Re: [apnic-talk] APNIC EC Election Review Panel
On 18/07/2010, at 10:14 PM, Matthew Moyle-Croft wrote:
>>
>
> Currently the voting system means that, roughly, votes represent the number of end users (ie. think about it - larger numbers of IPs represent larger end user bases).
>
well, actually, no. The current voting system doesn't represent that.
looking at Japan again and using:
http://www.apnic.net/services/become-a-member/who-are-our-members/member-search (BTW would be nice to have that as exportable CSV or put on the ftp site as csv!)
X Large: 2
V Large: 1
Large: 7
Med: 10
Small: 32
V Small: 7
Ass: 9
Giving a total of 503 votes, or 4.1% of the total 12201 votes in the APNIC membership.
Hardly reflective of the number resources (see previous emails for a link to some stats) or even the end users[1] in that economy.
At some point in the history of APNIC I dare say that the voting system would have represented the ideal you put forward. But I cannot see that as reflecting nor incorporating the larger set of APNIC stakeholders across the region today.
To break it down to the simplest terms, all the voting structure suggests at this point is that larger members, who pay more based on the size of their resource set, have more votes as a member. Nothing more.
So that critique of votes apportioned to the membership sets itself against both the 1 member/1 vote as well as the existing voting structure. The problem here is that while the membership class is derived from the resource holding and NIRs exist to aide in their local economy there will probably always be this contradictory element to the equality of the votes.
[1] I really dislike the using the term 'end users' as it is such a hard thing to quantify and generally includes counting errors ie right now, I am using 3 ipv4 addresses and 2 ipv6 addresses but I am 1 user..
> One APNIC member, one vote gives you a very different representation base: http://www.apnic.net/__data/assets/pdf_file/0003/19407/2009-APNIC-Annual-Report.pdf (page 3)
>
> Which is fairer? Which is actual proportional representation? Should Australians get 28% of the APNIC vote? How's that fair to Japan with 3% of the members due to an NIR?
Equally, if Australia were to form an NIR it would create some substantial changes. Potentially the entire economy with the largest holding of IPv4 would have 1 vote. Maybe that is what we need. Maybe we should be encouraging all economies to create NIRs and therefore all economies thus have just 1 vote - with a total of 56 votes.
Lets face it IPv4 is almost done. The amount of additional IPv6 applications people might need to make is going to be small. So the work effort required by APNIC in anything but a full blown trading model is equally going to be small. And should an IPv4 trading model be the result, the the fees from transfers should levied to recovering the service cost. In that is a multifaceted voting structure really the concern?
Or maybe we flip the thoughts the other way, do we truly need the numbering allocation function as it stands. Has it become too much of a cost (time/effort/money) to the industry? once ipv4 is done, why not centralise the entire function globally? Perhaps efficiency, aggregation, uniqueness should come from structural simplification? Do we need the geographical/regional constructs (RFC2050) post exhaustion? Perhaps there should be 1 registry, and all others are just deregulated registrars? Do we need the multiple simultaneous policy processes? I watch all the RIRs policy lists and I see more homogeneity than not. Perhaps one global set of meeting events is required. In terms of a consensus approach - the ietf seems to achieve that as an example without having separate regional organisations.
... hmm challenging thoughts. Is all that we are doing actually for the 'good of the internet'??
>
> It's all a compromise, but I don't see that changing it from the current system is any fairer.
Compromise is a pretty generous term :-)
But I share your concern that fairness isn't improved by what has been suggested so far. (not saying that the current system is fair, just that it is the incumbent)
>
> Travel, for an organisation that serves a very large part of the world is going to be high. As per my previous posts the %ages APNIC spends on various things is very much inline with other RIRs. So I doubt it's going to be something that really has a big effect.
umm.. what if the other RIR's are spending too much too? o.k. Unfair statement and unqualified. But maybe we should stop looking to the 'others' as the measure?
>
>
> It's easy to look at a big number and say "it's too much". But I suspect if you look at the number of trips involved it'd make sense.
I think this hints at a question of transparency and reporting. I wonder if all the trips that APNIC did and the number of people they sent were published with a matching set of reports as to what was achieved at the meetings that this whole issue might just go away??
>
> Certainly if people want more EC members and regional reps then this figure ain't going to go down.
>
>>
>> Membership fees are of less an issue to me than the Initial start-up fees of $3000+. I've been told APNIC were reviewing this fee yet I've seen nothing happen yet since I was told close to a year ago it was being looked at. Membership fees themselves I think are fair and equitable.
>
> Have you proposed a policy on it? Isn't so much of this actually being raised here in the wrong place? I'm not opposed to the membership startup fee as a higher bar to reduce the number of IPv4 allocations. Especially if combined with 1 member, 1 vote, it seems a way of making it easier to sign people up for voting needs.
>
As far as I recall, fees are not the responsibility of the membership or policy process exactly. The EC controls setting the fees. However such things can (I think) be raised on the ML or presented at the open mic.
Terry