[members-discuss] Charging scheme 2025 proposal (logarithmic)

Thank you for the contribution Sebastien.

Point of view of the other responses, to be able to say, if this model can pays the actual costs (~40m €), we need some data i think.

How can be tested or calculated, if this model generates enough budget or less budget?

--
Mit freundlichen Grüßen / Best Regards

Murat TERZIOGLU
PREBITS

Bochumer Str. 20
44866 Bochum
Deutschland

Telefon: 0234/58825994
Telefax: 0234/58825995

www.prebits.de
info at prebits.de

USt-ID: DE315418902


-----Ursprüngliche Nachricht-----
Von: members-discuss <members-discuss-bounces at ripe.net> Im Auftrag von Sebastien Brossier
Gesendet: Dienstag, 16. April 2024 15:51
An: members-discuss at ripe.net
Betreff: [members-discuss] Charging scheme 2025 proposal (logarithmic)

Hello all,

I propose to add the following model to the charging scheme 2025 voting options.


*1 - Introduction:*

This charging scheme is heavily inspired by APNIC. If you are not 
familiar with this, you can see an example here:
https://www.apnic.net/get-ip/apnic-membership/how-much-does-it-cost/member-fees-calculator/

The main idea is that each LIR pays according to its resources, but not 
linearly. You don't pay twice as much because you have twice as much 
resources.
The resulting fees are similar to what the other RIRs are charging, with 
infinite granularity (no categories).

It can be easily tweaked to reach any desired budget, and will remain 
viable when IPv4 has disappeared.

I have made IPv6 less punitive compared to APNIC, because RIPE has 
larger initial allocations.

Independent resources fees, sign-up fee, lack of ASN fee, remain as 
before in this proposal. I believe it is better to have a separate 
debate on these subjects at a later time.

The goal of this charging scheme is to lower the cost for members with a 
very low amount of resources, in order to attract newcomers and retain 
existing members. This way the RIPE NCC membership will remain numerous 
and diverse.


*2 - Charging scheme:*

(Warning: math incoming !)

IPv4_count = number of IPv4 addresses allocated
              (excluding independent assignments and legacy)
IPv6_count = number of IPv6 /56 subnets allocated
              (excluding independent assignments)

Base_Fee = 638 EUR
Bit_Factor = 1.31
Minimum_Fee = 500 EUR
Offset_IPv4 = 8
Offset_IPv6 = 24

IPv4_Fee = Base_Fee * Bit_Factor^(log2(IPv4_count) - Offset_IPv4)
IPv6_Fee = Base_Fee * Bit_Factor^(log2(IPv6_count) - Offset_IPv6)

Fee = max(IPv4_Fee, IPv6_Fee, Minimum_Fee)
         + 50 EUR per independent resource (excluding ASN)

My simulation, based on public data (2024-03-28), results in an average 
fee of 1900 EUR per LIR (+ 50 EUR per independent resource), so it 
should provide the same budget as the other options.
If RIPE NCC find different results with their simulation, they can 
adjust Base_Fee.


*3 - Examples:*

50 EUR per independent resource should be added to all these fees.

No allocations:			  500 EUR
IPv4 /24 and/or IPv6 /32: 	  638 EUR
IPv4 /23 and/or IPv6 /31: 	  835 EUR
IPv4 /22 and/or IPv6 /30: 	 1094 EUR
IPv4 /21 and/or IPv6 /29: 	 1434 EUR
IPv4 /20 and/or IPv6 /28: 	 1878 EUR
IPv4 /19 and/or IPv6 /27: 	 2461 EUR
IPv4 /18 and/or IPv6 /26: 	 3224 EUR
IPv4 /17 and/or IPv6 /25: 	 4223 EUR
IPv4 /16 and/or IPv6 /24: 	 5533 EUR
IPv4 /15 and/or IPv6 /23: 	 7248 EUR
IPv4 /14 and/or IPv6 /22: 	 9495 EUR
IPv4 /13 and/or IPv6 /21: 	12439 EUR
IPv4 /12 and/or IPv6 /20: 	16295 EUR
IPv4 /11 and/or IPv6 /19: 	21347 EUR
IPv4 /10 and/or IPv6 /18: 	27965 EUR
IPv4 /9 and/or IPv6 /17: 	36634 EUR
IPv4 /8 and/or IPv6 /16: 	47991 EUR

Largest LIR is just below 60 kEUR.

There are no categories, so your fee can be somewhere between these numbers.

If you think the fees are too high, I invite you to read the fee 
schedule of the other RIRs.


Thank you if you've read this far.


Best regards,
Sebastien Brossier

_______________________________________________
members-discuss mailing list
members-discuss at ripe.net
https://mailman.ripe.net/
Unsubscribe: https://lists.ripe.net/mailman/options/members-discuss/m.terzioglu%40prebits.de