Re: identity persistence and comparison issues

[Erik Nordmark <Erik.Nordmark@sun.com>]

> Without having thought this through very much I have a few questions on 
> this,
> 
> As you would need to keep state that would be very much like a cache. 
> Would you not open your self up for "cache-poisoning" very much like 
> todays DNS?

In principle,yes.

But the WIMP+AID idea wouldn't AFAIK have more of a problem in this space
than NOID or cb64 type approaches. All can use the return routability
implicit in establishing the state to place some trust (no worse than today's
Internet) in the initial locators, and then verify additional locators
before they are marked as "usable" in the cache.
WIMP+AID should be able to take the same approach.

  Erik