{Minutes} TTWG Meeting 2019-08-15

Thanks all for attending today’s TTWG meeting. Minutes can be found in HTML format at https://www.w3.org/2019/08/15-tt-minutes.html

In text format:


   [1]W3C

      [1] https://www.w3.org/

                Timed Text Working Group Teleconference

15 August 2019

   [2]Agenda. [3]IRC log.

      [2] https://github.com/w3c/ttwg/issues/55
      [3] https://www.w3.org/2019/08/15-tt-irc

Attendees

   Present
          Cyril, Glenn, Nigel, Pierre

   Regrets
          Andreas, Atsushi, Gary

   Chair
          Nigel

   Scribe
          cyril

Contents

     * [4]Meeting minutes
         1. [5]This meeting
         2. [6]TTML2 issues
         3. [7]Add a syntactic shortcut for an implied audio
            element. ttml2#1013
         4. [8]Constrain use of @type on data element for
            reference data embeddings. ttml2#1022
         5. [9]TTML1 tests
         6. [10]TTML Live Extensions Module
         7. [11]Charter status update
         8. [12]TPAC Planning
         9. [13]meeting close

Meeting minutes

   <nigel> nigel has changed the topic to: TTWG weekly webex.
   Today 1500 UTC. Agenda for 2019-08-15: [14]https://github.com/
   w3c/ttwg/issues/55

     [14] https://github.com/w3c/ttwg/issues/55

   <nigel> Log: [15]https://www.w3.org/2019/08/15-tt-irc

     [15] https://www.w3.org/2019/08/15-tt-irc

This meeting

   nigel: we are very few
   … I pull out a couple of substantive issues for TTML2
   … Test work
   … TTML Live extension module draft got pushed a couple of days
   ago
   … AOB: Charter update
   … any AOB?

TTML2 issues

Add a syntactic shortcut for an implied audio element. ttml2#1013

   <nigel> github: [16]https://github.com/w3c/ttml2/issues/1013

     [16] https://github.com/w3c/ttml2/issues/1013

   nigel: I assume Glenn that there is nothing to do except
   implement it editorially

   glenn: I marked it substantive because ...
   … it's asking for a default behavior that is not currently the
   case

   nigel: it's implemented in one place at least as requested
   … but the spec change is probably substantive, I agree

   glenn: you're asking that if no src is present, it default to
   speech
   … ,. and even if there is no audio

   nigel: when speak is used

   glenn: you could have a top level body element with an audio
   child

   nigel: if you add in an audio element referring to the speech
   at the body level, you can't independently control the gain of
   the audio programme down in the hierarchy
   … that's a common use case, duck the programme audio but not
   the speech
   … you need to bring the speech synthesis further in the
   document tree
   … but it becomes very verbose

   glenn: if you wanted to alter the gain on a leaf span, with
   speak=normal
   … could you not put gain on that span and it would control the
   gain of what got contributed and picked up at the high level

   nigel: no it would control the gain of everything mixed in
   … not the differential gain

   glenn: I think to study it more and start implementing a PR
   … I agree that right now it's not concretely defined
   … given the minimum implementations out there, if you have one
   implementation that does that, it is probably the right way to
   go

   nigel: we may have tests for that

   glenn: you contributed those audio tests

   nigel: yes
   … I'll check

   <nigel>
   [17]ttml2-tests/presentation/valid/ttml2-prstn-audio-speak-on-s
   pan.xml test

     [17] https://github.com/w3c/ttml2-tests/blob/master/presentation/valid/ttml2-prstn-audio-speak-on-span.xml

   nigel: this one does not contain audio
   … and that test passed in our implementation, it did speak it
   … the test matches what I'm asking to be specified now

   glenn: I'll add that to my work list

Constrain use of @type on data element for reference data embeddings.
ttml2#1022

   <nigel> github: [18]https://github.com/w3c/ttml2/issues/1022

     [18] https://github.com/w3c/ttml2/issues/1022

   nigel: raised by Glenn initially, edited by me
   … I thought this was merged

   glenn: I'm not sure if the constraint on the use of type has
   been
   … it may be that there is no further action
   … I need to check
   … I'll post a PR
   … if needed

   nigel: can you add details in the issue about what was merged
   already

   glenn: yes

TTML1 tests

   nigel: there was an issue 1, to add a readme and populate the
   repo
   … glenn suggested to add TTV tests
   … in the context that these tests are not CR tests
   … glenn did and you cannot identify their source
   … there is a comment regarding pending tests waiting approval
   … there are lots of validation and presentation tests, a script
   as well

   glenn: basically, I used the same structure that we have in the
   TTML2 test repository
   … I tweaked the readme file a bit
   … but it's the same structure and convention
   … the tests themselves don't identify the source
   … use the appropriate license
   … the next thing I need to do is take that test from IMSC on
   region timing test
   … and put that in using the same conventions
   … it's on my list
   … I want to review the original test suite
   … to make sure that these tests include the same content
   … since we did not have validation test concept in TTML1
   initially
   … there are about 300 tests that I added, coming from TTV and
   TTPE

   nigel: I agree we need to check that the TTML1 CR initial tests
   are in
   … we have a bunch of IMSC tests and they are a subset of the
   TTML1 tests
   … perhaps we should point to them too

   glenn: the IMSC test suite exist independently
   … it focuses on IMSC features that are new
   … as well as on the constraints model that IMSC puts in place
   … we could incorporate some tests from there with minor
   massaging
   … there is probably some redundancy that is not harmful
   … we'd have to look at the IMSC tests on a case by case basis
   … to make sure that there is nothing that would be problematic
   for TTML1 processors
   … in theory they should not be
   … moving forward we can incrementally add new tests

   nigel: any other question?

TTML Live Extensions Module

   <nigel> [19]TTML Live Extensions Module README

     [19] https://github.com/w3c/tt-module-live/blob/master/README.md

   nigel: I have pushed (after Cyril's approval to get past the
   branch protection) 3 documents
   … we can use them as a basis to raise issues
   … I have some editor's notes to make changes in the guide
   … the TTML live extension module is done, pending comments
   … also the TTML live carriage over websocket
   … the one that does need some work is the guide
   … this structure of having a guide separate from the main spec
   … is based on the request from the group to pare down the
   specification only to the normative parts
   … I'm interested if more needs to be pruned out or more added
   … we can shift content between documents
   … I plan to bring examples from the EBU document into the guide
   … I'd like to add this to the agenda for TPAC
   … so people have about a month to review it

   pal: is there a place where one can find a link to all TTML
   repos
   … I'm looking at the wiki, under publications, and can't find
   it

   <nigel> [20]TTWG Home page including repos

     [20] https://www.w3.org/AudioVideo/TT/#recent-activity

   nigel: the home page contains all the repos
   … I need to add the Karaoke, the Live module, ...

   pal: it'd be good to have all the modules on which we are
   working

   <glenn> try [21]https://github.com/
   search?q=org%3Aw3c+tt&unscoped_q=tt etc

     [21] https://github.com/search?q=org:w3c+tt&unscoped_q=tt

   <nigel> [22]Action on Nigel to add new repos to the home page

     [22] https://github.com/w3c/ttwg/issues/65

   nigel: in terms of details regarding the live extension, I
   created the new features for the live extensions in the TTML
   features namespace
   … I didn't do in the EBU namespace or some other namespace
   … I just use the feature namespace directly

   cyril: I did the same in the karaoke module

   glenn: so there is the definition of the underlying features
   (element types, attribute types) and you used the existing
   namespaces for that?

   nigel: I was talking about the feature designation

   glenn: you put that in the standard features namespace not the
   extension namespace

   nigel: right

   glenn: and about the underlying features?

   nigel: I kept them in the source EBU namespaces
   … there are only attributes, no new elements

   glenn: that might be a problem
   … to put the feature designation in the standard namespace
   while using non-TTML namespaces for the underlying features

   nigel: I did not think it would be a problem, but please review
   and explain why it would be a problem

   pal: EBU is giving up the control of those elements?

   nigel: yes, for this work

   pal: it'd be good to have confirmation for that

   nigel: not everything in EBU namespaces would move under W3C's
   control

   glenn: if you propose to include underlying features defined
   under EBU controlled namespaces, it makes me uncomfortable

   nigel: anything under these specifications is being contributed
   by EBU
   … no doubt in my mind
   … if we need to extend in the future, we might want to add it
   in the same namespace, but EBU might not be happy with that
   … is that the problem?

   pal: I'm not sure I see a scenario where we need to use names
   for other things that those that were transferred
   … extensions will be in W3C namespaces
   … I don't see a risk
   … but we need a formal communication from EBU explicitly saying
   we transfer control of these elements/attributes/features

   nigel: I agree
   … to glenn's point, I feel that the namespace of the feature
   designator can be separate from the namespace of the referenced
   vocabulary

   glenn: one thing that would make me more comfortable is if you
   used the extensions namespace instead of the features namespace

   nigel: that's not the point of how we do modules

   glenn: [explaining extension catalog]
   … I don't have a definitive comment to make right now
   … need to review carefully

   nigel: exactly

   glenn: can you include in your explainer the rationale for
   using the TTML namespaces

   s/for using/for not using/

   <nigel> for _not_ using the TTML namespaces for the vocabulary

   nigel: to preserve existing implementation

   glenn: also add the rationale for using the features namespace
   and not the extensions one

   nigel: because we are defining the module

   pal: on that point, that's for me interop standpoint
   … there would be no downside in having it in the extensions
   namespace

   nigel: I can't think of a problem

   pal: there is a features and extensions bucket

   nigel: I consider them separate by namespace only, no
   difference

   nigel: anyway, my main point was to highlight the fact that the
   specs were here for people to review before TPAC

Charter status update

   <nigel> [23]Proposed Charter (diff)

     [23] https://services.w3.org/htmldiff?doc1=https://www.w3.org/2018/05/timed-text-charter.html&doc2=https://www.w3.org/2019/08/ttwg-proposed-charter.html

   nigel: the thing to note is that the new TTWG charter as been
   put for AC review
   … please encourage your AC rep to look through that and provide
   feedback and vote
   … do look at the details of it
   … it has been tweaked

   pal: can you summarize them?

   nigel: the coordination and review bits
   … there was a discussion about readopting the template text

   pal: I'll do a manual diff

   nigel: the one issue that I had was that the template text said
   that before entering CR you have to do Horizontal Review
   … they clarified that by saying "first entering CR". That 3
   months means we need to ask for HR on new documents very soon.

   pal: if they insist on that, we should do FPWD

   nigel: but maybe we can do the HR on ED
   … any other question?

TPAC Planning

   nigel: we have a wiki page
   … I've not seen any edit
   … do we want to have a meeting in a call the week after TPAC
   … I'll take views, don't have a strong opinion
   … any views?

   cyril: no objection

   nigel: let's cancel the call the week after TPAC

   cyril: ok

   nigel: next week I'm not available
   … if anybody wants to chair let me know
   … otherwise it may get cancelled
   … ping me in the next 24h
   … Gary might be back

   pal: my regrets for next week

meeting close

   nigel: thanks everybody, adjourned


    Minutes manually created (not a transcript), formatted by
    Bert Bos's [24]scribe.perl version Mon Apr 15 13:11:59 2019
    UTC, a reimplementation of David Booth's [25]scribe.perl. See
    [26]history.

     [24] https://w3c.github.io/scribe2/scribedoc.html
     [25] https://dev.w3.org/2002/scribe/scribedoc.htm
     [26] https://github.com/w3c/scribe2/commits/master/scribe.perl

Received on Thursday, 15 August 2019 17:01:11 UTC