Minutes for yesterday's WG-CG call, and Agenda for next call

Hello WG and CG members!

Here are the minutes from yesterday's meeting:

https://github.com/w3c/socialwg/blob/main/meetings/2026/2026-04-16-CG-WG-joint.md

Some highlights and notes (and a little editorializing) for those reading
this email:

 - Evan Prodromou proposed a process
<https://github.com/w3c/socialwg/blob/main/meetings/2026/2026-04-16-CG-WG-joint.md#evans-notes-on-process>
for incremental WG updates to the AS 2.0 Core, AS Vocab, and ActivityPub
documents, particularly the v1.1-ish errata updates but also triaging
things that belong in a future "next" style update.  Discussion thereof is
in the "WG updates
<https://github.com/w3c/socialwg/blob/main/meetings/2026/2026-04-16-CG-WG-joint.md#wg-updates>"
section of the minutes. In particular we discussed approval process for PRs
and what current practices from other groups are. Seemed to me like many
other spec groups don't do a kind of "voting by group PR approval" but
rather an editor-driven process with specific "checkpoints" where diffs are
discussed at group meetings and the meetings function as a place for
raising objections.
- Look forward to diffs of above documents being published for group review
in the coming weeks, along with specific guidelines for what we'd like the
groups to be reading for
 - There are very interesting open questions in task forces. For example,
the E2EE task force has a lot of movement happening. I know there's a lot
of interest in what e2ee looks like in an ActivityPub ecosystem and I want
to encourage people who care to show up to those meetings and/or
participate asychronously via the TF Github repo
<https://github.com/swicg/activitypub-e2ee>. I'm using e2ee as an example
but you can find the repos for every Task Force on the SWICG repo. Here's a
direct link to a search for all Task Force repos sorted by last updated:
https://github.com/swicg/?q=taskforce+OR+task&type=all&language=&sort=

Additionally, I'll be looking into reasonably privacy-preserving
transcription solutions that work on Jitsi. The purpose here is to help us
be more inclusive to people with difficulty following English or following
audio more generally for whatever reason. There's a workflow Jitsi
recommends
<https://jitsi.support/developer/setup-jitsi-whispercpp-transcriptions/>
that involves piping audio locally to a local Whisper LLM and then piping
that to the Jitsi API. This would allow us to have automatic transcription
and subtitles for meetings without sending audio to a big tech platform.
The only parties processing data would be my local computer and the Jitsi
public server that hosts the meetings, which are each entities that already
have full access to meeting audio.

The Github issue with the proposed agenda for the May 1 joint call is here:

https://github.com/w3c/socialwg/issues/9

Please comment on the issue if you'd like something added to the agenda.

Thanks,
Darius Kazemi
Chair, W3C Social Web WG