Re: draft-gont-6man-managing-privacy-extensions-00.txt

[Ran Atkinson <[e-mail]>]

On 10  Mar 2011, at 02:34 , Dan Wing wrote:
> Nobody wants it removed in corporate deployments, either.  

That statement is far too strong; it simply is not true.

> Consider for a moment an IPv6-enabled telephone,
> on the desk of a Very Important Person at a company, ...

(Laugh.  I don't believe in that untrue example for a second,
but thanks for the humorous start to my morning.  :-)

Oh, and consider that Caller-ID, usually including a full name,
is already widely deployed in many countries.  It is so
much simpler just to read the Caller-ID string off of the
telephone handset.  

If one wants to intercept traffic, it is MUCH simpler to intercept 
the SIP traffic -- which is never encrypted in real-world corporate 
deployments because that would make VoIP too hard to debug.

> If we don't have IPv6 privacy addresses, we will also soon
> see NAPT66 (with UDP and TCP port rewriting) in order to 
> achieve the same result as privacy addresses:  trying to 
> obfuscate which host is communicating.

It seems pretty clear that Fred's NPTv6 is going to be
deployed in at least some locations, albeit for entirely 
different reasons.  I'm not sure if that meets your 
definition of NAPT66 or not.

Cheers,

Ran