- From: Nigel Megitt <nigel.megitt@bbc.co.uk>
- Date: Thu, 12 May 2022 16:17:30 +0000
- To: "public-tt@w3.org" <public-tt@w3.org>
- Message-ID: <273932BA-EDBE-40B9-B81D-88F3C081CA3A@bbc.co.uk>
Thanks all for attending today’s TTWG meeting. Minutes can be found in HTML format at https://www.w3.org/2022/05/12-tt-minutes.html
In text format:
[1]W3C
[1] https://www.w3.org/
Timed Text Working Group Teleconference
12 May 2022
[2]Previous meeting. [3]Agenda. [4]IRC log.
[2] https://www.w3.org/2022/04/28-tt-minutes.html
[3] https://github.com/w3c/ttwg/issues/217
[4] https://www.w3.org/2022/05/12-tt-irc
Attendees
Present
Andreas, Atsushi, Cyril, Gary, Nigel, Pierre, Xabier
Regrets
-
Chair
Gary, Nigel
Scribe
nigel
Contents
1. [5]This meeting
2. [6]DAPT REQs
3. [7]TPAC Planning
4. [8]Rechartering status update
5. [9]Meeting close
Meeting minutes
This meeting
Nigel: Today, we have a quick update on DAPT-REQs
… TPAC Planning
… Rechartering,
… We also have TT in low latency, and Behaviour with controls.
Drop them, or is there something to discuss?
Gary: Can probably drop them
Nigel: OK
… Any other business, or points to make sure we cover?
No other business
DAPT REQs
Nigel: We now have published the draft WG note, at
[10]DAPT-REQs DNOTE
[10] https://www.w3.org/TR/2022/DNOTE-dapt-reqs-20220510/
Nigel: And I saw that Atsushi merged the pull request that
means whenever
… we merge a pull request to the default branch, it will
trigger republication automatically.
… So as agreed, any PR on that document is considered a Call
for Consensus.
… Just like with IMSC-HRM.
… I've prepared a blog post about it which Cyril has kindly
looked over.
… Thank you.
… This is just something Chairs can do, so after this call I'll
give it another look over and then publish.
… It's basically a call to action to review and contribute.
Andreas: General question about the DAPT profile:
… Is there any relationship to IMSC?
… I understand it will be similar to IMSC, and written in the
same style.
… Any intent to take IMSC as a reference for the features
included or excluded?
Cyril: The intent is not necessarily to represent something
that can be rendered directly [visually]
… but there's an option to be able to associate styles and
rendering with content.
… Two links with IMSC. One is that the structural constraints
in terms of timing,
… attribute restrictions etc will be based on the same thing.
… Then for visual rendering, we want to base those semantics on
IMSC too.
… It should be possible, if the author of the script also
provides rendering indications then
… it should be easy to produce IMSC subtitles.
Andreas: That makes sense, thanks.
Nigel: I was going to say something similar.
… We may get some comments about timebase, I'm not sure,
… but I think we start with the position of media time only.
… I think the next stage is to start drafting the
specification.
Cyril: Yes, the action is for me.
Nigel: Any more on this topic?
No
TPAC Planning
Nigel: Gary, will you be able to fill in the form for the
Chairs?
Gary: Yes
Nigel: [runs through list of questions]
… Agenda topics? At the least, DAPT.
Cyril: If we have a meeting, I can show up.
Nigel: Me too
… Any other likely attendees in person?
Pierre: If there's something interesting on the agenda, then
covid permitting I will be there,
… completely driven by agenda for me.
Nigel: Thank you. Any more?
… In that case, for now, in-person is 3 for the form.
… Any constraints on days or times?
Pierre: What are the dates?
Gary: September 12-16
Pierre: IBC is concurrent, so for me,
Cyril: IBC is 9-12
Pierre: Right, so at the earliest, 14th allowing for travel,
for anyone there.
… Preference is more towards the end of the week than the
beginning.
Nigel: That's a useful data point - it may affect others too.
… Next question is about group overlap and joint meetings
… Maybe Media WG because of controls?
Gary: Yes
Cyril: Interested in both Media WG and TTWG so would like to
avoid non-deliberate overlap,
… but also having meetings close in time/day would be useful
Gary: Same here
… Any other groups to avoid overlap with?
Nigel: Sounds like a no
Andreas: There's the MEIG - would not be good to overlap
Nigel: Good point
Gary: That's already likely to avoid overlap because Chris
can't attend both at the same time
Nigel: One from me: I've suggested to the Audio Description
Community Group that we should
… hold a meeting, the intent is to gather input and build
momentum for DAPT amongst non-TTWG people.
… (and maybe non-W3C members too)
… From a time of day perspective, I think we need to support
… Vancouver, Europe and Japan, based on the responses so far.
Atsushi: I don't think there's a sweet spot for Vancouver and
Japan
Gary: We may have to do it early in the day
Nigel: Any other events of interest?
… Technical plenary with high level presentations
nobody
Nigel: I quite like that, myself
… Demos
nobody
Nigel: Hackathon
nobody
Nigel: Workshop
Pierre: Just a thought - TTML and WebVTT are used by a large
number of people.
… Which is awesome. Very few of them are in this group.
… One option, to increase in-person meeting value, is to ask
the question beyond this group.
… Are there hot topics? We could try to have a workshop with
users.
… Lots more work but more productive and fulfilling.
… Want to throw that out. If we want to try to answer bigger
questions or make
… progress on larger issues, we should cast the net beyond this
group.
Nigel: Really good point.
Gary: To add, a lot of those larger issues apply equally to
TTML/IMSC and WebVTT
… even if the implementation details might be different.
Pierre: Totally agree, a lot of it is people coming to terms
with Timed Text
… Maybe TPAC is an opportunity.
Nigel: OK let's tick the Workshop box and bear this in mind. I
can see it may well be worth the effort.
… Next one is Developer Meetup in the evening including local
community
nobody
Nigel: Training overview on W3C technologies
nobody
Nigel: Any others not mentioned?
None
Nigel: Thanks. Gary, have you got everything you need?
Gary: I think so. Still unsure about timing, but I think we
don't have to be specific yet.
Nigel: True - also worth noting the locations of likely
meetings.
Gary: They also ask about meeting outside regular meeting
hours, which I think we would have to do.
Pierre: Atsushi, how likely are you to be able to travel?
Atsushi: For now, quite unlikely due to requirements for
re-entering Japan, unless they are relaxed.
… Please set the meeting time for convenience at the venue - I
should adjust in any case.
Pierre: Let's say that we have a workshop. I think we might
actually want to avoid having a group meeting at all.
… We can have those whenever we want.
… If we are going to miss important participants because of the
time, maybe we can
… not have a TTWG meeting but instead have a workshop that
benefits from in-person discussions.
… It would be silly to go to Vancouver to have a meeting we can
do over the phone.
Nigel: I broadly agree, but note that there is an intangible
value to making some decisions
… in person around the table, thinking back over previous
meetings.
… Even if they could have been done over the phone.
… Anything else about TPAC for now?
Rechartering status update
Nigel: News!
… Two meetings, one with each formal objector.
… The discussion with Google resulted in a pull request that
would resolve their objection.
[11]Proposed edit to resolve Google FO
[11] https://github.com/w3c/charter-timed-text/pull/80
Nigel: Please review.
… Thanks Pierre and Gary for already approving.
… This PR changes the "For example" sentence in §3.1 Success
Criteria
… which, although it is an example, does include normative
keyword MAY.
Cyril: I think I understand the purpose of the change, but want
to check.
… The last change, source -> implementation, I understand, that
it says there's an implementation behind.
… Why was the word "single" removed?
Nigel: I don't think I can express the answer to that, it would
be for Chris Wilson to say.
Cyril: Thanks, it looks good to me too. Will approve.
Nigel: Does anyone have any concerns about this pull request?
Pierre: This is your pull request though, Nigel?
Nigel: I did the admin of opening the pull request and typing
the words, but the change was driven
… by Chris primarily, on the call.
Nigel: I'm sensing nobody has any concerns about this change.
… The next one is from Apple.
[12]Reintroduce the 'at least two independent implementations'
SHOULD from the previous version of the charter.
[12] https://github.com/w3c/charter-timed-text/pull/81
Nigel: Apple had a different take on it.
… They really want to push Charters to have a stronger
requirement for CR exit than the minimum
… required by the Process, and I think in time they would like
to raise that minimum in the Process too.
… So they're definitely unhappy about the idea that Content
alone, even though it may have come from an
… implementation, as they see that, might be one of the
factors.
… However, given that they have previously approved the old
wording, they said if we have the
… old wording as well as the new wording, that would be hard
for them to object to.
… They want to get W3C to a place where there are two
implementations that do the intent of the spec,
… e.g. for a caption format, paints pixels on a screen.
Gary: They also understood our wording differently - what does
"Content" mean,
… what does "validating implementation" mean. The old wording
is tried and tested.
Cyril: Sorry if I'm rehashing, but the suggestion from Apple
doesn't seem different from what we have.
… The Google change is clear about two independent
implementation, not necessarily of the same type.
… The second thing: I'm not sure we're in disagreement.
… If I understand correctly it's W3T to verify implementations
- they don't have to be open source,
… or on the market. That to me is equivalent to providing the
content, because a tool has been used
… to provide it. I don't see how the proposal is any different
to what we have.
Nigel: You reminded me of another key point of discussion:
… We haven't distinguished between content made for test by the
WG and
… real world content made by a bunch of other implementations
made by non-WG members.
Cyril: Chicken-and-egg - implementers want spec maturity before
making content.
… [asks a q about content]
Nigel: I think a large cohort of real world IMSC content
provided by multiple providers,
… all passing the IMSC HRM, would not be adequate from their
perspective.
… They want the spec to be shown to be implementable with the
same results based on the spec text
… by more than one person independently.
Pierre: I think it's bad to force Process changes via Charters.
… My interpretation is that they don't value the existence of
documents as a means of proving interop.
… They want multiple processors independently made. That's the
bottom line, right?
Nigel: Yes
Pierre: I don't agree with that personally.
Gary: Their issue with content is that to them it doesn't
represent someone reading the spec and
… understanding it, necessarily. They could have thrown a bunch
of angle brackets in a file
… and it just happened to work.
Pierre: I could argue the same about processors!
Gary: That was their issue with the validating implementation
bullet.
… We had a higher model of what a validating implementation
needs to be.
Pierre: My guess is if we accept Apple's input we will have
that fight again later.
Gary: You're right but we likely have a better story for HRM
particularly.
Andreas: Clarification: if you have an implementation built to
create samples, that's an implementation?
… Also, about the normative keywords, there is no MUST in their
PR right? It's a SHOULD.
Gary: Yes. They would prefer a MUST but are bringing back the
status quo.
Andreas: Possibly that hints at what Pierre said that they may
come back to this later.
Nigel: There are some options here:
… 1. Accept their PR
… 2. Reject their PR and tighten up the bullets to meet their
needs
… 3. Wait for them to come back with potential alternative
changes that would satisfy them.
Gary: They did say they would also circle back to the process
discussions.
Nigel: Of those, does anyone think we accept their PR, so we
can just move on?
Cyril: I need time to digest that.
… We've extended to when?
Gary: End of June
… If we don't come to a conclusion by then we'll have to extend
again.
Pierre: Maybe to leave us, what do you and Gary recommend,
having been in the discussions?
Nigel: Right now, personally, my recommendation is to think
about it. The PR was opened 21 hours ago. Plus there may be
other alternative options.
Gary: Right, there's no rush yet. Better to sit on it for a
bit.
Pierre: Thank you
Meeting close
Nigel: Thanks everyone. Regrets from me for the next call.
… [adjourns meeting]
Minutes manually created (not a transcript), formatted by
[13]scribe.perl version 185 (Thu Dec 2 18:51:55 2021 UTC).
[13] https://w3c.github.io/scribe2/scribedoc.html
----------------------------
http://www.bbc.co.uk
This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated.
If you have received it in error, please delete it from your system.
Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately.
Please note that the BBC monitors e-mails sent or received.
Further communication will signify your consent to this.
---------------------
Received on Thursday, 12 May 2022 16:20:03 UTC