[members-discuss] Re: New RIPE NCC Charging Scheme Models: Feedback Required

>> I'm supporting this kind of simplification of LIR size calculation, but
>> personally I feel that IPv4 allocations per category should be changed.
>> There's disproportion between number of addresses and category -

> As Daniel, I like this kind of simplification of LIR size calculation.
> But I am XS, and I think I will go Small or Medium... (depending of the way it is calculated).


	I think that the classification of a population into in a small number of classes for taxation or otherwise cost calculation owes to two basic reasons: (1) It dates back to when there existed no computers and therefore was not possible to make it perfectly fair for everyone without extensive manual calculations; (2) public laws, and quite often private entities' rules, being made by lawyers, who usually have a minimum background in mathemathics - therefore simply not enough knowledge among lawmakers in mathematics to make more precise schemes.

	Being the bunch of creative, numbers-oriented people that we (the LIRs and the RIPE) are, I think we could come up with something much better. I envision a formula which will give a calculation to the penny which is perfectly fair for each and every LIR.

	Such a formula doesn't have to be terribly complicated - no need for Fourier transforms or imaginary numbers. It may simply be a sum of the cost components, each with a factor to account for its weight in the overall cost.

	Cost components may include things like:
		- Amount of IPv4 space allocated
		- Amount of IPv6 space allocated (can be waived for a few years to help LIRs deploy IPv6)
		- Amount of IPv4 space assigned (i.e., usually generating revenue for the LIR)
		- Amount of IPv6 space assigned (same as before, and probably waived for a few years for the same reason)
		- Amount of AS numbers
		- Amount of unused (or undocumented) IP space (to give a good reason to keep the database updated!), 
		- Assistance to meetings (this one with a negative factor, of course)
		- Operating cost of the RIPE NCC divided by number of LIRs at previous year end (to cover expenses tied to managing a LIR which are independent of its size)
		- Labs projects sponsorship (negative factor for this one too)

	Each LIR would then be able to calculate well in advance its cost for the following year - so it would not be different from now, and therefore, we'd not face the opposition of our financial deparments. Quite the contrary - the cost would be somewhat more predictable, and the contention would be limited to the factors for each cost component. But because each factor would not fluctuate much from year to year, and even if they did the overall effect would be relatively small, they would not be as big a source of trouble as the proposed changes are now.

-- 
Alfredo Sola
ASP5-RIPE
http://www.solucionesdinamicas.net/