[arin-ppml] Against 2013-4

[Jimmy Hess]

On 6/4/13, George Herbert <george.herbert at gmail.com> wrote:
> I think the better question is, what are we trying to preserve in IPv6.
> Raw address space?  No point for the forseeable future (40+ year protocol
> lifetime? ...)
No  (well,  only to a limited extent)

> Routing table growth?  ... Yes.
Yes, Definitely
> Complexity of allocations?  Arguably yes.
Yes, Definitely.

In the case of IPv6;   slightly too small of an allocation can do much
more damage than slightly too large of one.

In other words:  the logic is almost the opposite of what IPv6 was.
A prefix longer than say a  /32;   should    not be allowed to be
issued,

without evidence of  "Justified non-need"  [  because it is at too
great a risk,  that the applicant might some day need more addresses
-- resulting in the use of an additional routing slot.]

"Justified non-need" --
In essence,  a guarantee  signed in blood that no more space will ever needed,
even in the event of a network redesign,  acquisition, etc.

Longer  prefixes should require  more justification that no more
addresses will be needed, than shorter ones.



--
-JH