Community Group Program

AI Agent Protocol

<Ian> Song: Thanks for having me; sorry I missed last call

Song's slides

[slide 12]

<Ian> ANP protocol site

(No questions)

CG Program topics

Ian: 8 new CGs since our last October meeting, 3 of which AI-related:

<Ian> AI-driven Web Standards Specification

<Ian> Agentic Arbitration Protocol

<Ian> Semantic Agent Communication

<Ian> Autofill

<Ian> Infinite Canvas

<Ian> Data Documentation

<Ian> Data Interoperating Architecture

<Ian> Cybernetic Avatar

Ian: Maybe we should organize a AI-CG get-together
… We are setting in place a system to have new CG chairs automatically informed about this meeting
… We had a discussion with PSIG about adoption of CG specs by non-W3C SDOs
… the CLA describes how specs move from CG to W3C WGs
… but we know of 25+ specs that have transitioned to other orgs, and we had had requests to make that process easier
… the Patent and Standards Interest Group (PSIG) indicated that for WHATWG, everything is in place, for the others, too complex to treat generally
… tooling support is a good start
… At TPAC, we presented mockups for CG Reports

Ian: we got a lot of feedback, incl making the language about "not being w3c standards" less hostile/more positive
… suggesting to make the style more distinct from W3C TRs
… and a suggestion to have a quick signal on "how mature" a report (in terms of a progress bar)
… which will be integrated in the next round of mockup
… with 4 stages (early, experimental implementation, standardization plan, transferred) with link to more detailed data below
… we also discussed metadata management based on work Denis, Dom and Xiaoqian started
… distinguishing frozen metadata in snapshots from live metadata in live editors draft
… with the exception of abandonment/transfer state that may need to be forced into snapshots if they were not getting updated to reflect the situation
… Open question on final reports: today, they're published on w3.org via a github pull request workflow
… with the goal of reducing confusion between CG/WG - should we move away from using w3.org for these final reports?

Elf: in SOLID CG, we submitted some of the reports for the LWS WG; we discussed whether they needed to be made final
… my understanding is that this wasn't so critical

<Ian> https://github.com/w3c/cg-program/blob/main/proposals/spec-lifecycle.md

Ian: the new spec lifecyle hopefully clarifies this, with the "transfered" state
… we would be interested in feedback

<Ian> Dom: The question of where to publish would be for snapshots (e.g., those associated with final spec agreements)

<Ian> ...the reason to have them on w3.org related to additional IPR commitments

Peter: having a place to host final specs on w3c domain is a good feature, to avoid relying on potential domains infrastructure used by CGs

wolfgang: important to have something reliable, stable, independent of github for all final snapshots,

<Ian> wolfgang: Searchability important

wolfgang: this also makes W3C s point of reference to find this type of technical specs

Debbie: +1 on having final reports on w3.org so we know where they are, to make sure they don't go away
… having a bit more of paperwork to get the final spec stage is a good forcing function for CGs

<Ian> Dom: We welcome feedback (here or in email) about the usefulness of this (relatively new) call

Next meeting

<Ian> 18 February at 16h00 UTC

<Ian> Anticipating Solid CG election procedure presentation at next meeting.

<PeterR> Feedback: The calls are excellent, presentations by new CGs are also a great idea to establish a CG culture.