Hi Michaela,

On Thursday 18 December 2008, Michaela Vanderveen wrote:
> Hi,
>
> I have reviewed this draft and have the following comments:
>
> Other reviewers think that the draft is reaching too far into the
> solution space, for a problem statement. I find the level of solution
> hints to be acceptable. In any case it seems that maybe it's a bit
> late for a WGLC?

Wandering into solution space in a problem statement document is not 
problematic per se.

But we're looking at a security problem, thus what is problematic is, 
the problem statement does not define the trust model from which we can 
observe what the problem is. 

As a start we could try to answer these questions: where is the NDproxy, 
who owns it, who operates it, who authorizes it, to do what? 

An answer could be: the NDproxy is on small residential network 
connected to an ISP by a CPE, it is owned by the customer but operated 
by said ISP, it is authorized by CGA owners on the link to send NS/NAs 
for said CGA owners, and by said ISP to advertize the prefix delegated 
to the CPE.

Is this a good answer? Is it the only good one?

--julien
_______________________________________________
CGA-EXT mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/cga-ext

Reply via email to