[address-policy-wg] Re: [policy-announce] 2010-01 New Policy Proposal (Temporary Internet Number Assignment Policies)

On 18/03/2010 16:06, Leo Vegoda wrote:
> I agree with James that by distinguishing between 16 and 32-bit AS
> Numbers in section 2, this text is inconsistent with the main ASN
> policy. I suspect that the RIPE NCC staff can reserve a couple of small
> AS Numbers for this purpose and there is such a large number of big AS
> Numbers that a reservation is probably not necessary.

The latter is certainly the case.  When drafting the policy, I didn't want
to prevent the RIPE NCC doing what they think might be necessary with
ASN32s.  Under the terms of the policy proposal, they're authorised to
reserve some ASN32s; whether they choose to actually reserve them is an
operational matter for them to decide.

As regards distinguishing between ASN16s and ASN32s, the fact of the matter
is that there are technical differences between the two, regardless of what
the current policies state.  Perhaps the main ASN policy needs to note this?

> The proposed policy text does not state that these assignments are
> subject to a contract, presumably because that is specified in the text
> of the main policy. Am I right in thinking that requesters would need to
> sign a contract to get resources under this policy?

Yes, correct.  This is covered by section 3.4 - "Compliance with Other RIPE
NCC Assignment Policies".  There is value in not repeating policy text
which appears in other documents, although it would probably be useful to
note the "IPv4 Address Allocation and Assignment Policies for the RIPE NCC
Service Region" document explicitly.

> While it is an
> implementation detail, it would be nice if the time frame was documented
> in the contract.

yes, good point.

> Also, if I have read it correctly, this proposal text removes the
> publication requirement for assignments made to experiments and academic
> research. What is the reason for removing this requirement?

Several reasons actually.

- I'm not sure of the intent of the original experimentation requirement,
but suspect that it was a quid-pro-quo: that in order to make it possible
for researchers to get large chunks of address space, they had to prove
that they were really doing research, and in order to prove it, there was
an obligation to publish the results.  But given that the ipv4 address
assignment landscape is shortly going to change irrevocably, it didn't make
sense (to me) to have two temporary assignment categories.

- given this, it would be discrimination or an encumbrance which did not
apply to other assignment categories.  This struck me as not being
particularly fair.

- most research journals will not accept articles which have been published
elsewhere or where there is a contingent requirement for the researcher to
publish elsewhere.  "Elsewhere" includes FoC publishing on the Internet.
This could create a chain of potentially serious problems for any
researcher who needed temporary address space for the purposes of their
work, given that the research cycle generally goes: grant application ->
work -> publish -> grant application based on impact factor of journal
where previous work was published.

- if there were such an encumbrance in place for just researchers, they
would probably end up requesting the address space on terms which had no
such encumbrance (e.g. time limited project).  By removing the encumbrance
requirement, people will be less tempted to be dishonest on the application
form.

> Finally, in the rationale section a., I think the phrase "a negligible
> effect on the final RIPE NCC IANA-supplied IPv4 address pool depletion
> date" should be "a negligible effect on the IANA-supplied RIPE NCC
> address pool depletion date" to improve clarity.

That would be clearer, yes.

Nick