Meeting minutes
Welcome Chris Needham as new co-chair of the Media WG
jernoble: Welcome to chris as the new WG co-chair
<chcunningham> Welcome Chris!
chris_needham: thank you to the previous chair, Mounir, and thanks to everyone in the WG
Autoplay policy detection status
Status of Autoplay Policy Detection
chris_needham: Gary asked about current status of the autoplay api
gary: still a pain point for developers; having an api to detect this policy would be useful
gary: sync vs. async was the previous hangup, the TAG was included, then the issue was dropped.
gary: is there anything else stopping work from progressing?
chris_needham: other implementors were busy, looking for editors for the specification. Paul from Mozilla has a volunteer for an editor
jernoble: no blocking issue from WebKit's POV
chris_needham: next step is to find out from Mozilla the identity of the volunteer'd editor
tess: small spec, so it's fine to have a single editor; if we need a co-editor, we can find an editor from Apple
chris_needham: that would be appreciated, thanks
EME status in the charter
How to gatekeep the EME "API to find existing sessions"?
chris_needham: Q's about "find existing session" API.
joey: the original proposal is being backed out, and alternatives are being considered; greg suggested yet another alternative on the charter repo
joey: greg and i need to get together to hash out alternatives; committed to doing privacy review as well
chris_needham: can this be considered the same feature as key rotation?
joey: no, they're similar, but not really the same problem.
tidoust: Interested in the charter space, rather than the implementation space. features need to be listed in the charter, and that's the driving question here
tidoust: there's a pull request on the charter, approved by joey, to remove this feature as a deliverable from the charter during this charter renewal
tidoust: is that acceptable to the WG?
chris_needham: there's a huge debate on issue #132, many voices think this was important.
JoeyParrish: the importance hasn't changed, it's useful, people want it, and some DRM providers are doing it already.
JoeyParrish: it's not necessarily hurting applications, since the current implementations do this outside of the specification.
JoeyParrish: its' a matter of documenting current practice
JoeyParrish: don't want to hold up the chartering process
jernoble: can we do this in a side spec
tidoust: can it just be an informative document, that's a note rather than a normative spec?
MattWolenetz_: is there implications around IP / patent exclusions
tidoust: if it's informative, it's not covered; a CG report will be slightly better because it would be covered by the IP policy of the CG
jernoble: shouldn't need to decide CG vs Informative note now, as neither blocks the charter
tidoust: starting the final review by W3C mgmt of the re-charter, if all goes fine we can send the draft to the AC for review. at the same time, we'll have to extend the current charter to cover the gap between the old and new charters.
w3c/mediasession#264 Add new actions to support video conferencing websites
Add new actions to support video conferencing websites
chris_needham: no representation from those on the thread in the issues; deferring
Status update on WebCodecs
chris_needham: can the editors share an update on the current status of WebCodecs
chris_cunningham: lots of feedback, incorporated into the specification
chris_cunningham: origin trial for Chrome has shown that current approach works
chris_cunningham: would like to triage the bugs into the "bug" category and "breaking changes" category so we can prioritize the latter
chris_cunningham: encourage WG to look at the issues in the latter list and provide feedback
<tidoust> See V1 Launch Blockers
chris_cunningham: video frame readback is a pain point for adopters
chris_cunningham: window vs. worker exposure is an open issue that has a lot of comments
chris_needham: now that the spec is inside the WG, how much would you like to discuss the issues inside the telecon?
chris_needham: consider this an invitation to bring these issues to the telecon to discuss and drive consensus on these contentious issues.
chris_cunningham: haven't prepared to bring these issues to the WG, and not all interested parties are represented
bernard: we would like the discussion to occur in the MediaWG rather than in WebRTC
bernard: makes the conversation difficult to manage, and the consensus needs to happen in the relevant WG
bernard: "breakout box" => "media stream track insertable" => ... (lost0
chris_needham: which group does that live in?
bernard: not an item of WebRTC WG; to some extent it's orthogonal to WebCodecs; not needed to render or to input (though convenient)
bernard: shares some elements with WebCodecs (due to shared objects like "VideoFrame")
chris_needham: Q re: the privacy review; is there anything you'd like to raise?
chris_cunningham: had a meeting with the PING; went through the review point by point and resolved to file individual GitHub issues or simple follow ups; updates needed to the Privacy Considerations section of the spec
chris_cunningham: other specifications have listed mitigations that the PING feel more straightforward that the PING think the WebCodecs spec can incorporate
chris_cunningham: many of the fingerprinting areas are not new/shared with other specs like WebRTC and MediaCapabilities; others are new and we are sensitive to that
bernard: in terms of incremental exposure; the list of supported codecs is already available from existing APIs
chris_cunningham: you can become aware of hardware or platform differences above and beyond supported codecs
MattWolenetz_: broader encoder support in these implementations than exists on the platform ATM
tidoust: WebGPU has not gone through privacy review yet; not the best example; neither WebGL
tidoust: are we looking at the right list for the blocking issues? the window-vs
… worker item is not in the blocker list
chris_cunningham: i do consider that issue to be a breaking change; not intentionally omitted from the breaking changes list
bernard: that original bug came in through breakout box; that wasn't intended and may be why that discussion started
tidoust: many untrained issues in the spec tracker; are those considered minor?
chris_cunningham: process is to do a weekly triage; some may have been missed
bernard: many of the issues are documentation issues (rather than implementation issues)[
bernard: many are just questions
chris_cunningham: there are other issues, like missing any mention of color spaces
bernard: many of the tags don't represent W3C spec status; some may be CR blockers but not V1 launch blockers
tidoust: that activity is great, but the volume of activity makes it hard to follow
chris_cunningham: I'd be happy to add a label to issues for extension vs. blocker
jernoble: is anyone using this api on the main thread? if not, could we just ship with the worker at first and add to the window?
chris_cunningham: i'll do this research and get back to the WG
chris_cunningham: this issue took us by surprise
bernard: usually this is a security issue, but there's no security issue at hand here
chris_needham: when you say "shipping a V1", you're referring to doing so outside of an origin trial?
chris_cunningham: yes.
chris_needham: around media session, mounir is currently named as a spec editor? is this something you're interested in continuing or are you looking for a replacement?
mounir: I'm happy to stay, but also happy to hand off work if there's another interested WG member?
mounir: ideally need a replacement editor, perhaps even two.