My thoughts about the proposed Charging Scheme Models A and B (2027–2031)
Dear colleagues, though I'm a few weeks too late I want to wish you a Happy New Year! As far as I understand there are currently two proposed models for the RIPE NCC charging scheme for the period 2027–2031. Model A is category based where Model B is formula based. Here are my thoughts: 1. I do not support Model B for the simple reason that I think a formula based model is too complicated and therefore intransparent. Members should be able to immediately see (on paper, not being required to use a calculator) how many fees they are required to pay. 2. I support keeping the flat fee model (only) in case the fees are not increased. If the flat fee is increased all few years due to higher costs, I do not support it any more because it will be a financial burden for smaller LIRs. For instance, I think it would be a imposition for smaller LIRs to pay €3,000 per year one day. The hosting business is not getting easier—it's getting harder. 3. If a category based scheme is used, I think many categories are better than less categories. The current Model A has 21 categories (counting the €500 category in case the LIR has no resources at all) which is good. However, I think Model A is extremely unfair: 4. First, from what I can see, the largest single allocation is IPv4 /10 and there is not a single IPv4 /8 or even /9 allocation. So it is doubtful that even the largest LIRs will fall into the uppermost categories. That means there are categories which from the beginning on will never be relevant. 5. Also there is a big gap between the smallest and the largest LIRs. That means that the vast majority of fees continue to be paid by the smallest, small and medium-sized LIRs whereas the large and largest LIRs still contribute only a small amount (in relative terms). For instance, I own about an IPv4 /19 equivalent (= 32 x /24s). With this new model, my fees are increased from €1,800 to €3,080 which is quite much. If I would have an IPv4 /13 equivalent I would not be a sole trader but own a large company. I would not have 32 x /24s but 2,048 x /24s which is 64 times as much—but with this new model I would only pay €6,488. And if I would have an IPv4 /10, which is equivalent to 16,384 (sixteen thousand) IPv4 /24s I would only pay €8,503 though I would have 512 times as much space. This is not fair. And there is no simple middle course. If the stated goal is to make the scheme more fair then the larger and largest LIRs unfortunately need to pay way more because they are the only group which is not fee sensitive. It makes (almost) no difference for them whereas it does make a huge difference for us. Thank you for your time. --Kayo
I suspect true "fairness" (ie, where former state telcos and their huge IPv4 estates like Orange and Deutsche Telekom pay some incredibly high amount) is (counterintuitively) probably not an ideal situation, as any serious changes in such customers can have a pretty big impact on the overall income of RIPE NCC. What RIPE appears to be going for here is that there is a much more distributed income over the all of the member base, rather than a situation where they have "whale customers" who pay an extremely large amount of money. A good demo is this logarithmic graph here (which admittedly I have not done the homework to double check that it is accurate, but at a glance it does look correct) https://better-ripe.github.io/calculator2025/comparison.html Having whale customers is good for some people/businesses however it can be a bit of a liability (to put it lightly) when one of those customers decides to leave, Looking at the amount of IPv4 allocated (although some of this could include legacy space which (if i remember correctly) has some billing nuances attached to it) here are the top 30 RIPE members by tag: totalIPs|accountowner | --------+------------------------------------+ 38073196|uk.microsoft | 28976719|de.telekom | 23695375|fr.telecom | 16842782|de.daimler | 16785429|uk.mod | 16323328|fr.sfr | 11615773|us.a100row | 11351910|uk.bt | 11190902|ru.rtk | 10806073|es.telefonica | 10420681|nl.surf | 9076830|uk.ntli | 8413212|fr.proxad | 7996417|it.tin | 7905085|uk.bskyb | 7847958|it.interbusiness | 7100733|fr.bouygtel | 6923294|tr.telekom | 6729240|de.arcor | 6706530|se.teliase | 6415417|de.mediaways | 6322408|se.swipnet | 5604355|it.vodafone | 5548370|sa.stc | 5513832|pl.tpsa | 5378848|uk.tm | 5334045|it.wind | 5247905|dk.teledanmark | 4810351|de.detemobil | 4685853|us.international2 | Now i'm not going to say that any of these businesses immediately look like they are "at risk" of transferring all all of their addresses to ARIN or whatever, but instead of you look at this list as a "list of companies that could conceivably pressure RIPE into doing things based on the fact that they are huge and ripe depends on them paying their bill" I think the story is a little bit more uncomfortable. I generally think that it's a good idea (no matter what you think about how RIPE spends their budget) . I would rather have a large population of companies paying the bill then a smaller pool of whale companies that could destabilise the org by leaving. I'm all up for stuff getting cheaper, but not if that causes stuff to not exist for the long term On Mon, 19 Jan 2026 at 15:10, Akayo <ripe@akayo.eu> wrote:
Dear colleagues,
though I'm a few weeks too late I want to wish you a Happy New Year!
As far as I understand there are currently two proposed models for the RIPE NCC charging scheme for the period 2027–2031. Model A is category based where Model B is formula based.
Here are my thoughts:
1. I do not support Model B for the simple reason that I think a formula based model is too complicated and therefore intransparent. Members should be able to immediately see (on paper, not being required to use a calculator) how many fees they are required to pay.
2. I support keeping the flat fee model (only) in case the fees are not increased. If the flat fee is increased all few years due to higher costs, I do not support it any more because it will be a financial burden for smaller LIRs. For instance, I think it would be a imposition for smaller LIRs to pay €3,000 per year one day. The hosting business is not getting easier—it's getting harder.
3. If a category based scheme is used, I think many categories are better than less categories. The current Model A has 21 categories (counting the €500 category in case the LIR has no resources at all) which is good.
However, I think Model A is extremely unfair:
4. First, from what I can see, the largest single allocation is IPv4 /10 and there is not a single IPv4 /8 or even /9 allocation. So it is doubtful that even the largest LIRs will fall into the uppermost categories. That means there are categories which from the beginning on will never be relevant.
5. Also there is a big gap between the smallest and the largest LIRs. That means that the vast majority of fees continue to be paid by the smallest, small and medium-sized LIRs whereas the large and largest LIRs still contribute only a small amount (in relative terms).
For instance, I own about an IPv4 /19 equivalent (= 32 x /24s). With this new model, my fees are increased from €1,800 to €3,080 which is quite much. If I would have an IPv4 /13 equivalent I would not be a sole trader but own a large company. I would not have 32 x /24s but 2,048 x /24s which is 64 times as much—but with this new model I would only pay €6,488.
And if I would have an IPv4 /10, which is equivalent to 16,384 (sixteen thousand) IPv4 /24s I would only pay €8,503 though I would have 512 times as much space.
This is not fair. And there is no simple middle course. If the stated goal is to make the scheme more fair then the larger and largest LIRs unfortunately need to pay way more because they are the only group which is not fee sensitive. It makes (almost) no difference for them whereas it does make a huge difference for us.
Thank you for your time.
--Kayo
To unsubscribe or manage your subscription, log in to the LIR Portal with your RIPE NCC Access account and go to the LIR Account page: https://my.ripe.net/#/account-details.
Scroll down to Membership Mailing Lists to update your 'members-discuss' subscription.
Having issues unsubscribing? More information about managing your subscription can be found at: https://www.ripe.net/s/members-discuss-subscription-options/
On 19/01/2026 15:32:29, "Ben Cartwright-Cox via members-discuss" <members-discuss@ripe.net> wrote:
...huge IPv4 estates like Orange and Deutsche Telekom pay some incredibly high amount) is (counterintuitively) probably not an ideal situation, as any serious changes in such customers can have a pretty big impact on the overall income of RIPE NCC.
Many reasons not to charge large orgs have been given by people who are not the large orgs. Let them defend themselves and tell us why they should pay proportionately less. I’d like to hear what the real choices are here. They may say "actually we have so many v4 we rent them out at EU1/month each so 10% to RIPE for the benefit of getting to make out like bandits on a public resource is a pretty good deal”
Having whale customers is good for some people/businesses however it can be a bit of a liability (to put it lightly) when one of those customers decides to leave,
We are only risking their current fee, not the potential larger fee, why discount ever getting a large fee and so asking other smaller orgs to pay a part of that large fee for them?
I would rather have a large population of companies paying the bill then a smaller pool of whale companies that could destabilise the org by leaving.
False dichotomy. There are not many of them, RIPE will lose more members to run out period LIRs cashing out their winnings. We are not voting on a forever scheme, if a few large paying memebers were to leave then we can change the model again to ensure the budget remains the same if we want that, and that fees per member do not become onerous. If we are voting for a forever scheme then the only safe choice is the current flat fee, it works no matter how many leave and at some point we will change the budget to keep the fee reasonable. To be clear, flat fee must remain a voting option each year.
I'm all up for stuff getting cheaper, but not if that causes stuff to not exist for the long term
I’d rather not pay more forever because I fear maybe paying a little more later. brandon
Let them defend themselves and tell us why they should pay proportionately less. I’d like to hear what the real choices are here.
The reality is most of the employees at the companies on the list I posted are not able to make statements around this kind of stuff (most of them being publicly traded or semi-government organizations), the age of where engineers are able to freely speak their mind when they employed by large companies that happened to have an internet department have mostly gone away if someone from such a company wants to go on the record on what their opinion as i'm absolutely fascinated on what it is going to be, but I don't actually believe that this going to happen because (having been in a regulated/publicly-traded organization that would have this kind of aversion to public statements) that doesn't seem to be a real big upside for a company to make a statement for something (inconsequential in the grand scheme of things) like this. --- I understand that we're not talking about changing something forever, however given that this general community has a pretty severe aversion to change (which definitely has upsides) such large changes in the building model are probably going to happen for the long run rather than the short one, especially in a situation where if we undo this the smaller cohort of the networks will get hit with a bigger bill, they are likely to be much more annoyed at than the current status quo On Mon, 19 Jan 2026 at 17:43, Brandon Butterworth <hostmaster@bogons.net> wrote:
On 19/01/2026 15:32:29, "Ben Cartwright-Cox via members-discuss" <members-discuss@ripe.net> wrote:
...huge IPv4 estates like Orange and Deutsche Telekom pay some incredibly high amount) is (counterintuitively) probably not an ideal situation, as any serious changes in such customers can have a pretty big impact on the overall income of RIPE NCC.
Many reasons not to charge large orgs have been given by people who are not the large orgs. Let them defend themselves and tell us why they should pay proportionately less. I’d like to hear what the real choices are here.
They may say "actually we have so many v4 we rent them out at EU1/month each so 10% to RIPE for the benefit of getting to make out like bandits on a public resource is a pretty good deal”
Having whale customers is good for some people/businesses however it can be a bit of a liability (to put it lightly) when one of those customers decides to leave,
We are only risking their current fee, not the potential larger fee, why discount ever getting a large fee and so asking other smaller orgs to pay a part of that large fee for them?
I would rather have a large population of companies paying the bill then a smaller pool of whale companies that could destabilise the org by leaving.
False dichotomy. There are not many of them, RIPE will lose more members to run out period LIRs cashing out their winnings.
We are not voting on a forever scheme, if a few large paying memebers were to leave then we can change the model again to ensure the budget remains the same if we want that, and that fees per member do not become onerous.
If we are voting for a forever scheme then the only safe choice is the current flat fee, it works no matter how many leave and at some point we will change the budget to keep the fee reasonable.
To be clear, flat fee must remain a voting option each year.
I'm all up for stuff getting cheaper, but not if that causes stuff to not exist for the long term
I’d rather not pay more forever because I fear maybe paying a little more later.
brandon
participants (3)
-
Akayo -
Ben Cartwright-Cox -
Brandon Butterworth