[address-policy-wg] RE: Question

Oliver Bartels wrote:

> On Mon, 24 Apr 2006 09:10:26 +0200, Marc van Selm wrote:
> 
>>>so, let's switch to discussing
>>>
>>>   http://www.ripe.net/ripe/policies/proposals/2006-01.html

[ ... ]

> I support your argument and the proposal, too.

<taking off all hats other than a long-term 'netizen and LIR manager>

I very strongly tend to support the proposal, with some questions and
qualifications (peculiar to the NCC) attached:

- As the proposal is worded now, it implicitely states that an applicant
can "walk in from the street" and ask for an assignment. However, some
time ago, we consciously supported the NCC to _exclusively_ do business
with LIRs. Whenever an entity requests resources that are not "directly"
tied in with an ISP, this entity has to find someone (an existing LIR)
to act on their behalf.

- Looking at the text further under the heading "Expiry for Assignments"
there is the possibility to the holder of such an assignment to become an
LIR

Given that setup, and the fact that (from an RIR's point of view there is
not too much difference in handling when cmparing PA and PI address space,
I would propose to *require* the applicant to become an LIR in the 1st place.

At that point in time when this LIR does no longer need the resources, and
has returned them properly, it can cancel the contract (and stop paying the
annual fee)

I have the very strong opinion that any entity requesting "core resources"
from or using services from an an RIR (big address blocks, AS numbers, rev.
delegations,...) should also contribute to the operational cost *and* be
alerted to the responsibilties of holding/using such resources.

Which then goes hand in hand with the privilege of providing guidance to
the RIR's operation by way of e.g. voting rights in the GM.

Potentially inventing a separate fee structure or dedicated size category
is a minor administrative exercise, I think.

Please note that the "nice" thing about that approach is to vent a lot of
steam regarding PA(good) and PI(bad).

Very simplistically put - this proposal is a way around the (dreaded) 200
"customer" rule, isn't it ;-)

Wilfried.