[ppml] Proposal for the creation of a working group.

[Edward Lewis]

Taking a quote potentially out of context...

At 10:00 -0400 10/10/07, Thomas Narten wrote:

>IMO, the process for getting globally-coordinated policies adopted
>locally within each region has signficant flaws. For starters, in an
>ideal world, it takes 2 cycles in each RIR to get something done.

I agree that the process of getting a "global policy" approved has flaws.

What can be done?

For one I think policies should be labeled as global or local and 
have different process paths in each RIR.  Not that the processed 
need to be similar to each other, but speaking from general ARIN 
experience and the IPv4 Countdown example, I don't think the one 
process ARIN has is the right fit for global policies.

It's not the fault of the ARIN process, the ARIN process fits well 
for ARIN region policies.  It's because the global policy has to 
cycle through each RIR - a result of the calendar of RIR meetings 
(they are clumped together without overlapping) and that each region 
has a different perspective.  (Vividly I recall witnessing the 
progression of a policy from ARIN to RIPE to APNIC that took 
surprising turns as it was presented from one place to the next, each 
time tuned to the previous audience.)

What I think is best is for the originator (instigator) of the policy 
to quickly assemble an ad hoc group representing all of the regions. 
The goal of the inter-RIR ad hoc group formation is to try to beat 
out the right words for the proposal that will fit into each region's 
Internet dialect.  The reason for this is that the current road 
filled with cycles of present, get beaten up, fix, and present to the 
next region.  There is cycling through the RIRs for the instigator, 
there are cycles inside the RIR for consideration by the audiences.

It would be good to be able to make the process more parallel, to get 
the edits from one region fed into the proposal in other regions. 
Otherwise I don't think we will see any more global policies passed - 
assuming we need them.  (And if we do see some, they might be 
approved way too late.)

I think too that because of regional differences in policy proposals 
processes, messages from one meeting to the next (e.g., from APNIC's 
to ARIN's) gets confusing.  The ARIN straw poll is often confused for 
a vote, having a proposal on the table is sometimes used as a 
statement of endorsement when presenting to the next RIR on the 
calendar.  It's good that we have cross attendees to the meetings to 
help straighten out mistaken messages, but cross attendees (whether 
RIR staff or others) are bearing a cost to do this.

I don't have a desire to make it easier for globally policy proposals 
to succeed.  But I think the current process is too slow for the ones 
we do need and it is rather confusing.  I think instigators should be 
able to do more ad hoc outreach (which is something they can do) but 
I also think that with in the ARIN region we have a separate (perhaps 
quite similar, but still) process for global policy proposals.

-- 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Edward Lewis                                                +1-571-434-5468
NeuStar

Think glocally.  Act confused.