Re: [sig-policy] IPv4 countdown policy proposal
David Conrad wrote:
Worse, I would imagine such a policy would generate a 2-year rush
from folks who will try to justify as much address space as they can.
Although perhaps I'm too cynical.
I don't think we should underestimate the chaos that might ensue in the
immediate pre exhaustion rush or it's aftermath. Inappropriate trading
of address space, hijacking of space, political intervention are all
possibilities with the possibility of fracturing the
internets universal connectivity in a way alternative dns roots could
never achieve.
An alternative approach would be to tie the required justification
for address space to the amount of free space available. As the free
pool gets smaller, the justification requirements (e.g., utilization
density, support for IPv6, etc.) increase.
IMHO we need a plan to mitigate the worst effects of IPv4 exhaustion but
the proposal as it stands is, to paraphrase David 'full speed ahead'
without even reorganising the deck chairs around the pool.
The pressure to migrate to IPv6 need not be solely commercial for
example the artificial (regulatory) 'drop dead' dates for the analogue
cell phone network in Australia or as I recall the VHF TV network in the
UK are excellent examples.
Simple RIR policies could, for example, mandate the deployment of dual
stack customer premises equipment rather than the current IPv4 only kit
that is installed pretty much everywhere in user land. (I do rather
shudder at the though of end users trying to configure them though)
The good news is that we have at least five years. The bad news is that
achieving consensus and taking steps to implement it will take at least
that long.
We need to do more, try harder and start now. Or of course we could
hoard space and sell it for heaps later :)
--
Robert Gray
bob at brockhurst dot co dot nz