[ppml] Big numbers

David Conrad david.conrad at nominum.com
Tue Apr 8 14:46:13 EDT 2003


On Tuesday, April 8, 2003, at 07:59  AM, Lee Howard wrote:
> Big bang equal infinite.

Took me a while to parse this.  Thought you might be making a 
grammatically incorrect cosmological statement.

> We can all think of worst-case scenarios where
> an attitude that IPv6 addresses are free and undepletable could lead to
> depletion.

I agree.  I am not suggesting v6 is infinite.  I am, however, 
suggesting that the scale of the numbers here is a paradigm shift in 
terms of how address allocation/conservation should be thought of.  
Thinking in terms of the historical /8 allocations to universities et 
al doesn't apply (at least in the sense of address conservation) in v6 

> The example I like to give is of IPv6 addresses replacing
> UPC codes, and every trip to WalMart results in another couple hundred
> addresses in the trash bin at the curb.

But wouldn't these be /128s?  Got oodles of /128s in a single /48 
delegated to WalMart... :-)

> If we treat v6 as if it's
> infinite, then the value of each address is infinitely small and can be
> infinitely wasted.

Of  course.  Again, I am not suggesting v6 is infinite.  It is a 
limited resource, albeit of vast size.  I believe allocation requests 
will be self-limiting, if for no no other reason, there is a cost to 
registration, even for very big numbers, thus there will be a simple 
constraint on the amount of address space allocated, namely money.


More information about the ARIN-PPML mailing list