[address-policy-wg] the post-mortem on 2008-09

Jim,

this message of yours explains for me that you have not really
understood why or what people are having issues with, with these
issues.

On Thu, Jul 28, 2011 at 2:15 AM, Jim Reid <jim at rfc1035.com> wrote:
> On 27 Jul 2011, at 22:17, niels=apwg at bakker.net wrote:
>
>> There has been a proper debate.  It's lasted three years.
>
> Well IMO, any debate lasting *that* long cannot be called "proper". A more
> honest description might be "ivory-towered" or "defective". There is
> something fundamentally wrong if we can't get a policy done in 3 years(!)
> and then have what appeared to be a consensus come off the rails at the very
> last moment. We, the RIPE community, should hang our heads in shame. Imagine
> the derision we'd rightly heap on other policy-making bodies if they had
> produced this outcome. And we all know a few of them.
>
> Please note I am not criticising the people who raised those last-minute
> objections at all. [Though it's a pity they didn't engage much earlier.] I'm
> actually relieved they intervened while the opportunity was still there.
> This had to be more preferable than declare a consensus, implement the
> policy and then have serious objections emerge. Though I admit both options
> are unpleasant. One's just worse than the other.
>
>> Trying to sneak it in via the back door of the AGM doesn't sound like a
>> great strategy to me.
>
> That's grossly unfair Neils. Nigel clearly asked for a documented decision.

The documented decision is there is no consensus on the current
proposals (and beyond).

If you want the RIPE NCC to cheat its policy process and "remove
consensus from the table", by going for a simple majority vote
instead, the RIPE NCC's policy process becomes void; a farce, and
clearly a waste of time. The bottom-up decision making process of the
RIR system will have failed, but not for the reasons you mention, but
for a complete top-down policy overrun.

Make no mistake, pushing such a (de)vast(ating) change to the Internet
architecture as we know today such as resource certification/RPKI/SIDR
through while completely ignoring any outcome from *any* PDP arm of
the NCC, is a great failure of the PDP in that it is indeed useless.


> In the absence of viable alternatives, the NCC AGM seems the obvious forum
> for that decision. Nigel just asked that the membership should have a vote
> on whether the NCC continues doing what's in the activity plan or stops. He
> was very careful not to say what that decision should be. Or even what the
> NCC membership should vote on because,  strictly speaking, he hasn't
> proposed a resolution for May's AGM.
>
> It should be patently clear RIPE cannot take a decision about address
> certification any time soon, if ever. So the NCC membership seems the best
> (or least worst) choice as a suitable forum in our service region that could
> actually take a decision on this issue, whatever that may be.

It is becoming increasingly apparent to me that these changes are so
large that their implementations are perhaps better left to the normal
democratically based law-making processes, had it not been for the
complications, let's say, of the union.
These architectural changes are so severe I'm starting to doubt we
really have the mandate to make them, especially given how the
Internet nowadays is so much "serious freedo^wbusiness".

> It's simply unacceptable to collectively shrug our shoulders at what has
> happened and wish the wreckage to vanish all by itself. For one thing, we
> have a duty to those in the community who have not followed the detail of
> 2008-08. They deserve an answer.

Isn't the answer pretty simple? "The proposed address certification
prototype (technology) has not passed the NCC PDP quality assurance.
There are serious issues with it that have not been addressed."

> So do our friends at the other RIRs.
> Uncertainty about address certification in our service region has an impact
> on them and their communities. There are also further global impacts. It
> would not surprise me if governments who are supportive of the RIR system
> and its bottom-up policy making processes take a rather dim view of what's
> happened too. There's a nasty question that needs resolving and soon: "how
> can you spend 3 years debating an important policy, letting if implode at
> the last minute and then just walk away?". It's these sorts of risks that
> have to be mitigated. The NCC will have a key role in that risk mitigation
> effort.

I do not see and have not seen your windmills.

> So with that context in mind, what do we do now?

Address the issues raised, or we happily do nothing at all?  Isn't
that how we normally achieve consensus? Isn't that how we have been
governing ourselves, bottom-up, previously?

> For some definition of
> "we". I think Nigel's suggestion is not just sensible, it has to be the next
> best (or least worst) option. Feel free to make better suggestions...

Clearly you rank from bad to good, the "no" to "yes" of 2008-08 and beyond.
So, since you repeat the same question, I will repeat the same answer:
the PDP did make a decision, it said "no, no consensus can be reached
on the current prototypes, come again with the issues better
addressed".

> We'll look really dysfunctional if we let address certification continue as
> our very own long-running and real-world version of Schrodinger's famous
> thought experiment.

>From your biased point of view, perhaps.  From my biased point of
view, not so much. Instead, we're looking pretty responsible, IMVHO in
the same general manner like the Norwegian government has been saying
they will fight terror with more democracy and openness rather than do
what the US did.

Best Regards,
Martin