Re: [homenet] About Ted's naming architecture presentation and document

[james woodyatt <[e-mail]>]

On Nov 21, 2016, at 15:11, Ted Lemon <[e-mail]> wrote:
> 
> Part of the goal of providing a naming infrastructure for the homenet
> is precisely to avoid what you are describing, James.   While it's
> true that consumer IoT manufacturers do seem to be using that model
> now, it's a broken model, and work is underway to obsolete it in the
> open source world.   Of course, that _does not_ mean that IoT devices
> will be publishing their services in the public DNS, but the dogleg
> model has many problems, not the least of which is that devices that
> use it and control power consumption are a significant risk for
> utilities.

This goes to the heart of my criticism of the Homenet Naming Architecture draft. If there is anything in any of the Homenet working group documents or pending drafts that contradicts the recommendations of RFC 6092 that amount in practice to a prohibition against passive listeners in the home network from being reachable by arbitrary exterior hosts, then I’m not seeing it. Could you provide me with a pointer to the relevant passage in the drafts? Without that, I can’t see how there’s really a strong case for doing any of this naming architecture work.


--james woodyatt <[e-mail] <mailto:[e-mail]>>