Meeting minutes
This meeting
Nigel: DAPT, minor TTML2 update, Charter 2025, anything else?
Chris: AOB around the WebVideo Map Tracks (WebVMT)
DAPT
Implementation Report
Nigel: I have created a wiki page
Nigel: It's a starting point. I've copied in the draft CR exit criteria and created a template based on what we did before
… It lists the extension features from DAPT, we can assess implementations based on tests for each feature
Cyril: This is looking good. How do we test features that are prohibited?
… I think #source-data as example
Nigel: We can create a document that does what you're not allowed to do, and check a validation implementation fails on it
… and a document that does it the right way, check it passes validation. Is that enough?
… The alternative is to rework the extensions so they're expressed as prohibitions. Equally weird, but testable
Cyril: Either way, not every type of processor can recognise that, it has to be a validation processor. So to exit CR we need a validation processor, not just an authoring tool and a presentation processor
… What kinds of processors do we need to include in the implementation report?
Nigel: (Reads the requirements in the Charter)
… You could have a content producing implementation that does source-data the right way, but it's a pretty weak test. Show the content it generates is valid, without using a validation processor itself, use manual inspection
… Any suggestions?
Cyril: A validation tool based on the XML schema, could be the easiest way to validate this feature
… I'd like to understand which features we think will be difficult to pass, because we'll have two implementations, specifically for features marked at-risk, e.g., variants around audio source
Nigel: The at-risk features are all things that are part of TTML2 already. With the way the CR exit criteria are drafted, and the list of features in the test list, none of the at-risk features are listed
… I'm not sure we have specific sub-features or extensions. Script Event Grouping is an example.
… We might find nobody wants to implement that, so we'd have to remove support for it, would be a significant editorial change
Cyril: I think it's good enough for now
Nigel: The BBC and Netflix implementations are just listed as examples
Cyril: It would be good to encourage people to list their implementations, even partial ones
Nigel: Yes, e.g, the EBU EuroVox
… The moment to do that is when we publish CR, announce and ask for implementation details
… We need the report to exit CR, not enter CR
Nigel: Anything else on this topic?
(nothing)
Pull requests
Nigel: There are three PRs. Only one is required for CR, the work on feature dispositions
… Thank you Andreas for your positive review. Does anyone want more time for review before I merge?
Cyril: I don't need more time
Andreas: It does need some time to look at, there's a couple of features
Nigel: Andreas and I have looked carefully at it. I'm comfortable merging it, we have no requested changes, and enough time has passed
Cyril: I'm fine with merging it
Nigel: There's an editorial PR to add an XSD, and there's another to prepare for CRS that needs rebasing
CFC Results and CR next steps
Nigel: Any objection to declaring that we have consensus?
Cyril: I support
Andreas: Also fine with me
RESOLUTION: Request transition to CR based on w3c/
Nigel: Atsushi, you raised this for horizontal review
Atsushi: I'm asking for security review. I hope to hear back shortly
Chris: It's not a recharter, it's a new group. I'd concur not to block on it.
Atsushi: Usually the review is required, but we should be fine for asking transition
Nigel: Is there anything else you need from the WG to request transition?
Atsushi: No, I'm just waiting to hear back from horizontal reviews.
… There's no strict procedure on requesting review and CR transition
Nigel: I might need to adjust the expected PR date in the CRS. Any suggestions for a date to use?
Atsushi: I personally want to finish during December, before end of year
Nigel: I want to suggest changing it to 31 Jan
… I'll do that
Nigel: There's a pull request to add an XSD to the repo, which I created with help from Ben Poor from EuroVox
… It tries to validate some of the prohibitions and constraints, in DAPT not in TTML
… It pulls in the EBU-TT Metadata XSD via a git submodule
… It's there for review and try out to see if it works
… The IMSC XSD just references from TTML2 and SMPTE-TT and others. The idea is you set up the validation environment to know where those XSDs are
… Getting that to work in my experience is difficult for implementations
… So the approach I used for DAPT is to create one XSD
… Hopefully that's easier to use
Andreas: Are you using XSD 1.0 or 1.1?
Nigel: It's 1.0, easier to get tooling that works with it
TTML2 - outdated banner
Nigel: The banner now points to the most recent version, as requested
… Thanks to Atsushi and systeam
Charter 2025
Nigel: I drafted the charter
github: w3c/charter-timed-text#89
github: w3c/
Nigel: I've incremented the date by +2 years. I've adjusted the wording around teleconferences
… I added mention of historical applications of TTML2
… I fixed links, e.g., to the current Process
… I added IMSC 1.3, reusing wording from previous editions
… I added DAPT, and mention that it defines registries
… TTML2 stays the same
… I added ARIB as an external organisation
… I haven't updated the charter history section
… Please review and comment on the pull request
… Any questions?
(nothing)
Nigel: We'll submit around the end of January?
Atsushi: We need to send advance notice, so I'll file a strategy issue in mid-December to highlight the charter discussion
Nigel: That's perfect, thank you.
Nigel: Anything anyone wants to raise about the charter?
(nothing)
AOB - Web Video Map Tracks (WebVMT)
Chris: This is another Charter out for AC review.
… I'm here as an AC Rep reviewing the Charter.
… It's the Spatial Data on the Web WG
… One of the specs listed as in scope for potential maintenance is the WebVMT spec.
… My question to you is whether you want this WG to be listed under the coordination
… section for that WG given that they're basing work on WebVTT.
… They don't list TTWG as coordination at the moment.
… Would you want to see that, and if so, I'll mention it in the AC Review.
Gary: I don't see an issue with having TTWG listed,
… but at the same time, it's clearly been working fine as is so
… I don't think we necessarily should make the change to disapprove not having it.
… Last time they needed a discussion we had that, and agreed no change to WebVTT on that occasion.
… Not having it listed is not a blocker on anything [i.e. the groups talking to each other]
Chris: As you say, it's fine to leave it as-is in coordination.
Gary: It could be "if you're making other changes feel free to add it but otherwise don't worry"
Chris: I could say that and leave it to their discretion.
Nigel: Would they prefer to have closer alignment and higher compatibility between the specs. At the moment it looks like WebVMT is inspired by WebVTT
Gary: WebVMT documents may not be valid WebVTT, e.g., by using the missing end time feature
Nigel: Is that something W3C should maintain as a constraint on these specs?
Gary: Maybe long term, it's not needed. If the formats diverge based on their needs, that's probably fine
Nigel: OTOH, if they think of some new spec change, we'd hopefully discuss as both groups might want it in their specs
… I think the purpose of coordination is to have a list of groups to be consulted when there's a significant change
… We don't list them in our charter coordination list. But i think it should be symmetrical
Atsushi: There are several meaning for the list: from just being interested, to having a more formal relationship
… It seems to me some W3C groups are listed as just being interested. Not sure on the formality of this section
… Not sure the Process needs a strict definition, just guidelines. It wouldn't factor too strongly in the team's review
Nigel: Another consideration relates to immersive captions. If we decide to work on adding 3D coordinates into caption formats, it relates to that group's use cases
… While we aren't doing volumetric subtitles and captions, it has come up before and could again
Andreas: We discussed maybe 4 or 5 years ago, things in the market might have changed now
… Could be a requirement for us
Chris: I think it's worth suggesting to them that TTWG is listed, without proposing a tight collaboration.
… More a "keep us informed", and as a reminder as much as anything to come back to this group
… should they have particular changes that might be worth including in both specs.
Meeting Close
Nigel: Thank you for scribing, Chris, we're 5 minutes over, let's adjourn.
… See you in 2 weeks everyone!
… [adjourns meeting]