Summary

  • The IESG announced on 28 August that the Security Dispatch working group had concluded, its mailing list would close and its function had merged into a rechartered DISPATCH venue covering ART, SEC and non-transport WIT topics.
  • The merger needs a proposal-level transfer record. It should preserve the old outcome, its consensus state, its non-binding authority class, the surviving archive, any destination and the conditions for a later return—without pretending that dispatch feedback approves standards.

One intake closes, another is already operating

The IETF did not close the door to new security work. It changed the door.

On 28 August, the IESG Secretary announced that Security Dispatch had concluded. Christopher Inacio and Deb Cooley were named as the IESG contacts. The notice said the old mailing list would close and that the group had merged with DISPATCH. The Security Area, it added, still encouraged new ideas to be proposed for dispatch.

The receiving institution was already in place. The IESG approved the current DISPATCH charter on 2 July. It gives one standing venue responsibility for proposed new work in the Applications and Real-Time Area, the Security Area and the non-transport part of the Web and Internet Transport Area. A joint interim on 30 June and a single DISPATCH session at IETF 126 in July show the transition operating before the formal closure notice.

Consolidation can be sensible. A proposal about identity, messaging or web application behaviour can cross area boundaries before anyone knows where it belongs. One intake reduces the need for a proposer to guess the IETF’s internal map. A shared room may also expose overlap sooner.

But a dispatch group’s durable output is not the protocol. It is the disposition of the proposal.

Dispatch advice is a routing act, not an approval

The new charter is unusually clear about this boundary. DISPATCH may point work to an existing working group, recommend a BoF, help form a new working-group charter, suggest possible Area Director sponsorship, create a discussion list, or advise that the idea be deferred or rejected. Its guidance is not binding. Feedback is not guaranteed when the chairs cannot determine consensus. Except for narrow administrative documents accepted with the relevant Area Directors, the group does not process documents.

RFC 7957 describes the same model. A DISPATCH-style group evaluates new work and finds a venue. It does not complete the proposed work. For borderline cases, the responsible Area Directors and chairs decide which venue should handle the proposal. A draft can also go to an existing working group or a BoF without passing through DISPATCH.

Those distinctions matter when records move. “Sent to a BoF” is not “chartered.” “Possible Independent Submission” is not a Dispatch action. “No action at this time” is not permanent rejection. “No presentation” is not a technical judgment. If a merger compresses all of those states into “handled by the old group,” later readers will invent authority the group never exercised.

The 30 June joint interim shows why exact state survives better than a narrative summary. Its chairs recorded about 50 participants and asked the lists to confirm rough consensus on six item-by-item outcomes. One presenter mentioned the Independent Submission Editor, which the record explicitly said was a possible path but not an action from Dispatch. Two items received no presentation. One received no action for the IETF. Another received no IETF action at that time, with a need for more independent implementation interest and a discussion venue. A fifth was directed toward a named BoF and organising effort.

Those are six different records. None should be upgraded, erased or silently reopened merely because the meeting used two group names and the successor uses one.

The missing object is a proposal transfer record

The closure announcement does not publish a crosswalk. That absence does not prove that a proposal was lost or that private operational work was omitted. The old public archive remained reachable at this article’s cutoff and displayed 1,699 messages. The question is narrower: can a proposer or later reviewer move from the old record to the current institutional state without reconstructing the transition from list searches, agendas and memory?

A public transfer record should give every materially handled proposal a stable identity and record:

  • the draft name and version, or another stable proposal identifier;
  • the final relevant old-list thread and meeting record;
  • the outcome text and whether it was proposed, consensus-confirmed, deferred or left without presentation;
  • the authority class—chair summary, group rough consensus, Area Director choice, BoF route, existing-WG adoption, Independent Submission choice or no IETF action;
  • the named destination or current owner, if one exists;
  • any condition for returning with substantially revised work;
  • status at the merger cutoff;
  • the surviving archive and correction route; and
  • any later DISPATCH thread, decision or superseding entry.

This is not a demand to reconsider every idea. A “no action” outcome can remain exactly that. Nor should subscribers’ identities or private correspondence be exposed. The record is about public institutional state, not a new dossier on participants.

Correct the procedural citation

The closure notice also contains a small but consequential record error. It says new ideas remain welcome for dispatch “per RFC7975.” RFC 7975 specifies a request-routing redirection interface for interconnected content-delivery networks. The approved DISPATCH charter cites RFC 7957, the Best Current Practice describing DISPATCH-style working groups.

The one-digit error does not invalidate the merger, and there is no evidence that it changed a decision. It should still be corrected or annotated. A conclusion notice may become the page people use to find the successor process years later. The public record should point to the method it actually invokes.

What the evidence does not show

The reviewed record does not establish that any named proposal is stranded, that the archive will disappear, that old advice is binding, or that a proposal is entitled to a new hearing. The IETF 126 materials page did not show session minutes at the evidence cutoff, but that does not prove that no outcome record exists elsewhere or will appear later. The proposed transfer record is Daniel Kade’s governance recommendation, not an announced IETF commitment.

The IETF has already made the structural decision: one cross-area dispatch venue now replaces a separate Security Dispatch entry point. The next test is evidentiary. A merger is complete when the new door works and the old decisions remain intelligible.

Sources