[address-policy-wg] 2013-03: Good enough?

Dear all,

Gert has prompted me to reply to the proposal on the table, to confirm
whether it meets my earlier objections.

Firstly, I'd like to apologise again for not paying closer attention
sooner. My thanks again to the co-chairs and to Tore for their hard work
and diligence in trying to achieve a consensus.

This proposal has gone through three drafts, and I doubt anyone has much
appetite for another. Since it seems there is not going to be unanimity,
the co-chairs have the unenviable job of determining whether all the
reasoned objections have been fully addressed, or whether consensus is
simply not available on this issue at the current time.

So as someone (Randy?) said weeks ago, the question isn't whether I like
this proposal, the question is whether it's good enough, that is, not so
seriously flawed as to justify standing in the way of a clear majority
so as to prevent a rough consensus?

The TL;DR is that I am now willing to withdraw my objection, but would
ask that you change the title to remove the words "No need", agree to
changes as to how this is presented externally, and to support further
work as part of a new PDP.

If that is agreeable (and I hope it can be done without going to Version
4, as I am not suggesting any change to the policy itself from Version
3), I would agree my points have been fully addressed and need not stand
in the way of announcing a finding of consensus.

In case others should be interested, I shall set out more fully my
reasons for changing my position before elaborating what I hope will be
a compromise everyone could support.

It is true that I have never been persuaded by the arguments advanced
for this proposal. Sylvain did a demolition job on them (2013-09-20
13:19 UTC+2), and I share many of the same views. However, I don't work
at the coalface of allocation requests, as many here who support this
policy do. I don't think my being un-persuaded on this is good enough
reason to block the policy.

Much of what McTim wrote I also agree with, especially this:
> Needs based distribution has been a cornerstone of the RIR system
> for the last 2 decades or more.  It has worked remarkably well, and I
> see no need to jettison it now just because there are fewer resources
> to distribute.  In fact, I see a greater need for it now!  I expect
> we will have to agree to disagree on this.

However he also said:
> The primary issue there is incompatibility with other regional
> transfer policies.

Gert replied that inter-regional compatibility was a closed issue:
> I do have the prerogative to ignore certain topics voiced as reason 
> for opposition to a proposal if I consider them suitably addressed,
> and there is sufficient support for a proposal otherwise.
> 
> This is what I did: I consider this point to be suitably addressed, 
> given the very specific fact that we do not have a cross-RIR
> transfer policy today, and this community has not shown interest in
> working on one.

I am in agreement with Gert on this aspect. Given that we do not have a
cross-RIR transfer policy, are not going to get one imminently, and that
reverting 2013-03 would be possible as part of the process of creating
one, I don't think incompatibility with ARIN is sufficient reason to say No.

That leaves the issues that were always the heart of my objection, the
political (and possibly legal) impact.

My reading of the original NCC Impact Assessment was that the
competition law issues were a concern, but not a grave risk: much more
serious was the political impact on the credibility of the RIR system of
throwing away all basis for a claim that the way the RIRs operate is in
the public interest.

The latest version of the Impact Analysis says

> External Relations
>
> Adoption of this policy would have a major impact on RIPE NCC
> External Relations, and would require the investment of considerable
> resources in messaging and defending/explaining this policy shift to
> stakeholders both inside and outside the technical community.
> Additionally, these activities would need to target not only
> stakeholders within the RIPE NCC service region, but also those in
> other RIR communities.

That's a pretty serious warning, and I think it corroborates everything
I have said on this subject.

On the previous draft, what concerned me most was the appearance of the
RIPE Community abandoning all claim to regulate allocation on the basis
of need. This would be tantamount to giving up the RIPE Community's
right to create new policies in future on the basis that market effects
are causing severe operational problems. Not being able to create new
policies to address new problems would be politically unacceptable: if
the RIPE Community would not do it, then governments would have to step in.

Version 3 provides the compromise Tore offered me: make dealing with
those future issues a problem for another day; "kick the can down the
road". Do not let those future problems stand in the way of this
proposal which is (by assertion) needed now - but retain a placeholder
so that we do not close the door on revisiting this issue in the future,
if we need to do so.

That placeholder is the new assertion in the policy of "fairness", an
undefined term. At the moment this either stands in as a loose and
probably unenforceable token for the current normative standards, or it
has no practical effect on everyday operations at all. But it does have
one key virtue: it asserts the community's right to adjust the current
policy on the basis of protecting what the community, by consensus,
agrees to be fair. That may turn out to be crucial, if we need to
respond to problems that may arise in the future.

Is that good enough?

I think it should be, if we are willing to work together.


So I propose the following compromise:

1. 2013-03 be retitled to remove "no need" from the title. Those words
are highly detrimental to the NCC's External Relations work, and add
nothing useful.

2. No other changes are made to the proposal itself.

3. The accompanying notes, which are not part of the proposal itself but
which are a part of the external messaging, should be re-written.
The current version of these notes unhelpfully emphasises the removal of
the requirement to demonstrate need, as if need were becoming
irrelevant. They should describe the effect of 2013-03, which is to
remove the *upfront documentation* of need (a form of ex ante
regulation), while retaining the concept of fairness, as defined by the
community, as a goal of allocation policy. It should be emphasised that
the community retains the right to introduce new policies in the future,
if deemed necessary, to ensure that this deregulatory initiative does
not bring about unfairness.

4. AP-WG should (later, but soon) consider a new policy, as part of a
new PDP, that asserts that the fairness goal makes it an objective to
maximise the availability of IP addresses for use, and asserts as a
community view that it is not fair to artificially restrict supply by
hoarding for speculative purposes.

The proposal might stop at that, and we could wait until there was
evidence of hoarding to create any enforcement mechanism. Alternatively,
we could consider introducing an ex post rather than ex ante
enforcement. i.e. where someone would have to complain that a given LIR
was hoarding before anything could be done, and the NCC could
investigate and if necessary that LIR (and only that LIR) would have to
justify their need. This would be quite different to the current
situation, where every LIR must prove their need up front.

Having in place a policy like that outlined in (4) would in my view
substantially mitigate the serious political risk referred to in the
Impact Analysis, while at the same time achieving the aim of reducing
mostly-unnecessary bureaucracy that is the goal of 2013-03.

Doing (4) as a new PDP would take it off the critical path for 2013-03,
and would also allow the principles expressed therein to be expressed as
an interpretive guide to other policies e.g. 2007-08, if thought
appropriate.

Of course, it is always open to any member of the community to draft a
new proposal for a new PDP. What I am asking is that the proposers of
2013-03 support that process, as being an appropriate way of ensuring
the reduction in application bureaucracy in 2013-03 does not have the
effect of undermining the core values the RIPE community has upheld for
decades, rather than criticising such a proposal as being a reversal of
"no need".

Version 3 of 2013-03 would allow us to move in that direction, if the
presentation of how it is introduced is cleaned up. So if these steps
are taken, I am willing to accept that my objection has been fully
answered and need not stand in the way of a finding of rough consensus.

Malcolm.
-- 
            Malcolm Hutty | tel: +44 20 7645 3523
   Head of Public Affairs | Read the LINX Public Affairs blog
 London Internet Exchange | http://publicaffairs.linx.net/

                 London Internet Exchange Ltd
           21-27 St Thomas Street, London SE1 9RY

         Company Registered in England No. 3137929
       Trinity Court, Trinity Street, Peterborough PE1 1DA