Meeting minutes
This meeting
Nigel: Today we have some comments on IMSC, a check-point on TTML2,
… and one AOB so far. Pierre will be with us but half an hour later than the beginning so
… we should rejig the order
… Is there any other business?
group: [no other business]
AOB - RFC8759
Nigel: I thought the group would be interested to know about RFC8759.
RFC8759 RTP Payload for Timed Text Markup Language (TTML)
Nigel: It provides a generic mechanism in the context of RTP for people to send
… streams of TTML documents.
… There are some constraints, like media time only, and the times are related to the
… RTP timestamps.
… This is formalisation of work previously discussed with, and reviewed by, this group.
Andreas: Congratulations for it. I think BBC put most work into this?
Nigel: Yes, the author works at BBC R&D.
… There was other supporting input as well from other parties.
Andreas: I wanted to ask how this relates to the IMSC live proposal, if this could be a
… good connecting point to push this?
Nigel: There's a key difference between what can be done with TTML over RTP and
… what can be done with the proposed TTML Live extensions.
… That difference is that TTML Live extensions can allow retrospective modification of
… the content on the TTML timeline.
… This cannot be done with TTML over RTP because RTP is a "now do this..." type of system.
… Conversely that RTP way is simpler for the same reason.
Andreas: For EBU-TT live there is the concept of a carriage specification.
… Is there the same concept for TTML Live?
Nigel: Yes there is
Andreas: Then you would say TTML over RTP would not be such a carriage specification?
… Or could it be one that uses a subset of the TTML Live extensions?
Nigel: Really good question. I did write the mapping down at one stage.
… I wrote something about this a while back:
TT-Live to TTML in RTP and back
Nigel: I should adjust that document now the RFC has been published.
… This document describes the conversion in each direction. It can be done.
Andreas: Thanks Nigel
TTML2 Implementation Report
Nigel: I think the main work now is tests.
… I saw Glenn's comment on one of the tests yesterday which I reviewed and agreed with.
Glenn: Yes, slow start getting going on those tests, but expect to see some PRs this week
… on that. I want to get them out there so we can get some implementations wrapped up
… and get moving on the proposed recommendation process.
Nigel: Also worth noting there has been some activity on the privacy review of IMSC and TTML2
… this week.
Glenn: Yes, I saw that. I think it won't take much effort. We can pull some small amounts
… of text into the appendix to address those points.
Nigel: I agree
Glenn: Those will be editorial changes for the proposed Rec.
Nigel: Yes
Glenn: Expect some pull requests for that process too.
Nigel: On the classification of change, they will be changes that have no effect on conformance.
Glenn: That's correct.
… That appendix is non-normative anyway.
Nigel: That will be useful or important for IMSC 1.2 also.
… I think we expect to update the ref from IMSC 1.2 to TTML2 to point to TTML2 2nd Ed
… and that's an agenda point for today.
… It could be that the resolution to the privacy issues consists of changes both to TTML2
… and to IMSC 1.2.
Nigel: One of the comments is that TTML doesn't mention anything about secure transport.
… But I think it makes no comment about transport at all.
Glenn: That's correct. We abstract out the transport by referring to the document
… processing context.
… It makes it easy for us to deal with this I think, by throwing it in the black box.
… We can make handwaving gestures to refer to this to say, if you're interested, pay
… attention to [blah blah].
Nigel: That's right. One of the suggestions is that we should change any URLs in examples
… to make sure the protocols are secure ones.
Glenn: Yes I will check that, it may be worth doing.
IMSC 1.2 - PING comments
Nigel: We received comments from PING via Nick Doty, which he has kindly
… raised as a GitHub issue.
… Let's cover those after the other two agenda points.
IMSC 1.2 CR Transition
Nigel: Where are we up to with transition?
Atsushi: Immediate approval was not gained. It will be discussed in tomorrow's slot, including
… recent updates.
Nigel: Does the team need me or Pierre to be part of that discussion?
Atsushi: I need to update following the comments from HR groups.
… Also comments on these reviews are welcome.
… Comments on the transition request issue are welcome from you.
Pierre: What's really important for us to understand is if the delay is due to lack of agenda
… time or because of concerns regarding comments?
Atsushi: It is the first of those. I needed to raise a comment by last Friday's slot.
Pierre: Now it is ready so it can be considered tomorrow?
Atsushi: Yes
Pierre: So there will be no update to the transition request?
Atsushi: We can add new information at any time.
Pierre: If the meeting is tomorrow then we have no more time to make modifications
… unless we do it now.
… I want to avoid a response tomorrow to say the request is not complete, let's wait another week.
Atsushi: I think the current transition request is complete and also refers to recent updates.
Nigel: Thank you.
… Then there will be a small modification to the document to reflect the new publication
… date.
Atsushi: I believe so.
IMSC 1.2 Reference to TTML2
Nigel: I wanted to check in with the group here.
… My assumption has been that we will update the TTML2 ref from IMSC 1.2 to point
… to TTML2 2nd Edition before going to Rec. Are there any counter-views?
Pierre: I don't have a different assumption, please file an issue.
Nigel: Happy to do that.
… Does anyone think we should not do it?
group: [silence]
Nigel: OK I think that is consensus to update the ref to TTML2 2nd Ed. Thanks.
IMSC 1.2 PING comments regarding privacy
Nigel: The reason it's useful to reference TTML2 2nd Ed is because some of the
… privacy comments may be addressed in part by reference.
Pierre: I think we need all the issues to be logged on IMSC GitHub.
… It is not practical to trace comments on other repos.
… We should not process anything that is not on the IMSC repo.
… I propose we let the W3C team know this.
Nigel: I'm not aware of any expectation to do so.
… Just for completion, Nick Doty raised one issue on IMSC 1.2 and one on TTML2.
… I think this these are all the issues we are expecting following the email review.
… I also think they made the right call about which repos to file them against.
… The generic resource fetching issue is filed against TTML2 where the features are defined,
… and the more specific domain related feature about fingerprinting regarding font
… matching is filed against IMSC 1.2.
CSS font-matching algorithm may introduce fingerprinting issues imsc#530
github: https://github.com/w3c/imsc/issues/530
Nigel: Did we actually introduce CSS font matching algorithm?
… I see at https://w3c.github.io/imsc/imsc1/spec/ttml-ww-profiles.html#text-font-source
… that we introduced:
… "A Processor MAY use the [css-fonts-3] §5 font matching algorithm for associating a font with a run of text."
… My question is, if this is an option, not a requirement, why wouldn't the CSS handling
… of the privacy issue be implied by reference.
Pierre: Just to point out that in §10.5 we mention the CSS font matching algorithm
… is also referenced via a defined term Font Matching Algorithm.
… Editorially we should improve that.
Nigel: Right, and that's in the HRM section.
… The HRM considerations are in my view concerned with document validation, and there's
… no requirement for the presentation processor to follow any steps in the HRM to
… render content.
… I would not expect a user-oriented player to execute the steps of the HRM.
Andreas: +1
Nigel: And therefore there's no privacy issue associated with 10.5.
… That takes us back to 8.5.3.
Pierre: To your earlier point Nigel, I don't see what action we can reasonably take.
… There are a lot of "mays" and "under discussion" and no proposed resolution.
Email that prompted this issue
Nigel: There are additional questions in the email that are not in the GitHub issue.
Pierre: We have generic text in TTML2 about loading of resources, I believe.
Glenn: There are some handwavy statements
Pierre: About resource fetching?
… In the absence of specific concerns we can only offer generic guidance.
Glenn: Exactly.
… I don't know what we can practically say.
Pierre: We can ask about specific issues with the TTML2 text.
Glenn: Ask for spec-ready text we can drop in.
Pierre: Exactly, that's what we should do.
… We can't tell CSS and HTML how to do fingerprinting mitigation.
<atsushi> +1
SUMMARY: TTWG thanks @npdoty for raising this. In the context of continuing discussions and without understanding any specific improvements we can currently make, we will proceed with no changes for the time being.
SUMMARY: Discussion of additional questions raised in the linked email to continue offline.
Meeting close
Nigel: Thanks everyone. [adjourns meeting]