All, The Meetecho recording, slides and minutes of the DNSOP interim-2026-dnsop-02 meeting on PQ DNSSEC are available via datatracker:
https://datatracker.ietf.org/meeting/interim-2026-dnsop-02/session/dnsop The chairs and PQ DNSSEC side meeting organizers think that while actual protocol or other standardization work can occur in DNSOP (or the new post-split WG once it is formed), the side meetings may still remain useful for implementation reports, measurements, non-IETF-draft studies, and the like. For convenience, I have added the minutes to this email below. Best, Peter --------- DNSOP WG Interim meeting interim-2026-dnsop-02 Date: Friday October 2, 2026 Time: 1300 UTC Chair: Benno Overeinder Only discussion during the mic line is covered here, not the slides You should definitely read the slides: Minutes by Paul Hoffman Meeting info: https://datatracker.ietf.org/meeting/interim-2026-dnsop-02/session/dnsop Full recording: https://meetecho-player.ietf.org/playout/?session=IETF-DNSOP-20261002-1300 Administrivia, Chairs Opening, Note Well, Review of current WG documents Mic line at the end of the presentations Peter Thomassen: Paul metioned non-NIST algorithms Would like an overview How to proceed: go to ML-DSA and work on TCP operations Start with ML-DSA now and maybe switch to something better later We can always add other algorithms later Has concerns how MTL affects online signing Push towards TCP might help make secure transports happen sooner Paul: Follow the NIST mailing list to see breaks in signing from the Chinese competition Stephen Farrell: Agrees with Peter Doesn't think optimizing for TCP will help increase number of zones that are signed Ben Schwartz: Finds MTL interesting We will push ahead in all directions at once DELEG will change the way we think about DNS over QUIC We have ML-DSA in production Wants to hear from validators about MTL mode MTL and online signing: it might still come out better than current Karl Dyson: Interested in understanding on very dynamic zones (one per second) Would MTL cause a full redesign of the zone? Mukund Sivaraman: Adopt ML-DSA Good support in libraries We would have a solution today Likes dropping any current algorithms We should measure resolver to authoritative traffic Simplicity and elegance might be better in the long term than complicated Paul: No need to adopt Bas' draft because it already has a codepoint in the IANA registry Benno: Should the work be done in DNSOP or in the Security Area? Can decide later Ben: When sizing authoritative server, size is not as important as DoS protection Authoritative fleet sizing is not that important Running over TLS was not as big of a cost as being ready for DoS protection Bas observation of single valid path is important Maybe need changes to validation rules or protocol Mohamed Boucadair: This work belongs in DNSOP Deb Cooley: As long as you are choosing algorithms that are well-vetted, you can do it in DNSOP If you are looking at new operations, you need more security review Maybe call in CFRG if you are using non-NIST algorithms Benno: More on the mailing list, and will be brought up again in San Francisco PQ-DNSSEC side-meeting chairs organize helpful side-meetings and helped organize this meeting. _______________________________________________ DNSOP mailing list -- [email protected] To unsubscribe send an email to [email protected]
