On 14 Apr 2016, at 17:23, Aled Morris <aled.w.morris@googlemail.com> wrote:
The other objection (Jim) seems to be "we should be all-out promoting IPv6”
I said no such thing. There are a number of ways for organisations to deal with IPv4 exhaustion. These include (but are not limited to) IPv6 deployment; closing your eyes, sticking your fingers in your ears and pretending the problem does not exist; NAT; ALG; acquiring addresses from the secondary market; buying an address holder with lots of unused space; etc, etc. IMO IPv6 deployment seems to be the least worst of these choices but whatever choice someone makes is up to them. After all, they should know which option works best for their network. YMMV. My objections to 2015-05 are as follows: 1) It will deplete the remaining pool of IPv4 space at a faster rate than is sensible or reasonable. 2) It will disadvantage new entrants who need IPv4 once the RIRs have run out sooner than they should have done. 3) It allows LIRs to continue to ignore the IPv4 run-out because they could just keep going back to the RIR and get yet another IPv4 allocation. 4) It disadvantages existing LIRs who have already taken action to address IPv4 exhaustion. 5) It enables new and unwelcome ways to scam address space and/or compromise the integrity of the RIPE database. 6) There’s no clear problem statement, let alone an explanation how this proposal solves whatever that problem might be or why other solutions are unsuitable. “Some LIRs want to grow their networks and ignore the IPv4 run-out” is not a sound problem statement IMO. 7) It will discourage address recycling. Why return unused space to IANA or the RIR if they’re just going to give it away to LIRs who can’t/won’t take proper steps to deal with the IPv4 run-out? There’s no mention of IPv6 in the above list.
recent LIRs are disadvantaged against established companies with large pools of IPv4 to fall back on.
Tough. That was then. This is now. The circumstances an policies that prevailed 10-20-30 years ago don’t apply today. We can only do the best (or least-worst) job of allocating the remaining IPv4 space in a fair and reasonable manner. 2015-05 does not achieve that. In fact it does the opposite.
It simply isn't possible, today, to build an ISP on an IPv6-only proposition.
Nobody was saying or even suggesting it was. At least I don’t think anyone was saying that. Similarly, it simply won’t be possible N years from now for someone to build their IPv6 only network and have connectivity to the legacy Internet unless they can get some IPv4 space.