[ppml] Proposed Policy: PI assignments for V6

[Kurt Erik Lindqvist]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


On 2004-12-08, at 20.49, Geoff Huston wrote:

>>> I am sorry but i think this is a huge waste.
>>> For the time being, most of mobile phones will not have a manet 
>>> behind, not one not many subnetworks. I think there must be 
>>> something very wrong if we assign a /48 to each mobile phone
>>
>> Do mobile phones even count as an "assignment"?
>
>
>
> Would personal LANs such as Bluetooth alter this perspective in any 
> way? The challenge here is to craft policies that look forward and 
> attempt to anticipate some of the more obvious near term technologies.
>
> Its also worth bearing in mind the larger picture of the total address 
> space - even if there was to be a /48 for each mobile handset, what's 
> the total address consumption. 1 billion handsets is 30 bits. Allowing 
> for an HD ration of 0.8 there is a further 8 bits - i.e. this is a 
> /10, or 1/1000th of the total address space.
>
> Of course this arithmetic allows 16 bits for a personal lAN. Maybe its 
> worth considering a slightly smaller subnet size - an 8 bit subnet 
> would imply that the total consumption would be a /18, for the entire 
> deployment across 1 billion such devices . So perhaps a /56 per 
> customer allocation in such context may represent a possible balance 
> between total consumption and recognition of this emerging personal 
> lan technology. Its worth considering in any case.

I think you numbers above are interesting. I have for some time been 
struggling to convince my self that a fixed /48 boundary for a "site" 
makes sense. I am less and less convinced.

- - kurtis -

-----BEGIN PGP SIGNATURE-----
Version: PGP 8.1

iQA/AwUBQbm1mqarNKXTPFCVEQLasACfUwjRpteunwUXoQM6DAgd2bJ+aCEAnRoU
HuRVkDAgvK0gTktnAMI0d647
=CsbV
-----END PGP SIGNATURE-----