Hi Ben,

On 17 Nov 2026, at 21:03, Ben Schwartz <[email protected]> wrote:

>> (a) authentication of servers is not always trivial (nameservers often have 
>> many names; addresses can be subverted since the routing system is not 
>> strictly trustworthy);
> 
> This authentication should be clearly defined via DELEG.  (I assume
> any discussion of authenticated ADoX is in reference to DELEG, as
> there is no other active specification for this.)

There are three:

 - RFC 9539
 - draft-hoffman-deleg-secure-transports-00
 - draft-johani-dnsop-transport-signaling-03

9539 uses port-knocking to find encrypted transport; 
draft-hoffman-deleg-secure-transports-00 uses signals published by zone owners 
in future delegations encoded using DELEG; 
draft-johani-dnsop-transport-signaling-03 uses signals published by nameserver 
operators.

>> (b) unsigned data might have been modified on a secondary server, or in 
>> transit between primary and secondary
> 
> In my view, DELEG should clearly specify that by making a nameserver
> "authenticatable", the zone owner marks that nameserver as holding a
> trustworthy copy of the zone.  We could support the idea of an
> "authenticated connection to an untrustworthy or vulnerable
> nameserver", but I think this is a confusing and counterproductive
> middle ground.

This is a brand new concept, but in any case it's orthogonal to the topic at 
hand which is user privacy.

>> (c) an insecure delegation from a parent cannot be trusted (the NS RRSet 
>> above the zone cut is unsigned)
> 
> DELEG doesn't have this problem.
> 
>> so the degree of integrity protection provided by DNSSEC and that provided 
>> by ADoX are not the same.
> 
> I agree!  But it can be quite a strong a form of authentication, and I
> believe it may prove to be much more widely deployable than DNSSEC.

You may well be right; I may well not disagree! But it's not what Leon's draft 
is about, unless I misunderstand its purpose.

>>> If we intend to support
>>> this arrangement, then I believe we should extend it with some signal
>>> to indicate ADoX-based authentication.  (We could even just overload
>>> the AD bit, but I suspect that would not have consensus.)
>> 
>> I'm not sure what ADoX-based authentication means, above. What would an 
>> authority server be signalling in a response to a recursive server by 
>> setting AD=1?
> 
> I'm referring to an "AD2=1" signal from recursive to stub, meaning
> "nothing in this response could have been poisoned on the path between
> me and the authoritative servers".

Isn't that what Leon's draft is trying to achieve in section 5.2, but with a 
richer signal?


Joe
_______________________________________________
DNSOP mailing list -- [email protected]
To unsubscribe send an email to [email protected]

Reply via email to