[arin-ppml] Change of Use and ARIN (was: Re: AFRINIC And The Stability Of The Internet Number Registry System)
owen at delong.com
Tue Sep 14 15:56:05 EDT 2021
> On Sep 14, 2021, at 10:32 , Joe Maimon <jmaimon at chl.com> wrote:
> Owen DeLong wrote:
>> Ironically, a timely move to dual stack would have meant that the IPv4 addresses needed for it were available, which would have allowed a smooth transition to v6 with new sites being deployed v6-only and old sites deprecating v4 at their leisure. Foot-dragging by those who saw no incentive to be community minded indeed took us beyond the date where this was feasible and transition mechanism hacks are the inevitable and inferior result of that foot dragging.
> Exactly. A timely migration using dual stack to avoid scarcity depended on something that was always extremely unlikely to occur, and look: it did not. This is the exact nature of the failure that is the IPv6 deployment.
>>> However, the longer this drags on the greater the odds that things we dont want will happen and irrevocably alter the landscape in ways we would not have wanted.
>> Or eventually, people will see something they want that is enabled by restoring the end-to-end model and that will drive demand for IPv6 and things we do want will happen.
> Keep on hoping. Just so happens that all trends are currently in the opposite direction.
>>> Worse for your analogy, electric car success can be wholly attributed to both new technology and manufacturers finally making cars that people actually wanted, and not just for the smug value.
>> OK, so what new technology can be deployed that will drive adoption of a new addressing scheme that doesn’t suffer from the shortages built in to IPv4?
> All I would like is for any innovators in the space to stop being hampered by the same ideologues who have brought us to this sorry state of affairs.
> Mechanisms that allow IPv6 only nodes to access the entirety|majority of the actual internet would be highly advantageous at this point. The reverse would be spectacular.
Both actually exist in varying degrees.
IVI provides some capabilities, but breaks in some circumstances. I confess I haven’t made thorough investigation of these.
NPT46 works in most cases, but does have some failure modes due to limitations on the IPv4 side of the equation.
NPT64 works in most cases, but does have some failure modes due to limitations on the IPv4 side of the equation.
DS-Lite has some features for addressing the issue, but also has some interesting failure modes. As i understand it, though I lack personal experience, again due to limitations in the v4 portion of the process.
Do you see the recurring theme in the failure modes?
>>> If I havent gotten the point across that this is what IPv6 actually needs to be successful any time soon, there is little reason for continuing this conversation.
>> You’ve done a great job of pointing out everything you think is wrong… You’ve done nothing to offer any positive solutions to any of the problems raised for either protocol.
> And look how difficult it has been. You want I should do everything??
If you abandon the former in favor of useful efforts towards the latter, I’d be fine, actually. I don’t want you to do everything, but doing something useful instead of just whining would be a pleasant change of pace.
> Before solutions can be offered that will actually be accepted, the ideology based opposition must be brought around and convinced and thereby silenced. We know this because we have seen this movie before.
Offer a useful alternative and my ideology is subject to improvement. Offering no useful alternative as you have done so far does nothing to shift my ideology.
I have lots of complaints about the flaws in capitalism. Since I don’t have a useful or superior alternative to offer, I don’t waste a lot of time whining about them.
> Many solutions and proposals have been offered by many people much more qualified than myself to do so, and were brought to the rocks due to dual stack and anti-nat religion.
Could this be due to the fact that things which merely extend the pain of address scarcity are not viewed as solutions by those of us who prefer an end-to-end model?
Creating more pain and/or prolonging existing pain in order to avoid a better end state is not viewed by the people you appear to want to convince as a solution.
>>> The goal should not be eventual IPv6 adoption by attrition. The goal should be meaningful solutions to actual issues, now and real soon. Hopefully sustainable ones, but the here and now means something to people living in it.
>> IPv6 adoption whether by attrition or other means is a solution to the actual issue of address shortage. Ideally, getting others to recognize this and move forward will provide a sustainable solution sooner rather than later, but the lengths people will go to in order to avoid change are impressive.
> Its a solution that doesnt solve anything anytime soon, which means more ongoing failure, issues and collateral damage.
How soon this finally happens or doesn’t is up to the very laggards you are championing.
> The lengths people will go to ignore change thats not very relevant to them are not impressive, just predictable.
The lengths people will go to to pretend that relevant change isn’t because they find it uncomfortable in some way are impressive. They might also be predictable.
More information about the ARIN-PPML