Jay,
On Aug 21, 2011, at 10:57 AM, Jay Daley wrote:
Bottom-up is qualitatively different from top-down, no matter how much the intended outcomes sound similar.
I'd argue the end result behavior is what matters, not how you get there.
While true, this is largely irrelevant since it isn't going to change.
I disagree. I think this is a matter of leadership.
What exactly would you have the RIRs in general or APNIC in particular do?
I would strongly argue that the way forward is reform of RIR participation to broaden it out, which will enable policy solutions to finally emerge, rather than abandoning the bottom-up policy process.
In theory, this would be nice. In practice, commercial competition is happening now. How long is this broadening (even assuming it is possible) going to take? How long will reforming the system take after this?
If this were the domain name industry, with all its competing interests getting involved, then we would see a much more radical policy for IPv4 such as:
If more competing interests are involved (which I presume would be a result of 'broader participation'), I would think address policy making would become even more glacial than it is now and 'radical' would be the exact opposite of the consensus-based outcome. However:
- all pre-RIR allocations given X years to sign up to RIRs or they would lose their allocations; enforced by exclusion from RPKI
Requires centralized control. How's that going to happen? And besides, the RPKI folk will tell you that use of RPKI is entirely voluntary so this threat is meaningless (if you believe them).
- a new pricing structure that increases exponentially as the size of the allocation increases
Why do you assume this (or something similar) won't occur, regardless of RIR involvement?
- 'use it or lose it' with strict enforcement
Requires centralized control. How's that going to happen? And define "use".
- much more detailed WHOIS requirements with strict enforcement
Requires centralized control. How's that going to happen? Hasn't happened in the DNS world after more than a decade and numerous attempts. Why do you assume things will be different in the addressing world?
See a pattern?
Well, yes, but probably not the one you intended :-). With the exception of pricing fluctuations, all of the radical policies you suggest seem to require some form of centralized control. What is occurring is actually the opposite: an increase in decentralization, at least in the functions associated with reuse of allocated address space and maintenance of registration data. I'd argue this decentralization is occurring because of the existing RIR system's inability to adjust in a timely fashion to an environment radically different than the one in which the original rules were defined and that this decentralization will only increase, particularly as IPv4 becomes more scarce. You appear to be arguing that a broadening of participation will fix that. I'm ... skeptical.
Regards,
-drc