[ppml] Big numbers

[bmanning at karoshi.com]

> The _REAL_ problem is whether or not that address space is or will ever 
> be routable.  We can't flat route the v4 space or even close to it, 
> multiplying that number by 2^96 doesn't help.  2^45 (or 2^48 or 2^32 or 
> 2^29, pick your big number) has the potential to be a swamp of 
> inconceivable proportions.

	actually, its pretty close.  slightly older view of
	IPv6 has the top/left 32 bits as fixed, the bottom
	64 bits as fixed.  the second 32 bits as the delegation
	space for which CIDR applies....  which is the same
	issue as flat routing of the IPv4 space as /32s.
	Newer versions of address architecture have "freed"
	up the constraints on the left hand side...
> 
> The question is, what does the ARIN community believe ARIN's role is or 
> should be in limiting the creation of this new, bigger, deeper, swamp?
> 
> Rgds,
> -drc
>