[rpd] [Last Call] Draft Policy Proposal - Hierarchical Names for New AS-SETs (AFPUB-2026-ASN-001-DRAFT02)

[NP Petronella]

Hi Hendrik,

I do not believe Tshepo’s position is contradictory, and I support his objection.

The fact that DRAFT02 contains an exception for restoring an AS-SET deleted by mistake does not settle the concern. The issue is not merely whether restoration is theoretically possible. The issue is whether the policy defines that restoration clearly, narrowly, predictably, and without creating unnecessary institutional discretion.

Section 3, as you describe it, requires exceptions to be considered on a case-by-case basis. That wording leaves the outcome dependent on interpretation. A policy should not force operators to rely on discretionary relief from a restriction that the policy itself creates.

Tshepo’s proposed wording is therefore not redundant. It attempts to distinguish a specific, verifiable restoration case from a broad exception mechanism. A pre-existing object, restored under the same key from an archived state with documented evidence, is materially different from giving the registry open-ended authority to decide which cases deserve exemption.

This is also consistent with the concern Alain raised in another policy discussion: broad terms, undefined evidence requirements, and promises that the community can correct excessive discretion through a future PDP cycle are not substitutes for clear policy now.

Saying that a problematic exception can be narrowed later is not a convincing reason to approve vague language today. The burden is on the present proposal to be sufficiently precise before adoption. The community should not be asked to ratify discretion first and repair it afterwards.

Nor is the choice simply between DRAFT02 and “zero protection.” Existing registry records, archived data, operational procedures, and a narrowly drafted restoration provision can be considered without accepting the wider policy structure. Rejecting one draft does not prevent the community from adopting a better and more limited solution.

The broader objection therefore remains valid: the proposal creates a mandatory naming restriction and then relies on discretionary exceptions to address the operational problems caused by that restriction. That is not predictability. It is administrative control presented as flexibility.

For these reasons, I support Tshepo and continue to object to the proposal as written.

Regards,

Nonhlanhla


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.afrinic.net/pipermail/rpd/attachments/20260730/e37fe406/attachment.html>