IETF debates DKIM2, list conduct, and protocol work

IETF lists saw heated technical disputes over DKIM2 design choices and SEAT security models, plus governance steps on posting rights and a new AI discussion list. Adoption calls and addressing debates also advanced in TEAS, deepspace, DNSOP, and RATS.

DKIM2 key validity and encoding choices

The IETF DKIM list debated a proposal for a key validity field in DKIM2, covering JSON versus tag-value encoding, key expiry tags, and null body recipes. Implementers pushed back on the added complexity. Email authentication developers should watch whether the design stays simple enough for broad deployment.

Last Call on PR-Action for Ken Kubota

The IESG opened a Last Call on an RFC 3683 PR-Action to suspend Ken Kubota's posting rights on four IETF lists for disruptive unprofessional commentary. Final community comments are being solicited. The outcome will set a concrete example of how the IETF enforces list conduct norms.

SEAT debates TLS binders against relay attacks

The SEAT working group argued over guidance text for draft-ietf-seat-use-cases, including whether shared secrets in TLS binders should be mandatory to prevent relay attacks, with citations to papers and formal models. The dispute turns on security modeling assumptions. Protocol designers in this space need to track the resulting requirements.

New list for AI in the IETF standards process

The IETF chair announced a new non-WG mailing list, ai-in-standards, for community discussion of artificial intelligence use in the standards process. The list is open for process and policy questions. Participants concerned with tooling and authorship norms now have a dedicated forum.

TEAS questions distributed power-aware TE

During the call for adoption of draft-many-teas-power-steering, the TEAS working group debated whether distributed power-aware traffic engineering is feasible at all. Participants challenged the draft's approach on technical grounds. Operators interested in energy-aware path selection will care whether the work proceeds.

Deepspace single versus multiple RIR models

The deepspace working group continued an active debate on single versus multiple RIR models for space IPv6, aiming to maximize address aggregation. The choice affects how routing scales for space networks. Address allocation structure remains a core open question for that effort.

New DNS RRTYPEs for UNECE and ISO codes

An individual draft proposes two new DNS RRTYPEs to carry UNECE and ISO registry codes verbatim for DIEM use. The DNSOP thread examined semantics and process fit. Applications that need structured external registry data in DNS may gain a standardized carrier if the work advances.

RATS adoption call for verifiable geo-fence

The RATS working group opened a two-week adoption call for draft-lkspa-rats-verifiable-geo-fence. Multiple co-authors and contributors expressed support. Attestation work involving location binding will track whether the draft is taken up.