[arin-ppml] fee schedule and allocation

[Robert E. Seastrom]

Randy Bush <randy at psg.com> writes:

>>>> The main justification for the ... minimum is to make sure that they
>>>> never have to come back for space
>>> I heard that exact sentence uttered almost 20 years ago, and let me tell 
>>> you... that worked out just fine!
>> Do you think we got it right this time, with 340 trillion, trillion, 
>> trillion addresses (or 35 trillion /48's)?  :)
>
> as danny cohen said last wednesday
>   first it was 8 bits
>   then times four to 32 bits
>   then times four to 128 bits
>   and he assumes times four to 512 bits

Unfortunately, glib quips that conflate multiplication with
exponentiation help perpetuate people's gross underestimation of the
vastness of the address space we're talking about here.

Let's restate that quote as:

first it was 2^8
then 2^(8*4) = 4.2 * 10^9
then 2^(8*4*4) = 3.8 * 10^38
and he assumes 2^(8*4*4*4) = 1.3 * 10^154

(the last equation suggests enough addresses to give 10^74 addresses
to each atom in the visible universe.  dealing with a routing table to
do this or pinging each address in non-universe-epochal-time is out of
scope for this discussion).

-r