On Thu, Nov 19, 2009 at 10:40 PM, Tony Li <[email protected]> wrote:

> Dae Young KIM wrote:
>
>  Question: In this arrangement, will the name (possibly concatenation of a
>> local ID and a local Loc) then be globally unique? Or is it the intention of
>> people that even the name may not be global?
>>
>
> This is now into the realm of variations among proposals.  Both are
> reasonable and workable alternatives.
>
> However, if one chooses a locally unique identifier, then there are more
> mobility issues.


Yes, definitely.


> If one chooses globally unique identifiers, then you have to have a
> mechanism to ensure uniqueness.
>

If we take the current assignment of IPv4 and IPv6 as identifiers, then we
can assure the global uniqueness of the identifiers. But, with awakening of
Christopher and Noel, I realized the ID should be exposed also in
inter-domain routing, extra to the AS # that I chose to take, then perhaps I
might not be able to keep the gateway routing table from explosion.

>
> In many cases, statistical uniqueness with a detection and recovery
> algorithm will be acceptable


Here I might seek your education, if you don't mind. Statistical? Detection
and recovery algorithm?

> .
>
> Tony
>

-- 
Regards,

DY
http://cnu.kr/~dykim
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg

Reply via email to