Attestation review turns heated as MPLS and mail drafts advance
IETF lists spent the day on a contested cryptographic review of early attestation bindings and a disputed MPLS adoption call over RSVP and segment routing. Parallel threads covered workload identity for ACME, mail attribution and DKIM policy, DNS ANY queries, SSH key exchange, and CDDL code generation.
CFRG review of early attestation bindings turns contentious
The Crypto Forum Research Group fielded a request for cryptographic review of a formal ProVerif analysis of cryptographic binding mechanisms for early attestation. The thread drew 58 messages from 11 participants and became heated, with conduct complaints and claims of withdrawn implementations. Readers following attestation designs should treat the review as contested rather than a quiet technical sign-off.
MPLS adoption call for RSVP-SR meets architecture objections
The MPLS working group ran an adoption call on draft-beeram-mpls-rsvp-sr-01 that ended 2026-11-17, producing 70 messages from 24 participants. Opposition cited conflicts with segment routing architecture in RFC 8402 and RFC 8426, and concerns about transit state. Operators considering RSVP and segment routing together have a clear record of why the draft is not uncontroversial working group work.
ACME draft proposes workload tokens without domain validation
A new Internet-Draft, draft-rasmussen-acme-wif-00, proposes OIDC workload identity tokens for ACME certificate issuance without domain control validation. The ACME list debated placement and scope across 33 messages from 7 participants. Certificate issuers that authenticate workloads rather than hostnames now have a concrete draft and an open argument about where it belongs.
Dispatch discusses an Agent-of header for AI-generated mail
The dispatch list discussed a draft for an Agent-of email header field intended to attribute AI-generated mail. Seven participants, in 26 messages, examined verification via DKIM and delegation. Mail operators and client authors have an early proposal for marking agent-sent messages, with authentication still unsettled.
DKIM list weighs DKIM2 and DMARC sender policy
The IETF DKIM list discussed a DKIM2 and DMARC sender policy proposal, including policy alignment and modification flags, in 61 messages from 16 participants. Security concerns centered on forwarding. Sites that forward or modify mail have reason to watch whether those flags change existing authentication assumptions.
DNSOP presses a harder stop on QTYPE ANY
DNSOP discussed strengthening the deprecation of QTYPE=ANY beyond RFC 8482, toward a MUST NOT, because of attack amplification and a lack of legitimate use. The thread ran to 17 messages from 10 participants. Resolver and authoritative operators still answering ANY have a sharper signal that the query type is unwanted.
SSH last call opens on strict key exchange
The SSH working group opened a last call on draft-ietf-sshm-strict-kex-02, ending 2026-12-08, to mitigate the Terrapin attack. Five participants posted five messages. Implementers of SSH key exchange have a dated comment window before the strict key exchange text moves on.
CBOR list revisits code generation from CDDL
A developer asked the CBOR list for best practices on code generation from CDDL, four years on. The working group chair noted ongoing work on modules and annotations, in a thread of 12 messages from 4 participants. Tool authors generating code from CDDL have an active discussion rather than a finished guide.