Re: draft-narten-ipv6-3177bis-48boundary-02.txt

[Kurt Erik Lindqvist <kurtis@kurtis.pp.se>]

On 14 jul 2006, at 06.05, Jason Schiller (schiller@uu.net) wrote:

> IMHO moving the boundary to a /56 is completely missing the boat on  
> what
> the actual problem is.  We discovered that classful addresses were a
> mistake and moved to variable length subnet masks.  Let's not  
> repeat this
> mistake by not having variable length subnet masks in ipv6.

Agreed!

> That rounds up to 16 subnets and doubles up to 32 subnets.  That would
> require 5 bits for subnetting or a /59
>
> I'm not trying to suggest that this is the right number, but rather  
> that
> we should take this sort of approach.

I actually think that is a policy to be determined by the RIRs...

- kurtis -