Dear Christian, Thanks for your explanations. You most certainly know that, for now, LIR's incomes derive primarily from the amount of IPv4 space they can rent to their customers. Any notion of "volume based" fairness between members must therefore take this into account and factor it in directly. Otherwise it's just saying that a someone with a small garden behind his house has to pay the same taxes that someone who owns half the country. Even worse, the small-garden-guy who also has an allotment garden will pay twice as much as the rich land lord. In what european country would that be fair, I beg you to tell me ? Or perhaps the RIPE returned in the 18th century ? Not only is your volume-based proposal unfair, it is also quite dangerous. It lead people to think that the RIPE has moved to some kind of nobility/gentry/aristocratic way of seeing things, only defending the big LIR interests when most of the LIRs financing the RIPE are small LIRs (or am I mistaken here ?). It also means that we, small LIRs, have to go buy legacy space if we want to ensure our business stability in the future, because one day, the option to "keep the current plan" might not be there. Given the recent examples of own *un*understanding the RIPE legal department have become over time, I can understand that some LIRs might feel threatened in their business and their future by such poor proposals. I would like to share my opinion on a plan that would be more consensual and fair without being so disruptive : From my point of view, volume-based fees should count only for a small amount of the membership fees, at least at first, because : - we should leave time for the members to adapt to this change, it just wouldn't be fair either to them to disrupt the current model in less than a year's time - causing large unused ressources holders to return large pools of IPv4 won't be beneficial for the move to IPv6 - we don't want to scare away legacy ressource holders and prevent/hinder the move to ROA/RPKI We could start with something as low as €0.01 per /24 of IPv4 and announce that we will increase it each year (€0.02 then 0.05 then 0.10 then ....), letting us time to see if it suits the majority. Even for de.telekom, it would only be 1132€ volume based fees per year, which is quite literaly a drop in the ocean for them (and will still be in 4 years at €0.10 per /24 IPv4). And yet it would set things moving in a "more fairness between members" direction. IPv6 should IMO follow the same rules. /32 ? /29 ? /48 ? I have no idea, these are too big numbers/factors to get a proper idea. Perhaps /32 at €0.001 ? Other allocated ressources/DB entries (ASN, ORG, etc..) should IMO be considered part of the membership, with limit at a certain amount per LIR in order to prevent abuse. Prices are just examples that I thought were reasonable, but can be changed/adjusted, you get the idea. Regarding your other arguments : You say that you are not willing to indicate that IP is a sellable product, yet you provide links to the brokers on the RIPE website : https://www.ripe.net/manage-ips-and-asns/resource-transfers-and-mergers/brok... Not only do they (re)sell IPv4 but they also make money from it and get (free?) referal service from the RIPE. You imply that RIPE costs are directly related to the numer of database entries. Do you have some backing numbers on the amount of requests ? Whois for example ? I guess you're likely to receive on average 256 more whois request on a /16 that on a /24, even if the /16 is a single entry in the database. Maybe small LIRs "cost" more on support/training and big LIRs "cost" more on database requests ? Seing the actual numbers would perhaps help calm things down, I think. Best regards, Arnaud Le 2019-04-18 18:14, Christian Kaufmann a écrit :
Dear members,
First of all, I'd like to thank you for the feedback we received from everyone so far, and special thanks to the people who gave some more context and explanation. Trying to arrive at a charging scheme that will please everyone is not an easy task.
The reason the board proposes two charging schemes is because some members requested a real alternative and difference to the existing "one LIR account-one fee" version we have right now and that is more volume based.
This came up previously in the charging scheme task force discussions but also from individual members via emails or through personal contact. Nigel and I promised at the last two GMs that we would present a new one before the May GM this year.
So what was the board's thinking in proposing these two models?
Firstly, many people like the existing model and the board believes that it covers the spirit of what some members want by maintaining the financial stability of the NCC while keeping fairness and equality in mind. The board also does not want a price per IP model because this would have tax implications (the RIPE NCC does not sell IP addresses and the charging scheme should reflect this) and we feel it is not in keeping with the idea of a membership association.
We have also found in the past that having more than two options does not work well from a voting perspective. This would add considerable complexity to the voting in which resolutions must be approved by more than 50% of voters to be adopted.
The second charging scheme option is one that the board believes offers a real alternative while staying away from the price per IP aspect.
The board's thinking in making the Option B proposal is that every registry entry consumes resources such as customer support time, database memory, registration time, etc. regardless of the size of the allocation. A /24 and a /12 are not so different in this regard so we see this as fair in terms of the work required by the RIPE NCC to maintain the registry. The reason we suggest to charge IPv4 and IPv6 in the same way follows the same logic - there is no tax designed to move people to IPv6. We did not want to have a political, policy-driven charging scheme because the board believes this is the work of community rather than for the board or membership to decide on.
I understand that the "volume-based" description could be seen as misleading and I apologise for the misunderstanding here. The proposed model is based on registrations and not per IP as we do not want to indicate that IP is a sellable product but rather the RIPE NCC should charge members for the registry services it provides.
The new charging scheme was also not proposed so that the RIPE NCC could make more money - it takes the current budget and calculates backwards to achieve the amount required to run the RIPE NCC. It is just a different model to share the current cost among members.
Despite concerns that were raised on this list, the board took the request of some members to propose a new model very seriously and we spent quite some time to discuss and model the current scenario by trying to be as fair as possible and sticking with the principles of a membership organisation.
Again, we are very thankful for your input and the feedback on the two models. We will continue to monitor discussions and we will of course present on the Charging Scheme 2020 at the upcoming GM. We encourage you to register your vote so you can have the final say on the two proposals.
Best regards,
Christian Kaufmann RIPE NCC Executive Board Chairman
_______________________________________________ members-discuss mailing list members-discuss@ripe.net https://lists.ripe.net/mailman/listinfo/members-discuss Unsubscribe: https://lists.ripe.net/mailman/options/members-discuss/arnaud.brand--ripencc...