On Tuesday, 20 October 2015, remco van mook <[email protected]> wrote:
> > > On Tue, Oct 20, 2015 at 5:49 PM Elvis Daniel Velea <[email protected] > <javascript:_e(%7B%7D,'cvml','[email protected]');>> wrote: > >> Hi Remco, >> >> On 10/20/15 5:27 PM, Remco van Mook wrote: >> > Hi all, >> > >> > (no hats) >> > >> > I think this is a very bad idea*. The whole reason the final /8 policy >> looks the way it does (and is as far as I can see working *exactly* as >> intended) is so late entrants to this Internet game have a fair chance of >> establishing themselves without having to resort to commercial alternatives >> for IPv4 address space. >> We started this discussion at least one year ago. We had a few >> presentations at RIPE Meetings and there were a few discussions on this >> topic on the mailing list. It was obvious that many of the new members >> registered after 2012 need more than the default /22. > > > Doesn't everyone? There's a reason the minimum allocation size pre-runout > was never smaller than a /21. As said, the purpose of final /8 is *not* to > keep doing business as usual - those days are over and are never to return. > Adding additional discontiguous prefixes form the final /8 pool to existing > LIRs, aside from being bad engineering, does not provide a scalable > solution; at the end of the race you now have two separate /22s and as you > managed to run out of the first one, you'll run out of the second one as > well. At the same time it's one less company that is able to get their own > onramp to the IPv4 internet. > > >> Additionally, there are a couple other RIRs (APNIC, LACNIC) that have a >> similar policy and it seems to be working just fine. >> > For established LIRs, adding a trickle of additional address space >> probably won’t make a jot of a difference for their business and is likely >> not going to optimise the utilisation of those final scraps. The final /22 >> is *intended* to be used as a migration tool to IPv6, and is a crucial tool >> at that. I consider it a Very Good Thing Indeed that this region had the >> foresight that IPv6 won’t happen overnight once IPv4 runs out** and as long >> as we’re still talking about IPv6 adoption and not IPv4 deprecation, that >> tool should be available for as many organisations as possible. >> Well, a /22 every 18 months may be helpful to those that need to work >> with only 1024 IPs.. That was the signal I received over the past months. >> > > A /22 every 18 months will give 'newish' LIRs (but not the 'newest') a > single extra /22. Come round 2, there will be none to be had. To me, this > looks like an extra final cigarette when you resolved to stop smoking. The > policy text was and is unambiguous, you knew you were only getting the one > allocation, there should be no surprises there. Stop smoking already. > > >> > >> > Finally, introducing this kind of change in policy at this point in >> time could well be argued as being anti-competitive and would end us up in >> a legal mess. >> Can you please detail how it would be argued as being anti-competitive? >> This would apply to *all* members and each member would have access to >> it, provided they have not yet transferred (parts of) the allocations >> already received. >> > > It's anti-competitive to the people who are looking to sign up in 2018 or > so. There's another word for companies that keep new entrants out, but I'm > pretty desperate to keep that word out of this discussion. Legislation > takes a dim view. > > >> I hope you understand what we want to achieve. Give a chance to those >> that have registered as LIR after Sept. 2012 to receive a *bit* more >> space from the central registry (as the prices for small allocations via >> the transfer market is really high). >> - would you agree with an other way to achieve this? If yes, please >> share your thoughts on how this proposal could be amended. >> > > Well, sure, why not. I think it's a very bad idea for a whole pile of > other reasons, but if you were to draft a policy that would allow > additional NEEDS BASED allocations to existing LIRs from address space that > gets RETURNED to the RIPE NCC that is outside the final /8 pool (so > basically, allocated pre-2012), that would sound very reasonable, fair and > good for competition. > On the other hand, why not over is over, even a bit v4 is wasted in the end in a world of v6, who cares? > > Best, > > Remco > -- -- Kind regards. Lu
