Re: [ietf-privacy] Fwd: draft-huitema-dnssd-privacy-01.txt

["Christian Huitema" <[e-mail]>]

(Moving this conversation to DNS-SD mailing list)

On Thursday, June 23, 2016 2:53 AM, S Moonesamy wrote:
> 
> Hi Tim,
> At 05:18 22-06-2016, Tim Chown wrote:
> >We're encouraging discussion of privacy considerations in the WG. As a
> >result, we now have a draft (see below), including an initial proposal
> >for a solution, for which we'd welcome wider review. The draft also
> >addresses mDNS/DNS-SD privacy within single subnet scenarios.
> 
> One of the privacy issue identified in the draft (Section 2.4) is device
> fingerprinting.  In Section 3.1, it is proposed to solve the privacy
issues
> described in Section 2.1 by obfuscating instance names.  If I had to pick
one
> privacy threat for that I would choose "correlation".  Obfuscating service
names
> would not address that.

Section 3 describes an initial design that was then abandoned. I guess that
in the next revision we could just remove that section entirely.

On the other hand, the proposal was indeed to use different obfuscated names
at different locations.


> If I understood the draft correctly, the solution "to prevent tracking
over time
> and location, different string values would be used at different
locations, or at
> different times".  QR-codes are used to generate a shared secret and
establish
> trust between two or more "friends".

The private discovery service relies on pre-existing pairings. The pairing
solutions are only drafted in very vague terms in the draft. I really wonder
whether we should go define a complete pairing protocol. Is that in-charter
for DNS-SD? What about competing with existing solutions over Bluetooth,
Wi-Fi, and certainly many more?

-- Christian Huitema