[arin-ppml] ARIN-2011-3: Better IPv6 Allocations for ISPs - Last Call

[Charles O'Hern]

On 4/24/11 7:13 AM, Joe Maimon wrote:
> 
> 
> Owen DeLong wrote:
>>
>> On Apr 22, 2011, at 8:47 AM, Joe Maimon wrote:
>>
>>>
>>>
>>> ARIN wrote:
>>>> The ARIN Advisory Council (AC) met on 13 April 2011 and decided to
>>>> send a revised version of the following draft policy to last call:
>>>
>>> I remember when policy proposals where abandoned, citing as a factor the complexity in the proposal. I suppose they just were not complex enough.
>>>
>> I believe the phrase in those cases was "unnecessary complexity".
> 
> A subjective hedge. In any event, I personally believe that any proposal with mathematical equations fits the definition of "unacceptable complexity". Can you supply a web page app
> that will let average folk punch in easy to identify numbers to produce results in line with the formulas?
> 
> If a proposal requires a calculator, the complexity involved needs a fair bit of justification.

Any proposal about numeric addressing that doesn't include mathematics is overly simplistic to the point of not accomplishing anything.
Any 'average folk' who are also ISP network administrators should be able to do powers of two.  Luckily a good portion of the general public can do arithmetic.
If you require a calculator to follow any of this, you probably shouldn't be participating in addressing policy.

>>
>>> Do we want the general public to be able to read this NRPM thing we are crafting?
>>>
>> Yes.
> 
> We have long ago failed at that, and the direction we are taking now will not be making anything better.

Any member of the general public has access to the means to comprehend any part of the existing address policy or any part of this proposal.  I'm vehemently opposed to holding back
any policy because of a perception that the public can't do math.  Following that argument we should replace all the complex word thingies with brightly colored pictures.  I will
not condone kowtowing to the idiocracy.

>>
>>> This proposal does nothing to change a fundamental and longtime ARIN policy defect.
>>>
>> Huh?
> 
> It is very simple.
> 
> If you use IP resources as quickly as you can, so long as the justification passes muster (and there is an evolving latitude involved in that), you get rewarded with more
> resources, faster.

I agree in concept.  But that defect is inherent in the way IETF divvied up the bits with the /64 limit to subnet size.

> As a trivial example:
> 
> Suppose you number /30 serial links. Completely justifiable. You will be rewarded for doing that by using your resources faster and thereby qualifying you for more of them sooner
> than if you had uses unnumbered, /31, or rfc1918.
> The only time policy has ever taken direct aim at this was for virtual hosting. It is not clear to me how policy can address this defect without taking direct aim at methodologies
> and technologies, which I dont consider a good idea either.
>>
>>> To the extent you can get away with it, profligacy is rewarded.

The way IPv6 is written, profligacy is encouraged.  Whether this will be seen as a good thing in 40 years or not, remains to be seen.  Its also outside of our (ARIN's) demesne to
change, at least with how things work right now.

-- 
Charles O'Hern
Network Operations

TCSN - The Computer Shop Netlink
1306 Pine St. Paso Robles CA 93446
1-(805) 227-7000  1-(800) 974-DISK
http://www.tcsn.net  abuse at tcsn.net