On Fri 2017-04-28 08:57:39 -0700, Joe Touch wrote: >> -- existing HTTPS isn't going to move to a new port, which means >> traffic on that port would be suspect to a would-be censor or a >> pervasive monitor, right? > > All traffic on the Internet is already subject to censorship and > monitoring. I can't see how you can traverse someone else's link or > router without that possibility, regardless of port.
To clarify any miscommunication, i said *suspect* (in the sense of
"suspicious"), not *subject*. One of the aims of the draft is to make
the two protocols less distinguishable from one another to a network
attacker who might prefer one over the other (and consider the other to
be suspicious). A new port assignment fails to meet that goal.
Of course users are subject to censorship and monitoring by networks
that they rely on. This power imbalance between network operator and
user is part of the way that digital transport works. We can try to
even out this power imbalance by demonstrating mechanisms for the user
to make their traffic indistinguishable from other traffic and opaque to
the network operator -- that way the operator cannot target them for
censorship or surveillance without incurring heavy collateral damage.
The goal of this WG is to provide confidentiality to the DNS querier
against a network attacker. If every mechanism that we produce can be
identified by a network adversary who wants to violate that
confidentiality, then the adversary can simply kill that traffic,
causing the user to fall back to cleartext DNS or to leave the network
entirely. This draft is trying to defend against that scenario, which
is why i brought it here first.
Your points are noted, thanks, and i'll branch the discussion and start
talking with the HTTP folks soon.
Regards,
--dkg
signature.asc
Description: PGP signature
_______________________________________________ dns-privacy mailing list [email protected] https://www.ietf.org/mailman/listinfo/dns-privacy
