On Thu, 22 Jan 2015, Dave Wilson wrote:
Hello James,
On 22/01/2015 12:38, Kennedy, James wrote:
Wouldn't relaxing the text (as initially suggested) to require the LIR to have
*any* form of IPv6, rather than removing it altogether, be more beneficial to
general IPv6 adoption?
I fear having no IPv6 requirement at all may encourage the LIR to look into
alternatives, such as NAT or the transfer market.
Honestly, I don't think it would, no. We'd get some anecdotes like we
already have, but nothing systematic.
This proposal looks harsh because all a proposal can do is changing the
policy text and, taken on its own, that appears to be a negative change.
Might as well do something, right? But "something is better than
nothing" is just not effective; in fact the thrust of my talk at RIPE68
is that it's worse than useless.
Here's one example for this particular case: the number of v6
assignments is not currently a useful measure of interest in IPv6,
because it's polluted by assignments to people who only applied because
they had to do so to get a /22.
As Gert said, the RIPE NCC is asked to send very clear signals about
IPv6 to future applicants. That's something that doesn't belong in the
policy text, but is absolutely pertinent to this proposal, and my
feeling is that it'll be an overall improvement. More importantly, it's
something that can be worked on over time so that we do end up with a
systematic improvement.
(So to be clear: I support the policy as is in last call.)
+1
Cheers,
Daniel
_________________________________________________________________________________
Daniel Stolpe Tel: 08 - 688 11 81
[email protected]
Resilans AB Fax: 08 - 55 00 21 63
http://www.resilans.se/
Box 45 094
556741-1193
104 30 Stockholm
