Re: Planning to close Procedural Memory Knowledge Representation Community Group due to inactivity unless we hear from you

Hi Ian,

Fair questions, so the plain answers first.

Conversations: the list was active in February and March, 36 messages from five people (Christoph, Adam Sobieski, Milton, Chris Lilley and me), and it has been quiet since Milton's message of 1 April.
Meetings: none so far.
Contributions from others: Milton is drafting the charter and a survey of priorities.
So the group itself has been quiet, and as chair that is on me.

One correction: it is not a personal repository.
The work lives in the projects themselves, reviewed and merged by their maintainers (HandBrake, MKVToolNix, mpv, Universal Media Server), and in review at KDE, Mesa, FFmpeg and the Wayland protocols.

It is also closer to the web than it looks, and a little history shows why.
YouTube launched a 3D player in July 2009: anaglyph, side by side and interleaved, switched on per video.
It lived in the Flash player, and when YouTube moved to HTML5 the 3D modes did not come along.
YouTube still accepts stereoscopic uploads today, and its ordinary player shows them flat.
I lived that era: I pitched 3D to Joe Penna (MysteryGuitarMan) in 2010, and his 3D videos went on to millions of views.
The history, with sources: https://github.com/danielcamposramos/awesome-stereoscopy/blob/main/stereo-video-on-youtube.md

Today the web's only stereo vocabulary is inside WebXR: the Layers API has "stereo-left-right" and "stereo-top-bottom" layouts, for a headset session.
Headsets and display glasses are bringing stereo to ordinary people again, and many of the same people own 3D televisions and monitors from the 2010s.
Outside a headset session, a page has no standard way to say that a video or a canvas is stereo, so the screen next to the headset shows it flat.

The specifications themselves are not new.
OpenGL has had stereo buffers since 1992, HDMI 1.4a defined its 3D formats in 2010, and H.264 and HEVC carry the frame packing inside the stream.
On Linux they were adopted only in pieces: Intel's driver has handled HDMI 3D since 2013, but no desktop and almost no player ever used it.
That is what we have been fixing, from the kernel up to the desktop, and the next step on our ladder is the browser: stereo video and WebGL applications on any screen, not only inside a headset.
With all respect, the web has not adopted them either: browsers ignore the frame packing a stream carries, and outside a headset session no web specification asks them to use it.
So, in a small way, we are doing the web's homework too, and this group is where we would like to hand it in.

For PM-KR it matters because Knowledge3D, as already explained, represents knowledge in space, and knowledge in space reaches people only when their screens can show depth, in a browser as much as in a headset.

So this is what I propose for the group, on the list:
1. A short report on stereo 3D, established specifications adopted only in pieces: what exists, what Linux and the web never adopted, and what a browser would need. I will post it to the list for discussion by the end of October.
2. Milton's charter draft and priority survey, on the list, for the participants to answer.
3. A first call once the survey is in, open to all 24 participants.

Please give us until the end of October to show that activity in the group itself.

Best regards,
Daniel Campos Ramos, chair, W3C PM-KR CG

Received on Tuesday, 6 October 2026 16:53:51 UTC