Meeting minutes
This meeting
Nigel: [reviews the agenda]
Nigel: Any other business or points to raise within the agenda items?
(nothing)
Apply streamlined publication to all of Note track documents
Nigel: I think we're only waiting on tt-profile registry. Atsushi, how is this looking?
Atsushi: I opened a PR
Atsushi: We're in streamlined publication already for png-hdr-in-pr, and I've opened PRs for the other Note track documents
Nigel: Any objections to going ahead and merging?
PR to add streamlined publication to TTML Profile Registry
(none)
Nigel: Atsushi, do you want to merge?
Atsushi: I'll recheck it
DAPT
Test Suite
Nigel: I don't think I've seen any updates on the test suite issues. Cyril, you said you'd hoped to review some of them
Cyril: I had an early look, also asked a colleague. I have feedback on script represents that I haven't had chance to input on yet
… There should be a test for the linear whitespace syntax
… I intend to add a comment about that
… That's the only feedback for now
Pull requests
Explicitly permit daptm:represents on tt, body, div, p and span #294
Andreas: I had a quick check. It seems there's a source issue, but this was a while ago so need to check again
Nigel: As far as I know this is the last substantive issue we plan to make a change for
… Once we've done this we should request a new CR snapshot, which will trigger another 60 day exclusion period
… Anything else on DAPT?
(nothing)
Cyril: Netflix wants to create some content, to demonstrate DAPT, give real world examples people can refer to. Are the W3C repos available?
Nigel: I think a Netflix repo makes sense here, we could link to it from the test repo. Would that work?
Cyril: I think so. But in other cases like AOM we've shared in their repos, so wanted to check what to do here
Atsushi: We could accept examples in the test repo, things that aren't mandatory in the tests
Nigel: Yes, in an examples folder, and mentioned in the ReadMe
… Cyril, what would the licence for these be?
Cyril: Creative Commons non-commercial, sharealike, or somesuch
Atsushi: An example could be more complex than typical tests to show real world usage, so can be valuable
Cyril: The test report is about testing the formal features of DAPT, but the examples I have include proprietary or dummy metadata from Netflix which wouldn't be tested per se
Nigel: There is an issue for a test that includes proprietary metadata, where the implementation is supposed to say it's valid
Cyril: Similarly, if we want an IMSC example of showing how to generate subtitles from DAPT
… We could also have GitHub submodules in a Netflix repo
Nigel: I'm happy with Atsushi's suggestion to have examples in the test suite, it makes them easier for people to find if they're in one place
Cyril: Any restrictions on the licence?
Nigel: Would the W3C Software Licence work?
Atsushi: It should be fine to apply an open source licence, as the test repo isn't an official W3C spec repo
Nigel: The IMSC content is dual licenced under BSD 3 clause and W3C Test Suite Licence, so you can choose
Nigel: Cyril and Atsushi, please check on the licences
Atsushi: I believe many WGs use other licences like MIT or Creative Commons for examples
Pierre: w3c/
… This is under a BSD 3 Clause Licence, so that's possible
IMSC-1.1_Text_TestContent (BSD 3 clause licence)
Cyril: Was that required to be BSD?
<atsushi> https://
Pierre: No, it was offered to W3C and they accepted it. So I suggest doing that here
Atsushi: The Web Platform Tests uses the BSD 3 Clause licence, so I believe there's no strict requirements for tests and examples for W3C
… WebXR has an examples repo that uses MIT licence
Nigel: I think it's up to Netflix to determine which licence they're happy with
Cyril: I'll get back to you
IMSC 1.3
Nigel: The PR preview isn't working!
… Can someone look at it?
Atsushi: I'll check it later
Nigel: Are we ready for FPWD publication?
Pierre: Nearly, I think we should allow until the end of the month at most to make a decision on whether to remove the Image Profile
Nigel: On 24 April we said we'd consider any liaison responses
… We have a response from EBU. No concern raised about dropping Image Profile. Is that the same for you, Pierre?
Pierre: Yes, maybe some potential use. But any potential user would have feedback on the PNG version. I haven't had feedback saying people want to use it or improvements they want to see
Nigel: Has anyone else heard views?
(nothing)
Nigel: The EBU response said they don't use Image Profile, and no members use it in their operations
… They said it would be useful to include DAPT metadata in IMSC documents
… I need to analyse it, but I think it can just be done
… It's foreign to IMSC, so IMSC already allows that
… Might be useful to have an informative section in DAPT and IMSC 1.3 to give an example
… So we have a decision to make about dropping Image Profile. We've done the outreach, not heard anyone request it. Pierre created a PR
Pierre: My concerns are with alt text, this was permitted with IMSC 1.1 and 1.2, but the alt text is there to support image, it would be weird to keep that in Text Profile if the main application is actually image
… It would be the first time prohibiting a feature previously allowed so now IMSC 1.3 wouldn't be a strict superset
… We might find lots of nits in the spec on removing Image Profile
Nigel: [Reads the ittm:altText section in IMSC1.2]
… That suggests to me there may be a use for doing it for text
… I might suggest a quite strict approach, choose one of two things
… Either keep it as is in 1.3, or we say it's a breaking change and so we need a semantic version upgrade to 2.0
Pierre: Would be fine if we limited to that, but worry about other deprecated things
Nigel: I was only thinking of that one
Nigel: A major version change could allow us to remove other permitted deprecated things
Pierre: Most of the text of the feature will be gone. If we ever need to review Image Profile, we'll need to revise that text again
… The description of the feature would be much shorter, as most of it is focused on alternate text for img
Nigel: The other permitted deprecated things are in Text Profile: Z-index, aspect ratio, and progressively decodable
… We might end up breaking existing documents
Pierre: And inadvertently, that's the issue I see
… I'll add it back
… I'll do another editorial pass, then will be ready for a CfC for FPWD
Nigel: I haven't finished looking through it yet, so will finalise my review
Nigel: Any other comments?
(nothing)
PROPOSAL: Remove Image Profile from IMSC 1.3
Pierre: I'm thinking of this as publishing a new version that's only a Text Profile
Nigel: Okay, I'll modify the proposal:
PROPOSAL: Remove Image Profile from IMSC 1.3 and create an IMSC 1.3 Text Profile as a standalone document
Nigel: Is everyone OK with this approach? I will highlight it as a CfC in the minutes.
Pierre: Thanks to everyone getting the feedback
TPAC 2025 Planning
Gary: I haven't spent much time on this
Nigel: Last meeting we said do similar to last year. There was discussion in MEIG too.
Chris: Nothing changes - still thinking of a joint meeting on the Monday with APA perhaps later
… in the week but we haven't discussed with APA.
… I'm wondering whether this group would want to meet later in the week or if we could get all
… our business done on the Monday and Tuesday and then we could try to fit APA in on the Monday/Tuesday
… schedule. I don't know what APA will be able to accommodate, or if they have their own plan.
Nigel: Sounds like an action for Chairs - should I get in touch with the APA Chairs (Matt and Janina)
Chris: Good idea, start a thread between you and me and them.
Nigel: And Gary of course!
Gary: The GitHub issue is relatively minimal because there are only 4 slots to request / say is acceptable per day.
… The CSS WG said they wanted 8 slots and they're all good.
… We can also say we'd like all our slots on one day and not overlap with MEIG and Media WG
Nigel: Do we also ask for joint meetings?
Gary: Yes we can ask for that too.
… Once the issue is created we can add comments if we get more information
… They want us to do this as soon as possible but before the 20th, but if we need to we will have
… one more meeting scheduled before the deadline.
Nigel: I propose we open the issue, ask for 2 slots on Monday and Tuesday, and say we don't mind which slots, avoid overlap with MEIG and Media WG, then update the issue after we've talked with APA
Chris: Only 2 slots?
Nigel: How long are the slots?
Chris: 90 minutes
Nigel: Oh, then 4 then!
Gary: Some are 2 hours
Chris: You don't know what you'll get unless you're very specific
Gary: We can ask for 4 slots and say it would be nice not to be split across multiple days
Chris: Yes
Gary: I might not be there in person, so a morning slot is better for people in the US
… Timing would be 7 or 8pm east coast, 4 or 5pm west coast
… It'll be standard time in US again
Chris: So select early slots only or run the meeting all day?
Gary: Request early slots on Monday and Tuesday and we can adjust agenda topics to fit
AOB
Nigel: We're now under our new charter
Meeting close
Nigel: Thanks everyone, we're slightly over time, let's adjourn. See you in 2 weeks.