Summary

  • W3C Strategy issue 563 opened the Web Real-Time Communications Working Group recharter on 21 July 2026. It remains open and in charter refinement even though its recorded 25 August target has passed.
  • The draft says there is no substantive change in scope, but it identifies six items as discontinued work and presently describes SFrame as the replacement for WebRTC Identity.
  • A security reviewer said SFrame does not preserve Identity’s identity, target-peer and media-isolation properties, and asked for either a successor or an explicit narrowing of scope.
  • A WebRTC co-chair agreed the technologies are unrelated, attributed the retirement to lack of implementor interest and said the Working Group has no successor deliverable for those properties.
  • Open pull request 869 changes the text to say SFrame covers only limited use cases. A property-disposition crosswalk should record the uncovered residue rather than turn retirement into proof of equivalence.

Retirement became a property question

W3C’s rechartering record for the Web Real-Time Communications Working Group began with a modest premise. Strategy issue 563, opened on 21 July, says the proposed charter makes no substantive change in scope. It nevertheless retires several drafts as mature material moves to specifications further along the Recommendation track. The issue remains open, carries the labels “In Charter Refinement” and “Security review completed,” and still displays 25 August as the expected end of refinement.

The unfinished status matters. Neither the charter language nor its corrective patch is final. The security exchange is newsworthy because it caught a category error while the words can still be repaired: ending work on a document, replacing a technical mechanism and disposing of its security properties are three different acts.

That distinction is easy to lose in a retirement table. A line can name an old document, a newer technology and a future disposition in a few words. Readers then infer that the new item carries the old item’s functions. The WebRTC review produced unusually direct evidence that this inference would be wrong.

What the draft proposes to discontinue

The public charter draft lists six discontinued-work items. Four are normative drafts for which the table describes a migration or replacement destination. Two are non-normative use-case documents described as unmaintained. The table also preserves exclusion-draft dates and originating charters for the normative work, making retirement part of the group’s institutional and patent history rather than mere repository housekeeping.

For WebRTC Identity, the current public wording goes further than saying the draft has lost support. It says SFrame, added through WebRTC extensions, should replace it and cover data as well as audio and video. That sentence binds the reason for discontinuation to a functional replacement claim.

The older document did not merely describe media encryption. The 2018 Candidate Recommendation defined a way to validate the identity associated with a peer and let the browser constrain media to an identified target. Its isolation model could withhold content when the asserted identity or target conditions were not met. Whether browsers implemented that model widely is a separate empirical question; the current record does not establish broad adoption.

SFrame has a different job. RFC 9605 defines protection for encoded media frames in multiparty conferencing while leaving key management to applications. Its security considerations expressly say it does not itself provide per-sender authentication of media data. That does not make SFrame deficient. It means the two documents locate trust and enforcement in different places.

The security review forced the distinction

The 24 August security review asked the charter authors to account for three classes of property: identity, assurance that media reaches the intended peer, and browser-enforced media isolation. If those guarantees had a successor, the reviewer wanted it named. If the Working Group no longer intended to pursue them, the scope should say so explicitly.

The response from a WebRTC co-chair was unusually clear. SFrame and WebRTC Identity are unrelated technologies, he wrote. Identity was being deprecated because of a lack of implementor interest, not because SFrame replaced it. He also said there was no successor deliverable for the identified properties within the Working Group.

Those statements do not invalidate discontinuation. A standards group is not required to maintain every design indefinitely. Scarce editor attention, implementation demand and test capacity are legitimate constraints. But the rationale now becomes lack of adoption rather than equivalence, and the residual properties become an explicit absence rather than something silently inherited by SFrame.

An open patch narrows the claim

Pull request 869 was opened on 25 August against the charter draft. Its proposed sentence says WebRTC Identity will be discontinued for lack of interest and that SFrame in WebRTC Encoded Transform covers only a limited set of the document’s use cases. The commit is concrete, reviewable and responsive to the objection.

It is not yet the operative result. At the evidence cutoff, the pull request remained open and unmerged. The public draft still carried the broader replacement language. The correct state description is therefore “correction proposed,” not “charter fixed,” “decision reversed” or “replacement rejected.”

Even if merged, the patch would solve only the immediate false equivalence. It would not by itself say which Identity properties are abandoned, which might be useful elsewhere, or what evidence could reopen the question. “Limited set of use cases” is more accurate than “replacement,” but still leaves the limit unnamed.

Discontinued is a document state, not functional equivalence

The W3C Process supplies a disciplined way to close unfinished Recommendation-track work. When a group no longer intends to advance a technical report, it should publish a Discontinued Draft without substantive change and explain why. A group may later resume the work within charter scope by publishing a new Working Draft.

That state does useful work. It tells readers not to expect the document to advance and preserves a reason for the stop. It does not prove that every property has migrated to another specification. Nor does a replacement label prove implementation, interoperability or adoption. Those propositions require their own evidence.

This is where Heng Lu’s minimum-specification discipline is useful. A common record should carry only what coordination requires, while implementation and voluntary adoption supply future proof. Applied to WebRTC retirement, the charter need not settle whether every unused property deserves revival. It should avoid claiming that a new label creates equivalence which running systems and named requirements have not demonstrated.

The crosswalk that should survive the charter

A compact property-disposition crosswalk would preserve the decision without enlarging the Working Group’s mandate. For each discontinued document, it should state:

  • the source document and publication state;
  • the security or interoperability property being discussed;
  • the implementation evidence currently available;
  • whether the property is migrated, replaced, abandoned, deferred or still unassigned;
  • the exact destination deliverable and its owning group, if one exists;
  • the evidence for functional equivalence and the residual properties it does not cover;
  • the applicable exclusion-draft and patent-review lineage;
  • objections and their recorded disposition; and
  • the event that would trigger reassessment or resumption.

For WebRTC Identity, the record could therefore say that SFrame overlaps with a limited set of protected-media use cases, while peer identity, intended-target assurance and browser-enforced isolation have no named successor in the Working Group. It could separately state that the document is being discontinued for lack of interest. That is neither an endorsement of the old design nor a criticism of SFrame. It is a faithful description of state.

The separation also protects future work. If applications later demonstrate demand for a target-bound identity property, participants can see what was left behind without pretending that the entire Candidate Recommendation must return. If no demand appears, the retirement record remains honest and complete.

What the record does not prove

The available evidence does not show that WebRTC Identity was broadly implemented, that browsers removed a deployed feature, or that SFrame is defective. It does not establish a patent-policy breach. It does not show that the proposed charter has been approved or that pull request 869 has merged. It also does not prove that the missing properties should remain within this Working Group rather than move elsewhere or be abandoned.

What it does show is narrower and more useful. A public security review identified a mismatch between a replacement claim and the properties of the cited technologies. A co-chair agreed with that diagnosis. A textual correction is open. The remaining governance task is to give the uncovered properties a disposition that can be read independently of the retirement label.

Sources

  1. W3C Strategy — WebRTC Working Group recharter issue 563
  2. W3C — draft 2026 WebRTC Working Group Charter
  3. W3C charter-drafts — pull request 869
  4. W3C charter-drafts — proposed correction commit
  5. W3C — active WebRTC Working Group Charter
  6. W3C — WebRTC Identity Candidate Recommendation
  7. RFC 9605 — SFrame security considerations
  8. W3C Process — abandoning unfinished work
  9. W3C Patent Policy
  10. Heng Lu — Minimum Initial Specification, Localized Future Decision, and Voluntary Adoption