Hi, On Sun, Mar 18, 2018 at 09:10:11PM -0300, Nicolas Antoniello wrote:

I'd really appreciate real comments on pros and cons about the proposal other from non constructive comments.

I'm fairly sure Randy tried to politely bring across the message that "we had a global registry first, and then split it up into regional IRs, because that's what made sense, and still does". Besides that, discussion on that policy proposal takes place over at the LACNIC list, not here. Gert Doering -- APWG chair -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279

Hi, On Mon, Mar 19, 2018 at 09:08:56AM +0000, Randy Bush wrote:

and i am less sure i want to strongly assert that the split still makes sense. do the regional empires really improve the operators' life?

For an regional European operator, I certainly value being able to talk to someone in my timezone, talk to someone who understands the way Germans do business, etc. - read: the interaction with a regional IR (RIPE NCC) works well for us. "Timezone" and "cultural understanding" seem to be the important bits. Where the *numbers* come from, in the end, does not matter much - it could be a global pool everyone draws from (well, it is, but we pretend it isn't), but the local interaction is important. Also, this being a policy list, being able to discuss and agree on policy inside a region is complicated enough. Getting agreement globally on *details* seems to be near impossible ("principles" sort of worked out).

and, while indeed this is being discussed in lacnic, it does affect the ripe region.

True, but since a global proposal needs to reach consensus everywhere, we can be a bit lazy and wait to see how much momentum it gains in Lacnic, before formally entering the discussion here... Gert Doering -- NetMaster -- have you enabled IPv6 on something today...? SpaceNet AG Vorstand: Sebastian v. Bomhard Joseph-Dollinger-Bogen 14 Aufsichtsratsvors.: A. Grundner-Culemann D-80807 Muenchen HRB: 136055 (AG Muenchen) Tel: +49 (0)89/32356-444 USt-IdNr.: DE813185279

Hi everybody, On 26.03.2018 2:56, Max Tulyev wrote:

If this GIR runs parallel to existing RIRs and in competition with all them - that's a very good idea, I support it.

Another one IR is good idea and should be created. GIR with RIRs - this is good idea too. GIR should be created and be independent. And members of RIR should have ability where to have there resources support. I support. Juri.

On 26/03/2018 13:06, Jim Reid wrote:

This is a remarkably bad idea. It’s also likely to be unworkable.

First, it’s not clear what problem (if any) this proposed new RIR would solve. Where’s the use case(s)? What are the requirements and why aren’t these being met by the existing RIR system? Each RIR produces address allocation policies which meet the needs of their respective communities. Have these somehow become defective? And if the regional policy in LACNIC (say) isn’t working for that part of the world, why can’t that be fixed directly without introducing this new (allegedly virtual) RIR and all the extra complexity that will create?

Second, how will policies for this new RIR be created/implemented/maintaines/decided? How will they be aligned with those of the existing RIRs? How/when does an RIR decide to use this new RIR's address resources instead of its own to handle a request from an LIR?

Third, why will addresses from this proposed new RIR be “better” than those issued by the existing RIRs? What makes these addresses “special”? Who needs them? Why?

Fourth, why would an LIR choose to pay more for these “special” addresses instead of just using the ones it’s already got from its RIR?

Fifth, this proposal has great potential for unnecessary mission creep, adding more moving parts, forum shopping and so on.

All very good points. All I can really add is to amplify on point 2: Proposing the creation of a "global registry" isn't really about saying "Let's have a GIR too". That's the simple bit. The real meat of proposing a GIR is saying "let's have a new institution, that has - THIS structure - THIS funding model - THIS secretariat/support/NCC equivalent - THIS type of PDP - THIS model for who participates in the PDP (both in theory and practice) etc. Until you have a proposal (at least in outline) for what that looks like, you don't *have* a proposal at all, just a vague idea of address management by Coca-Cola*. Malcolm * classical reference. -- Malcolm Hutty | tel: +44 20 7645 3523 Head of Public Affairs | Read the LINX Public Affairs blog London Internet Exchange | http://publicaffairs.linx.net/ London Internet Exchange Ltd Monument Place, 24 Monument Street London EC3R 8AJ Company Registered in England No. 3137929 Trinity Court, Trinity Street, Peterborough PE1 1DA

Hi, Dne pondělí 26. března 2018 13:35:47 CEST, Staff napsal(a):

Hi everybody,

On 26.03.2018 2:56, Max Tulyev wrote:

...

If this GIR runs parallel to existing RIRs and in competition with all them - that's a very good idea, I support it.

Another one IR is good idea and should be created.

GIR with RIRs - this is good idea too. GIR should be created and be independent. And members of RIR should have ability where to have there resources support.

I support.

I don't think that it is such a good idea. First of all, I can see the problem of such organizations which resident in multiple RIR regions, however I do think that I can be solved by bilateral agreement between current RIRs, rather than creating "GIR" (something between IANA and current RIRs. I can also see that someone might see it as an opportunity to get yet another resources, which they cannot from current RIRs. However there is no more IPv4 in IANA pool, so we would have to talk about IPv6 only "GIR" with only 32b ASN (in contrast with LACNIC policy text). And when I look at IPv6 policies at RIPE region (at least), there are quite open-minded with their allocation size. So do we really need yet another RIR? In my opinion No. It would solve just marginal problem which does have simpler solution. The solution might be an Inter-RIR status (e.g. source: RIPE-INTER-RIR) based upon agreement between LIR and multiple RIRs (in which case the resources would be assigned/allocated from one of them). Example: 1) AfriNIC based LIR would like to operate part of its network in RIPE region 2) LIR asks AfriNIC for approval to operate outside of RIR region and provides documentation with reasoning and corresponding RIR in which region LIR would like to operate 3) AfriNIC decides if the LIR's proposal is fine. 4) AfriNIC asks RIPE: Is it OK? May that LIR in this case operate this network in your region? 4) If both RIRs agrees on LIR's proposal, the AfriNIC marks LIR's resources accordingly (like moving it to separate DB or something like operates in: RIPE) Certainly no RIR would volunteer their IPv4 pool to new "GIR" as LACNIC proposal suggest and there is no more "global" IPv4 pool available... Sincerely Martin Hunek

Hi, Some comments. It's good to have new kind of IR, not for requesting resources, but for managing them and keep things in order. Current RIRs are old enough, too bureaucratic, old school and so on. Also you don't need to take care about new GIR if you prefer to stay with your local RIR. Ability to make new IR will push on RIRs to make better services and less members fees. New IR will make more possibilities for new database management. Because current RIRs keep there databases in not optimal way because those databases keeps there structure from the time Internet was born. They are old enough, but they are too heavy to be rebuild-ed. It's a good opportunity to create something new, useful for ISP and their customers. Of couse this database should be redistributed and protected globally. Juri On 26.03.2018 15:28, Martin Huněk wrote:

Hi,

Dne pondělí 26. března 2018 13:35:47 CEST, Staff napsal(a):

...

Hi everybody,

On 26.03.2018 2:56, Max Tulyev wrote:

...

If this GIR runs parallel to existing RIRs and in competition with all them - that's a very good idea, I support it.

Another one IR is good idea and should be created.

GIR with RIRs - this is good idea too. GIR should be created and be independent. And members of RIR should have ability where to have there resources support.

I support.

I don't think that it is such a good idea. First of all, I can see the problem of such organizations which resident in multiple RIR regions, however I do think that I can be solved by bilateral agreement between current RIRs, rather than creating "GIR" (something between IANA and current RIRs.

I can also see that someone might see it as an opportunity to get yet another resources, which they cannot from current RIRs. However there is no more IPv4 in IANA pool, so we would have to talk about IPv6 only "GIR" with only 32b ASN (in contrast with LACNIC policy text). And when I look at IPv6 policies at RIPE region (at least), there are quite open-minded with their allocation size.

So do we really need yet another RIR? In my opinion No. It would solve just marginal problem which does have simpler solution.

The solution might be an Inter-RIR status (e.g. source: RIPE-INTER-RIR) based upon agreement between LIR and multiple RIRs (in which case the resources would be assigned/allocated from one of them).

Example: 1) AfriNIC based LIR would like to operate part of its network in RIPE region 2) LIR asks AfriNIC for approval to operate outside of RIR region and provides documentation with reasoning and corresponding RIR in which region LIR would like to operate 3) AfriNIC decides if the LIR's proposal is fine. 4) AfriNIC asks RIPE: Is it OK? May that LIR in this case operate this network in your region? 4) If both RIRs agrees on LIR's proposal, the AfriNIC marks LIR's resources accordingly (like moving it to separate DB or something like operates in: RIPE)

Certainly no RIR would volunteer their IPv4 pool to new "GIR" as LACNIC proposal suggest and there is no more "global" IPv4 pool available...

Sincerely Martin Hunek