per se, does not have the trust problems that seem to motivate
>> CT; the Web PKI does. That PKI has always had a serious problem because
>> any TA can issue a cert for any Subject, irrespective of the Subject name.
>> because DNSSEC intrinsically incorporate the equivalent of PKIX Name
>> Constraints, it does not suffer from that specific problem. That's not to
>> say that mis-issuance is not possible in DNSSEC, but rather that its
>> effects are more limited.
> 
> I've already said that DNSSEC fundamentally has strong naming
> constraints, whereas the TLS web server PKI doesn't (and worse: has
> been deployed with none).
> 
> However, I don't think it necessarily follows that having name
> constraints -> no need for CT.
> 
> CT is about keeping CAs honest.  In the TLS web server PKI there are
> very many CAs to keep honest, therefore anything that helps automate
> that task is greatly helpful.
> 
> DNSSEC "CAs" also could use being kept honest, even if none of them
> have yet failed to be honest.

If I understand your point (perhaps I don't) the type of ``honest[y]'' that you are talking about (in the Web PKI) refers to a CA vouching for a name binding that is illegitimate. How do you imagine this is possible in DNSSEC?  I could (for example) stand up a DNSSEC signed zone for someone else's zone, but because key verification and key learning are tied to the DNS delegation hierarchy, no resolver would learn of my doppelgänger zone, right

Eric