W3C

Timed Text Working Group Teleconference

30 Apr 2015

See also: IRC log

Attendees

Present
nigel, Andreas, courtney, pal, plh
Regrets
dsinger
Chair
nigel
Scribe
nigel

Contents


<trackbot> Date: 30 April 2015

<scribe> scribe: nigel

This meeting

nigel: AOB?

group: no AOB

Action items

action-390?

<trackbot> action-390 -- Nigel Megitt to Update the multirowalign and linepadding ttml files to have shorter lines of text -- due 2015-04-17 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/390

action-391?

<trackbot> action-391 -- Nigel Megitt to Update the multirowalign and linepadding pngs to match the ttml -- due 2015-04-17 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/391

nigel: I was hoping that Andreas could review these before I close them.

Andreas: I had some problems with Mercurial but this is on the top of my list to review. I'll come back next week with these.

https://www.w3.org/wiki/TimedText/IMSC1_Implementation_Report#Tests

https://dvcs.w3.org/hg/ttml/file/tip/ttml-ww-profiles/testsuite/linePadding/linePadding1.ttml

https://dvcs.w3.org/hg/ttml/file/bb46ee1c9c18/ttml-ww-profiles/testsuite/linePadding/

nigel: You can browse directly using the repository - see above links.
... I should possibly remove the .png files that aren't 16x9 or 4x3 because they're wrong and not referenced anywhere.

action-392?

<trackbot> action-392 -- Pierre-Anthony Lemieux to Tweak the examples with multiple regions so that they appear in a more intuitive order. -- due 2015-04-17 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/392

nigel: Pierre, you've adjusted the multiple region examples?

pal: Yes.

close action-392

<trackbot> Closed action-392.

action-393?

<trackbot> action-393 -- Nigel Megitt to Send pal details of implementation and tests passed so he can create a table in the implementation report -- due 2015-04-17 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/393

nigel: We discussed this last week and plh uploaded the videos. I think that's all done now.

<plh> http://media.w3.org/2015/04/

nigel: https://www.w3.org/wiki/TimedText/IMSC1_Implementation_Report#Results
... The two BBC prototype tests are now on the above results page.

close action-393

<trackbot> Closed action-393.

action-400?

<trackbot> action-400 -- Nigel Megitt to Send smpte dispositions from action-378 to smpte (mdolan) -- due 2015-04-30 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/400

nigel: I sent Mike our dispositions on the SMPTE email on Monday. Just for info, when David Singer
... reviewed the draft he noticed that the TTML2 features may be subject to further change when we
... go to wide review, so we added an extra sentence in the introduction to highlight that point.
... Obviously Mike already knows it anyway.

close action-400

<trackbot> Closed action-400.

nigel: There are lots of open actions remaining - does anyone want to call out any specifically that may
... have had some work but not been updated on the tracker?

action-389

<trackbot> action-389 -- Andreas Tai to Check with ebu group if we can have a combined event at ibc with ebu group and imsc/w3c -- due 2015-04-17 -- OPEN

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/389

andreas: We are having internal discussions on if we can have joint presentations at IBC for IMSC and EBU-TT-D.

action-398?

<trackbot> action-398 -- Andreas Tai to Contact erik lindstrom re htmlcue -- due 2015-04-17 -- OPEN

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/398

andreas: I wrote to Erik re the generic HTMLCue idea and he responded today that there has been
... some ongoing work with Glenn and it was considered for use in blink. The idea was to transform
... TTML documents to ISDs and then generate HTML code from them. That was the basic idea. Maybe
... the document itself shouldn't be proposed directly to the group but could be the basis for a proposal.
... The next thing is to check if Opera or Erik is able to work with someone else in the group on that
... proposal. I understand this is something we can discuss with the HTML WG at the next TPAC meeting.
... Possibly this group should scope more what our requirements should be for this internal cue concept.

close action-398

<trackbot> Closed action-398.

Andreas: The idea was that we make some kind of proposal for the HTML WG. I think we decided we
... want to go ahead with this. Is that right?

pal: Yes, I think there was broad consensus that it would be a good idea.

andreas: I think so; I can imagine that Erik or Opera would be willing to participate in this if there are
... other collaborators, and to use their input as a base. I'm certainly interested but have not enough
... knowledge at the moment to volunteer.

pal: This has been a long running project. As I recall, Glenn had put a lot of effort into trying to get this
... done and the obstacles were not technical. So do you feel that those obstacles have been overcome
... and this is mainly technical work? I'm just trying to understand the status. I'd love to be able to help.

andreas: Mostly my information is based on what Glenn said. I think the main obstacle is time on Glenn's
... side, that was my interpretation. This hasn't yet been proposed to HTML WG but was a ticket raised
... on Chromium, but it didn't get positive feedback.

pal: So the current strategy is to write an HTML extension and submit it to the HTML WG. I'm happy to help.
... Is there a draft? Do we have a beginning?

andreas: I'm not sure. He brought Opera's proposal in as a good start. Possibly we can say that pal and I
... are willing to help and if we can work with Erik then we can plan on how to push this forward.

pal: Maybe we can take this on with just the 4 of us and see where we go.

nigel: Great, we need to turn this into actions. Philippe, what's the normal way to publish HTML extensions?
... Do we have to do it as a Recommendation, or is a Note okay?

plh: There's no standard way to do it. It could be published by HTML WG under the Media Task Force
... or here - if you want it to be a Recommendation from this group then you need to update the charter
... and send it to AC for review. It's normal for HTML extensions to be Rec track documents.
... This reminds me that the Media Task Force met 2 weeks ago and discussed transcripts. There is
... interest there to do further work on transcripts.
... Figure out where the best place is to work on it, in HTML WG or here, and then adjust the
... charters as needed. Adding a deliverable to a charter isn't too difficult to do.

nigel: It seems to me that we need two pieces of work: one to define the generic HTMLCue structure
... and the other to define the mapping from TTML/ISD to HTMLCue.

Andreas: Do you think the HTML WG would define the HTMLCue and and we would define the mapping?

nigel: Yes, that's what I was thinking. Even if we write the HTMLCue spec and give it to HTML WG as an
... extension that would be fine.

pal: I tend to agree with nigel that for both implementation and correctness the HTML WG should publish
... the generic HTMLCue part.

plh: The best thing is to pitch it to the Media Task force and the HTML Accessibility Task Force. They
... both work on media related issues. Once you guys have a proposal I'm happy to help.

nigel: I'd like to make an 'in principle' proposal first before we put in loads of effort that would be
... rejected.

plh: Certainly as much detail on use cases and proposed approaches for solving them would be useful.

<scribe> ACTION: atai2 Kick off discussions with Erik, Glenn, Nigel, Pierre on HTMLCue [recorded in http://www.w3.org/2015/04/30-tt-minutes.html#action01]

<trackbot> Created ACTION-401 - Kick off discussions with erik, glenn, nigel, pierre on htmlcue [on Andreas Tai - due 2015-05-07].

Action-394?

<trackbot> Action-394 -- Courtney Kennedy to Determine who is responsible for putting together the dispositions of comments document and get a schedule from them. -- due 2015-04-17 -- OPEN

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/394

Courtney: I talked to dsinger - clearly this is under way and he's actively running it. I didn't get a schedule from him.
... I'm happy with the progress report so far. He sent a spreadsheet - he's going to update it and maybe
... make it into a web page. Other than that everyone's aware - it's all been on the public list.

close action-394

<trackbot> Closed action-394.

Issues

issue-382?

<trackbot> issue-382 -- Require a computed non-indefinite begin time -- open

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/382

pal: There's been a lot of offline discussions on this issue. To summarise: there's a difference in opinion
... on exactly how the Root Temporal Extent works and what it means and the interaction between
... TTML1 and SMIL. So the very premise is still murky; therefore it's been difficult to make progress.
... My thinking on this is that I don't think IMSC 1 can do much to resolve ambiguities in SMIL and TTML1.
... If they are there they should be resolved at some point. The best IMSC 1 can do today with the
... least impact is to recommend that begin and end are used on div or p. That does not resolve the
... existing ambiguity but unambiguously captures when elements are active within an IMSC 1 document.
... That would not be a SHALL but rather a SHOULD. In the majority of TTML documents I've seen so far
... begin and end are used on p or div. We can continue to try to dig, and maybe reach a resolution.
... The other is to add a recommendation in IMSC 1 to specify begin and end on ps and divs. EBU-TT-D
... actually requires begin and end on p or span. Or the third option is to do nothing.
... I'm not optimistic that we can resolve the ambiguity.

andreas: I agree with pal on this. I thought we might conclude to recommend the use of timing
... explicitly in the document. So I think that was the second option.

pal: Yes, actually we had discussed a SHALL. I'm not sure that's appropriate because it sounds like
... there are applications where it's correct to omit begin and end times even though the vast majority
... of cases include them.

andreas: In the discussions I've had with people not involved in the specification process they like as
... many SHALLs as possible. If we know there are some problems and there are no clear use cases
... not to specify begin and end I can still see reasons to specify a SHALL not a SHOULD.

nigel: The use case comes from me actually in thinking about live cases. For distribution use cases I
... agree that begin and end times should or even shall be present, but the scope of IMSC is not
... completely restricted to distribution. I would accept a compromise that qualifies the requirement
... with it only applying when distributing the documents in a wrapper format.

pal: I am not sure that the times must always be present so that's why I'm proposing a SHOULD.

andreas: It's always a specification issue whether you know there's a use case or you don't know that
... there isn't a use case. I'd be happy to go with a should if that's what you think.

pal: If we're happy with this then I'd add a simple line under #timing to say begin and end should be
... present on at least one of div, p and span.

nigel: That works for me. You do need to allow for the possibility of timing on span.
... Are you also planning to add a note on the potential ambiguity?

pal: The best I could write is maybe: "if begin remains indefinite then the begin corresponds to the begin
... of the root temporal extent"... but that's where the disagreement is so I'm not sure what to write.

nigel: As far as I know this issue has been raised only by Pierre. If we just make this change then
... maybe if there's an implementation problem we can deal with that as a specific problem to solve.
... That sounds like a resolution to this issue.

pal: When I've done that then that will allow me to move ahead with the CR2 and the list of changes.

plh: I'll help you with that. Are there any substantial changes?

nigel: Yes, we reviewed them in the F2F and minuted them.

plh: If you make substantive changes you have to go to LC.

nigel: That's not what we understood - there is no LC in the new process.

plh: Ah my mistake you're on the 2014 process. We will have to ask for Director permission to go to CR
... again.

IMSC Test Suite and Implementation Report

nigel: My BBC colleagues spotted that the example video on the frameSync test has an edit list that
... moves the frames on by 2, which makes their gstreamer implementation fail.

pal: I've sent you the source for that video - maybe your colleagues could see if they can create an
... alternative version of the video without the edit list.

nigel: Okay I'll see what we can do.

WebVTT review feedback

Courtney: There's nothing more to report on this right now.

TTML2 Editorial Actions

nigel: We're out of time for this meeting but if you'd like to volunteer for any editorial actions please
... email me.
... Thanks everyone! [adjourns meeting]

plh: regrets for next week (AC)

Summary of Action Items

[NEW] ACTION: atai2 Kick off discussions with Erik, Glenn, Nigel, Pierre on HTMLCue [recorded in http://www.w3.org/2015/04/30-tt-minutes.html#action01]
 
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2015/04/30 15:17:24 $