[arin-ppml] Draft Policy ARIN-2025-6: Fix formula in 6.5.2.1c

[Eric C. Landgraf]

On Feb 16 12:00, arin-ppml-request at arin.net wrote:
> Date: Mon, 16 Feb 2026 21:09:26 +0700
> From: William Herrin <bill at herrin.us>
> Subject: Re: [arin-ppml] Draft Policy ARIN-2025-6: Fix formula in
> 	6.5.2.1c
>
> On Mon, Jul 14, 2025 at 7:49?AM Owen DeLong <owen at delong.com> wrote:
> > Assuming /48 as the PAU is not true to the policy. The policy was
> > intended to encourage LIR/ISPs to implement /48 PAUs by limiting the
> > size of allocations to providers that issued smaller end site
> > allocations.
>
> 1. Make ARIN's implementation conform to the written policy. I.e.
> start asking ISPs what size assignment they've selected as their
> default.
> 
> 4. Do a broad rewrite of section 6.5.2.1, which will result in ARIN
> creating an implementation of the new policy. We do, after all, have a
> decade more experience with the practical operation of IPv6 than we
> did when this policy was written. Surely we can improve it.

I would like to see ARIN implement the policy as written (or at least
there should be a good reason not to do so). In practice, the end result
may well be the same besides exact netmask, because of sparse
allocations, and the minimum /32 allocation for most providers.

Generally I think §6.5 needs a rework. It's pretty opaque, has very
strong "IP scarcity mindset" rules, and is not as consistent in
terminology as it could be. At minimum, we can get rid of confusing math
and be intentional about our use of LIR versus ISP. I suspect we could
come up with a set of rules that is reasonably simple and covers both
LIRs and direct allocations, and finally excise the word assignment from
the section. Personally, I think there are issues with how these rules
play in the Edu space, as large state universities often sit somewhere
between LIR and End-users (also complicating application of §6.11).

	Eric C. Landgraf