Minutes from Media Timed Events Task Force call 26 May 2020

Dear all,

The minutes from today's Media Timed Events / WICG DataCue API call are now available [1], and copied below.

There were a few action items:

1. Kaz to publish the use case and requirements document as an updated IG Note
2. Rob to comment on WHATWG HTML issue #5297
3. John and Chris to organise follow up with media companies

As mentioned, W3C members are welcome to join DASH-IF calls on this topic. The next call is Friday 29th May at 14:00 UTC. Please contact Chris for dial-in details.

Kind regards,

Chris (Co-chair, W3C Media & Entertainment Interest Group)

[1] https://www.w3.org/2020/05/26-me-minutes.html

--

W3C
- DRAFT -
Media and Entertainment IG - MTE-TF
26 May 2020
Agenda

Attendees

Present
  Kaz_Ashimura, Chris_Needham, Estella_Oncins, Gary_Katsevman, John_Simmons, Larry_Zhao, Yajun_Chen

Regrets

Chair
  Chris

Scribe
  cpn

Contents

Topics
  Publishing an updated IG Note
  Add TextTrackCue end time representing end of media
  Text track cue event timing accuracy
  issue 5297
  Should we close the TF, or are there other topics to work on?
  DataCue use case review
  Next call

Summary of Action Items
Summary of Resolutions

<scribe> scribenick: cpn

# Publishing an updated IG Note

Chris: CfC to publish: https://github.com/w3c/media-and-entertainment/issues/36
.... There was some support, so we can go ahead and publish
.... Kaz, would you be able to help with the publication steps?

Kaz: Sure, we can use echidna for publication

Chris: Do I need to do anything, as editor?

Kaz: No, I'll use echidna to publish

Chris: The draft is here: https://w3c.github.io/me-media-timed-events/
.... Things have moved on, new APIs, we could include, but i'm happy to move on

# Add TextTrackCue end time representing end of media

Chris: https://github.com/whatwg/html/issues/5297
.... Should get input from Rob
.... Some implementers not comfortable with setting and end time beyond the media end time
.... Can discuss here and update the proposal

# Text track cue event timing accuracy

Chris: https://github.com/whatwg/html/issues/5306

<RobSmith> I'm online. Please email me a link to meeting

Chris: Proposal here is to change HTML to recommend 20 millisecond accuracy of event firing

<RobSmith> Thanks

Chris: Some, e.g., Apple, saw this as a quality of implenentation issue
.... Not sure when Chromium's new implementation will be released, but this wll be an improvement
.... I'll draft a PR with proposed wording

# issue 5297

Gary: Eric said if media duration is infinite, ok for cue end to also be infinite, but not if media duration is finite

Rob: Why is that the case?

Gary: Right now, can set a duration beyond the media duration

Rob: What's the benefit of that additional restriction

Chris: Maybe this matches the WebKit implementation?

Gary: For id3 cues, they set the end time to be the media duration, and if another cue comes in, that cue is created through the end of the video and the previous cue's end time is updated to the start of the new cue
.... for captions or in-band metadata

Rob: So what precludes using an infinite end time?
.... in the WebVMT use cases, the cues persist to the end of the media, drawing lines on a map, both persist to end of the media

Chris: We shuold respond to the WHATWG github issue
.... I'll also follow up with Eric directly to ask how we can have the conversation

Rob: The proposal was a short step, based on existing media duration support for Infinity

# Should we close the TF, or are there other topics to work on?

Chris: We have WHATWG and WICG topics to follow up, but anything else for the IG TF?
.... We could also look at rendering aspects?

John: It seems that doing an analysis of the original goals and what has been accomplished would be good to do before that
.... There could be additional issues that arose that need more focus, or maybe a refresh or change to the TF
.... Or it could be a time to wrap up, or we could solicit more input from others in the media industry
.... Part of the problem is that we don't have enough high bandwidth communication with media companies
.... It could take the form of a summary of what's been done that could be shared with DASH-IF or other groups, ask for input on next steps

Chris: We have a good relationship with DASH-IF

John: There's also activity in IETF on HLS, so soliciting input from people there would be worthwhile. People are in CTA WAVE, DASH-IF, IETF
.... For example, Warner Media, not so much here, but they're in the HLS interop discussions, looking at this same set of problems, good to get feedback

Chris: How to bridge the gap between the media companies and browser implementers

John: We could organize something to present the results, under whatever umbrella. It might cause people to join in, or bring improvements, or lead to new work

Rob: I'd echo that, we want to get people engaged
.... Bullet chat is interesting, there could be overlap with that and CSS things i've been doing, good to explore and collaborate

Chris: I think there's still a lot to do, so we can reevaulate the TF, and bring in other interested people

John: Let's talk separately

Chris: We've talked about having something ready for TPAC, working a schedule back from there

# DataCue use case review

Chris: We need good use cases to convince implementers
.... https://github.com/WICG/datacue/blob/master/explainer.md
.... for the ad insertion use cases, we need input to validate this

John: What's the status of DataCue in Media WG? Some points of view expressed at last TPAC. Is it on the plan for Media WG? Has there been a review from Media WG in the explainer?

Chris: Last TPAC was the last time it was discussed with those people

John: We could talk with the media companies who'd want to use this. There's appetite from CTA WAVE members, Warner Media, Hulu. Not sure if the right path forward
.... If the media companies say they want the functionality, it would get prioritised.
.... The IG is their proxy, their advocate, so we need their feedback on the explainer
.... We can discuss offline. I feel some time pressure on the DataCue issue. The use cases are quite varied, also related Type 1 and Type 3 playback interoperability. What events are exposed to the web app for a Type 1 player?
.... Let's find a way to have that conversation. I can prepare materials to share

# Next call

Chris: Monday 15th June
.... Regular MEIG all on Tuesday 2nd June

Kaz: WoT WG are planning an online meeting mid-June, may overlap with the MTE TF call on 15th June

Chris: Please let me know

[adjourned]