[members-discuss] Input from Membership on RIPE NCC Charging Scheme Model
- Previous message (by thread): [members-discuss] Input from Membership on RIPE NCC Charging Scheme Model
- Next message (by thread): [members-discuss] Input from Membership on RIPE NCC Charging Scheme Model
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Daniel Pearson
daniel at privatesystems.net
Fri Sep 23 15:27:46 CEST 2016
Robin, There is one big question. How are you going to prove, that my network, which you won't have an inside view on, is indeed dual stacked. I can request an IPv6 block, I can tell you I'm using it but if you go and try to prove that you would end up over stepping the bounds as RIPE. Daniel~ On 09/23/2016 08:22 AM, Robin Johansson wrote: > Hi, > > No, not really. > The legal/financial perspective has been that RIPE is a non-for-profit > organisation. > This proposal doesn't imply that the aggregated membership fees should > exceed the running costs for the oranisation. Only how they are > distributed among the members. > > Much of the discussion has been that it's unfair that new LIRs with > only a /22 ipv4 have to pay as much as all others. > A new LIR is in a good position to dualstack every single device from > the start, meaning that if they can end up with a lower fee than the > older LIRs, who have millions of devices to dualstack before they > qualify for the lower fee. > > The real difficulty would be how to measure the actual distribution. > > It likely that this model doesn't cause a much larger cost for the > large players than today, but it also means that a new provider that > choose to offer ipv6-only or dualstacked services get away at a much > lower cost. > > It makes more sense to promote the adoption of todays technology, > rather than clinging on to things of the past (IPv4) that can't be > changed enough to make a difference anyhow. > > Example of weights > singelstacked ipv6, factor 1 > dualstacked, factor 10 > singlestacked ipv4, factor 1000 > > Over time the amount of singlestacked ipv4 LIRs will go away, which > means that we slowly move back towards the same equal fee structure we > have today. > > /Robin > > > On Fri, 2016-09-23 at 07:27 -0500, Daniel Pearson wrote: >> Robin, >> >> Not to be rude, but we've already explained why this is not possible >> for RIPE to do several times in this thread both from a financial and >> legal perspective. >> >> Daniel~ >> >> On 09/23/2016 05:30 AM, Robin Johansson wrote: >>> Hi, >>> >>> If we are to base membership fees on resources then the only way >>> that makes sense today is to make it really expensive if you're not >>> giving your subscribers ipv6 addresses. >>> >>> Could even have it with multiple tiers >>> majority of subs singlestacked ipv4: really expensive >>> majority dualstacked: fairly cheap >>> majority of subs singlestacked ipv6: really cheap >>> >>> This makes it very easy for all the new "small" LIRs to reach the >>> fairly cheap fee, as they don't have a lot of subscribers to >>> dualstack. And it gives incentive for every LIR to at least >>> dualstack, maybe move away from ipv4 all together. Also to ensure >>> that their subscriber base have modern equipment capable of handling >>> ipv6. >>> >>> The final /22 ipv4 is enough to serve huge numbers of eyeball >>> subscribers, if used wisely for supporting services and nat64 pools >>> (or similar technology). >>> And as more and more services get ipv6 the number of subscribers >>> served through those pools can be increased even further. >>> >>> /Robin >>> >>> On Thu, 2016-09-22 at 18:51 +0200, Tim Armstrong wrote: >>>> >>>> Rather than bickering over the last scraps of IPv4, saving smaller >>>> LIRs a few hundred euros, attempting to somehow screw the older >>>> LIRs, or three at the same time. Wouldn't our time be better spent >>>> working out ways to improve end user adoption of IPv6? >>>> >>>> I'd like to propose RIPE set up a fund (summer of code style) for >>>> the implementation of native IPv6 support in open-source software >>>> (such as cloudsta k, etc) and simplifying end-user adoption. >>>> Perhaps we should even offer a free public IPv6 tunnel service for >>>> natural persons similar to the service currently offered by >>>> hurricane electric. >>>> >>>> -Tim >>>> >>>> >>>> On 22 Sep 2016 6:13 p.m., "Floris Bos" <bos at je-eigen-domein.nl >>>> <mailto:bos at je-eigen-domein.nl>> wrote: >>>>> On 09/22/2016 01:57 PM, Daniel Pearson wrote: >>>>>> I'm not saying that a discussion is bad, but I'm simply saying >>>>>> that most of the discussions are being had are not based on fact. >>>>>> >>>>>> To my knowledge RIPE doesn't have a list of members categorized >>>>>> by assignment size, so this is something that someone would need >>>>>> to parse the RIPE db for, it's all public record so it can be done. >>>>>> >>>>> >>>>> Counting all allocated IPv4 each LIR has, and converting it back >>>>> to CIDR: >>>>> >>>>> CIDR Number of LIRs >>>>> >>>>> <= /24 1 >>>>> <= /23 4 >>>>> <= /22 6051 >>>>> <= /21 1582 >>>>> <= /20 1638 >>>>> <= /19 1547 >>>>> <= /18 1040 >>>>> <= /17 709 >>>>> <= /16 386 >>>>> <= /15 293 >>>>> <= /14 134 >>>>> <= /13 110 >>>>> <= /12 80 >>>>> <= /11 64 >>>>> <= /10 25 >>>>> <= /9 14 >>>>> <= /8 6 >>>>> <= /7 2 >>>>> >>>>> IPv6 only 241 >>>>> >>>>> >>>>> If we were to take ARIN's fees as example where up to and >>>>> including /20 is less expensive than RIPE's current fees, 9276 out >>>>> of the 13686 LIRs with IPv4 would pay less. >>>>> Not just new ones... >>>>> >>>>> Total income would be similar. >>>>> >>>>> >>>>> Yours sincerely, >>>>> >>>>> Floris Bos >>>>> >>>>> >>>>> ---- >>>>> If you don't want to receive emails from the RIPE NCC members-discuss >>>>> mailing list, please log in to your LIR Portal account and go to >>>>> the general page: >>>>> https://lirportal.ripe.net/general/ >>>>> <https://lirportal.ripe.net/general/> >>>>> >>>>> Click on "Edit my LIR details", under "Subscribed Mailing Lists". >>>>> From here, you can add or remove addresses. >>>>> >>>> ---- >>>> If you don't want to receive emails from the RIPE NCC members-discuss >>>> mailing list, please log in to your LIR Portal account and go to the general page: >>>> https://lirportal.ripe.net/general/ >>>> >>>> Click on "Edit my LIR details", under "Subscribed Mailing Lists". From here, you can add or remove addresses. >>> >>> >>> ---- >>> If you don't want to receive emails from the RIPE NCC members-discuss >>> mailing list, please log in to your LIR Portal account and go to the general page: >>> https://lirportal.ripe.net/general/ >>> >>> Click on "Edit my LIR details", under "Subscribed Mailing Lists". From here, you can add or remove addresses. >> >> >> ---- >> If you don't want to receive emails from the RIPE NCC members-discuss >> mailing list, please log in to your LIR Portal account and go to the general page: >> https://lirportal.ripe.net/general/ >> >> Click on "Edit my LIR details", under "Subscribed Mailing Lists". From here, you can add or remove addresses. > > > ---- > If you don't want to receive emails from the RIPE NCC members-discuss > mailing list, please log in to your LIR Portal account and go to the general page: > https://lirportal.ripe.net/general/ > > Click on "Edit my LIR details", under "Subscribed Mailing Lists". From here, you can add or remove addresses. -------------- next part -------------- An HTML attachment was scrubbed... URL: <https://lists.ripe.net/ripe/mail/archives/members-discuss/attachments/20160923/5416e35c/attachment.html>
- Previous message (by thread): [members-discuss] Input from Membership on RIPE NCC Charging Scheme Model
- Next message (by thread): [members-discuss] Input from Membership on RIPE NCC Charging Scheme Model
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]