On Jun 10, 2021, at 11:35 AM, Paul Wouters <[email protected]> wrote:
>> We propose that this be the actual,
>> long-term solution.
> 
> I understand the desire but I don’t agree as this signal is insecure, and 
> foresee TLDs abusing this as potential nation state monitor / privacy leak.

Please say more. I don't see how this proposal leaks anything that could not be 
trivially determined by probing.

> It is also dangerous when used via resolvers.

Please say more. To me, it is only useful for resolvers.

> I still prefer something with DS than can be signed, and validated by the 
> child as their intend via CDS. With transparency monitoring.
> 
> If we are using overloading, might as well overload securely.

If you write up a draft, I'm happy to send responses to particular statements 
in the draft. I don't see how such a DS could be specified in a way that would 
get more than a trivial amount of deployment. I would be happy to be wrong, 
given that DS is signed in the parent.

--Paul Hoffman

Attachment: smime.p7s
Description: S/MIME cryptographic signature

_______________________________________________
dns-privacy mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/dns-privacy

Reply via email to