[arin-discuss] IPv6 as justification for IPv4?

John Curran jcurran at arin.net
Thu Apr 18 13:28:11 EDT 2013


On Apr 18, 2013, at 11:19 AM, Dmitry Kohmanyuk <dk at intuix.com> wrote:

> I think we can describe and name those pricing models by function of total IP addresses allocated.
> 
> So, "fee per address" would be _linear_ (cost = X * addresses);  
> "fee per bit of block size"  would be _logarithmical_ (this includes current scheme: cost = K * log2(addresses) + N)
> and fixed fee regardless of size of allocation (a-la RIPE model) is _flat_ fee (cost = N).
> 
> I am not doing advocacy of specific cost model (now or in the future) - but a formula is very useful to illustrate them.

Dmitry - 

The taxonomy of terms is quite useful.  I'll note that some approaches are
actually going to be a combination of these elements, and you need to add 
another potential element -

Rename _linear_ to _address linear_, i.e. _address linear_  is cost = X * addresses
Add _block linear_, which is cost = L * number of blocks

I would describe RIPE's new model as _flat_ plus _block linear_

FYI,
/John

John Curran
President and CEO
ARIN

 


More information about the ARIN-discuss mailing list