Re: [hrpc] "Paul Vixie and Peter Lowe on Why DoH is Politically Motivated"

[Eliot Lear <[e-mail]>]

We run this risk of this discussion getting quite 'meta' (no offense to 
Facebook).  Yes, DoH and DoT are both public protocols, and in 
themselves I don't think they are particularly harmful.  It is precisely 
when they bypass local infrastructure when things get messy.  You may 
say, “but a private mechanism can do the same thing.”  This is true.  
But when a private mechanism does the same thing, we can in fact place 
controls on the infrastructure to either prevent it or reduce it 
substantially, "simply" by binding network access to name resolution 
(ok, it's not THAT simple but it can and has been done).  But that 
doesn't work unless the the local resolvers are used.

Do motivations matter?  Certainly.  That's how work gets done. If 
motivations can be impacted, that's how work is directed.  But I tend to 
agree with you that it is the outcome that matters by orders of 
magnitude more, as far as we are concerned.  And here, we really can't 
say if there is a good outcome or a bad outcome – yet.  It depends on 
what information is lost, and whether Bad Guys are covertly using newly 
created channels, and whether you are even able to identify that.  The 
jury is still out (a phrase that has some interesting meaning today).

More data needed.  But where is that data?

Eliot


On 16.11.21 20:18, Ted Lemon wrote:
> On Tue, Nov 16, 2021 at 11:36 AM Vittorio Bertola 
> <vittorio.bertola=[e-mail]> wrote:
>
>     That's exactly why the difference between DoT and DoH is political
>     way more than technical: I trust you that both may be used in ways
>     that make them hard to block, but for DoT it is an unintended
>     consequence, while for DoH it is an intended objective of the
>     designers, stated in the specification itself and in several
>     public statements.
>
> And this is a great illustration of the good reason why we talk about 
> political versus technical in the IETF: because if we consider only 
> the political and not the technical, we can wander off into the weeds 
> doing things that don't matter.
>
> In this case, you're drawing a distinction of intent, but what matters 
> is the effect of the protocol, not what was intended. As an example, 
> we've seen a few people assert that DoH is bad because it makes it 
> possible to bypass the local resolver in a way that is difficult to 
> detect (which is actually debatable). But in fact it's trivial without 
> the DoH protocol to do the same thing—it's just not standardized. So 
> the difference between DoH and this non-standard protocol that I 
> speculate might exist is merely that one is documented and the other 
> isn't.
>
> So if we imagine that we are somehow protected when we don't document 
> the protocol, we are whistling past the graveyard. This is the 
> important distinction that we lose when we reduce this to a political 
> discussion, and it's the reason we shouldn't.
>
> _______________________________________________
> hrpc mailing list
> [e-mail]
> https://www.irtf.org/mailman/listinfo/hrpc