On Jul 13, 2012, at 9:40 PM, Nigel Titley wrote:
Andrea, in the previous round of discussions we said why we can't use an "n euros per address model".
To re-iterate the argument, if we are seen to be "selling" IP addresses by the Dutch tax authorities then we lose our special tax status. This will immediately cause a rise in the cost of running the RIPE as we will be liable for Dutch corporation tax. Up until now the membership hasn't wanted this.
I am sorry I was not here in the previous discussions. I understand your point, but I personally see RIPE and its role as something that goes beyond a "company", as a matter of fact RIPE is something whose policies go far beyond the Dutch laws … and even though I do not have to propose a "technical" solution I am still convinced that the technicalities of the dutch tax system should not impact something so important (I understand that this is easy to state as a principle, but hard to apply as a matter of fact). The principle in my mind is quite simple: IPv4 is a limited resource, everyone would love to have a /8 allocated tomorrow morning, but that is simply not possible. With all limited (shared) resources in the world the only way to limit their misuse is a magic and hated word: taxes. This is what happens to use of water, to pollution, to a lot of other things: the more you use/pollute/occupy the more taxes you pay. Now I understand that it is probably impossible for legal reasons to put it in the form "pay 0.1 euro per IP", but still I think it would make sense to push toward a "the more you use the more you pay" model, and not toward "everyone pays the same" scheme. I see that this proposal goes toward the "everyone pays the same" schema. My opinion is that RIPE should go in the opposite direction. Since I represent a "medium" LIR (we have about 16K addresses allocated) the two models do not change much on our side. By the way I can say that the RIPE fee is a marginal cost in our budget (you bet a few gigabits of transit on Tier1 networks costs a couple of orders of magnitude more). So there is no personal/institutional bias on this. I would say that a model like "up to a /21 total allocation you are small and pay 1; up to a /18 you are medium and pay 8; up to a /15 you are large and pay 64; up to /12 you are fat and pay 512…" would be a fair approach, technically not different from the preexisting one and pushing toward resource saving and evolution (knock knock… time to shift to IPv6??? or want to pay 500k euros??). Really, I suppose these are known and already discussed arguments, I apologize if being the last coming here I reiterate them… but still, that's what my opinion is. Regards, A.