<p dir="ltr"><br>
On Apr 7, 2013 7:19 PM, "Paul Vixie" <<a href="mailto:paul@redbarn.org">paul@redbarn.org</a>> wrote:<br>
><br>
> ...<br>
><br>
> cb.list6 wrote:<br>
>><br>
>><br>
>> On Apr 7, 2013 12:49 PM, "Paul Vixie" <<a href="mailto:paul@redbarn.org">paul@redbarn.org</a>> wrote:<br>
>> ><br>
>> > i know that it's a popular viewpoint -- many folks feel that the time for needs based allocation is over and that the invisible hand of the market is now capable of optimizing the holding of address space and the aggregation level of that space into routing table entries.<br>
>> ><br>
>><br>
>> Popular viewpoint go far in a bottom up process such as arin. In fact, the whole thing is a popularity contest.<br>
><br>
><br>
> i said it was popular, not that it could win a popularity contest.<br>
><br>
><br>
>> > so i thought i'd chime in: i consider that case to be extremely unmade as yet. even though i am in most other ways a free-marketeer. as stewards of a public resource ARIN has always been guided by RFC 2050 which requires recipients of these public resources to justify their need, no matter whether these resources are coming from a central pool or a private transfer.<br>
>> ><br>
>> > paul<br>
>><br>
>> Does that mean you require an update to rfc 2050 to move ?<br>
><br>
><br>
> not at all. i think RFC 2050 was and remains correct in this regard. i'll "move" when and if my mind changes on the matter.<br>
><br>
>> I noticed this <a href="http://tools.ietf.org/html/draft-housley-rfc2050bis-01">http://tools.ietf.org/html/draft-housley-rfc2050bis-01</a><br>
>><br>
>> ...<br>
>><br>
>> Should 2050bis ask rir not do this fair policy? From what I read in 2050bis is that is says the rir can make their own policy and 2050 is dead.<br>
>><br>
>> Do you read it differently?<br>
><br>
><br>
> i read it to accurately explain that not every RIR still follows the needs based justification described in RFC 2050. it's a description of the current RIR system. 2050bis does not "ask" RIRs to do anything, it's a description of what they actually do.<br>
><br>
><br>
>> As it stands, speaking from experience, the justification story in v4 and v6 drives design choices. That is an unfortunate fact and negatively impacts system design.<br>
><br>
><br>
> i'm intrigued by this statement. i hope you are willing to share some of your experiences as to how needs based justification has negatively driven some design choices.<br>
><br>
> paul</p>
<p dir="ltr">I just wrote a page of explanation and deleted it. </p>
<p dir="ltr">If I have to explain it, you would not understand. And you do not understand today's data networks at all. I feel bad and outrageous saying that. But, given hundreds of millions of mobile phone users behind cgn today, perhaps your question is outrageous</p>
<p dir="ltr">Note that att and vz have both rolled cgn to their dsl subs. </p>
<p dir="ltr">Yet arin is not exhausted. </p>
<p dir="ltr">Interesting?</p>
<p dir="ltr">CB. </p>
<p dir="ltr"> </p>