{Minutes} TTWG Meeting 2020-05-07

Thanks all for attending today's TTWG meeting. Minutes can be found in HTML format at https://www.w3.org/2020/05/07-tt-minutes.html


In text format:

   [1]W3C

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


                Timed Text Working Group Teleconference

07 May 2020

   [2]Previous meeting. [3]Agenda. [4]IRC log.

      [2] https://www.w3.org/2020/04/30-tt-minutes.html

      [3] https://github.com/w3c/ttwg/issues/111

      [4] https://www.w3.org/2020/05/07-tt-irc


Attendees

   Present
          Andreas, Atsushi, Cyril, Gary, Pierre

   Regrets
          -

   Chair
          Gary, Nigel

   Scribe
          nigel

Contents

    1. [5]This meeting
    2. [6]IMSC 1.2
    3. [7]IMSC 1.2 - ARIB liaison
    4. [8]IMSC 1.2 - how to make progress?
    5. [9]WebVTT
    6. [10]TTML2 2nd Edition Implementation Report
    7. [11]AOB TPAC Planning
    8. [12]Meeting close

Meeting minutes

  This meeting

   Gary: Today we have IMSC 1.2 - more discussion on the ARIB
   liaison.
   … WebVTT, more on the Unicode normalisation, though I haven't
   had chance to follow up.
   … Then TTML2 2nd Ed IR, Test Suite and a test to talk about
   … in AOB, more TPAC planning, and are there any more AOB items
   to discuss today?

   Nigel: Atsushi reminded me we need to fill in the form, so even
   if we don't get to discuss
   … it today I will try to fill in something based on the
   discussion we have already had.
   … Also IMSC - how can we make progress?

   Pierre: Yes, I would like to discuss that. We are stuck because
   of the HR business
   … and we need to take decisive action.

   Nigel: OK let's cover that in IMSC 1.2

  IMSC 1.2

   Nigel: Two things today: status of the ARIB liaison and our
   response, and how we make progress.

  IMSC 1.2 - ARIB liaison

   Nigel: On the ARIB liaison, we left it before that Atsushi
   would make an informal contact,
   … and I think if I understand correctly that the result is that
   we should raise GitHub issues.

   Atsushi: In summary we can copy/paste their liaison comments
   into GitHub and discuss
   … them there. ARIB can't discuss them from there, so they would
   like a reply in liaison form.

   Nigel: OK then someone needs to create the issues.

   Atsushi: Yes, I will do that.

   Pierre: I also could do it, I don't mind.
   … So we have confirmed that they are happy to make the issues
   public even though the
   … liaison is confidential?

   Atsushi: I specifically pointed to the public repo and asked if
   it was okay. They are fine
   … for their comments to be made public.

   Pierre: Perfect.

   Atsushi: They need to conclude discussions internally
   collecting all comments from members
   … so that is why it is hard for them to contribute one by one
   on GitHub.

   Nigel: Did you get any sense of whether or not they might want
   to adopt IMSC if their
   … comments were addressed?

   Atsushi: They only said they want to get a reply.

   Nigel: OK. Anyway it is great news that they want to engage
   with us.

  IMSC 1.2 - how to make progress?

   Nigel: It seems that Pierre and I (at least) have the feeling
   that we are a bit stuck with
   … the HR issues.

   Pierre: What's on my plate is doing an introduction but we
   decided to do that once we
   … had conclusion on the other issues. I've not seen any review
   on PRs, for example.
   … I have not seen any reaction from the APA to our proposed
   resolutions.
   … We should avoid making changes that make noone happy.
   … One way would be to call a meeting and go through the issues.
   I think you proposed
   … that before and they did not take you up on it?

   Nigel: I'm not sure from recollection.

   Nigel: We moved all the privacy issues to TTML2?

   Pierre: Correct. There's a TTML2 PR open and a comment that it
   ought to be generalised.
   … I don't know how to generalise it, but it is a TTML2 issue.

   Nigel: We have 5 open issues from APA for IMSC 1.2 and one for
   a future version.
   … There are 6 pull requests open.

   Pierre: The relevant ones are #526, #527 and #542 that are APA
   related.
   … For example #527 has a summary from us saying we want review
   by APA, from April 2.

   Nigel: I'm sure they're super busy but this is really slow for
   us.

   Pierre: I'm happy to reach out to the Chair of that group and
   to Philippe and to say this
   … is not a good way of working for us.
   … My recommendation is to ignore the issues and proceed. Making
   changes without the
   … commenter being involved is terrible. We will end up with the
   worst of all worlds.
   … If they don't have time the safest thing is to move ahead and
   deal with it later when
   … they have time.
   … I'm happy to craft an email or to have the Chairs or all of
   us do it.
   … The more I think about it this is really the issue - if they
   don't have time we should defer
   … and move on.

   Nigel: This is no fun for us - we have vNext stuff to consider
   and we're in stasis waiting
   … to do it.

   Nigel: Any other views?

   Andreas: I agree

   Nigel: I think this is probably better coming from the Chair.

   Pierre: Happy to review that email and please CC me.
   … I don't think any of the issues is fatal so it is fine to
   defer them.
   … I'm really worried about the introduction myself.

  WebVTT

   Gary: I did not get a chance to make progress on the Unicode
   issue so unless there are
   … thoughts from others I think we can skip it today.

  TTML2 2nd Edition Implementation Report

   Nigel: We have tests and the IR itself here.
   … Glenn merged a pull request on the tests recently, and
   there's one more outstanding one.
   … That is w3c/ttml2-tests#235, which I reviewed on 15 Nov, and
   reminded Glenn about
   … again a week ago and earlier today.

   Cyril: If you look at line 16 the same comment would apply.

   Nigel: Oh yes, I didn't notice that.

   Cyril: Your comment is this is not testing the xml:id, but the
   semantic about nesting?

   Nigel: Yes

   Cyril: Makes sense to me, yes.

   Nigel: I've added that comment to the pull request.

   Cyril: What about the references then, if you change the inner
   xml:id, what about the used
   … attribute value? Oh actually line 14 has
   use="#nestedContentProfile" and there's a comment
   … that it's invalid as a fragment identifier.

   Nigel: Now I'm confused about what this is trying to test.

   Cyril: Yes he seems to be testing multiple things.
   … Maybe we should have nested profile elements without xml:id,
   and test that nesting
   … is not allowed.
   … I don't know if we need to test twice the same id, it doesn't
   make sense.
   … And then test that a `use` attribute refers to an absent
   profile element (in the invalid case).

   Nigel: Thanks for the input, I'm not sufficiently confused that
   I will go back and check the
   … spec change that was made and what this test is trying to do.
   …
   … Moving on to the Implementation Report,
   … last time Cyril you volunteered to do a pass over the IR
   assuming that the tests would
   … be as they are if these pull requests get merged. I'm not
   sure if you managed to do that?

   [13]TTML2 2nd Ed Implementation Report

     [13] https://www.w3.org/wiki/TimedText/TTML2SecondEditionImplementationReport


   Cyril: I haven't had time yet, but I will do it.

   Nigel: The next agenda item here is w3c/ttml2#1207 which is to
   add more examples
   … to the Introduction.
   … Last time I put this on the agenda when Glenn had registered
   regrets he objected to it
   … being discussed in his absence. He's not present today and
   wasn't last week when it was
   … on the agenda, and did not object again, so I think we need
   to make progress.
   … Cyril reviewed this a week ago and I resolved the comment
   about trailing whitespace.

   [14]TTML2 Pull request #1207

     [14] https://github.com/w3c/ttml2/pull/1207


   Nigel: Are there any other views on this pull request?
   … This pull request was opened 3 weeks ago and had some review.

   Andreas: I didn't review the complete example in detail but in
   general it is super helpful
   … to have text and examples that clarify that putting region on
   body isn't always good.
   … People don't go into the detail about content being pruned if
   it has a different region,
   … to body in this case, so I think it is good to try to clarify
   and fix this.

   Nigel: Thank you for that.

   Cyril: I'm just reading the paragraph in the intro. It says
   each paragraph is flowed into
   … "its targeted region" but it means its own one but here
   there's only one. It should say
   … "each paragraph is flowed into the target region" not "its
   target region" - it is misleading.
   … This is in the existing text, not what's in the pull request.

   Pierre: TTML2 has an introduction, that has an example that's
   not practical - misleading
   … is too strong a word, so this is an attempt to correct that
   by including a more practical
   … example. Is that correct?

   Nigel: Yes.

   Pierre: I'm wondering if the original example should be
   removed. If the problem was the
   … first example that is not incorrect but also not
   representative. Should it be removed and
   … replaced with this more complete example?

   Nigel: I wouldn't object but I suspect Glenn might.

   Pierre: As long as the new example is correct, it's fine with
   me.

   Nigel: Obviously I believe it to be correct but it would be
   helpful to get extra review.
   … [shows pull request]

   Pierre: This is a more advanced example?

   Nigel: Yes

   Pierre: If anything the complexity may make people think a bit
   more carefully.
   … "The location of content spatially is independent of its
   timing." strikes me as incredibly
   … abstract and mind-blowing.
   … The way I've tried to describe this is that the author
   defines region and selects what text
   … and when the text flows into those regions.

   Nigel: Yes

   Pierre: So there are multiple steps. First define regions, then
   specify text, and then specify
   … when that text flows in what region. This is how I try to
   summarise when people ask
   … how this all works. Figure out which region text in a
   document flows into, and then figure
   … out when, separately.
   … Looking at this, I'd be tempted to replace the first sentence
   by something along those
   … lines. Then the example illustrates that.

   Nigel: I'm happy to try to make that change.

   Andreas: I also agree, I found that first sentence a bit
   confusing.

   Pierre: It's true, right, you can specify the position and
   timing independently, but that is
   … incredibly abstract, but for people especially with
   experience of, say, 608, will understand
   … the concept of regions.
   … To answer the question, I don't have any objections to adding
   examples at this point.

   Nigel: Thanks all for that.

   Andreas: I also think it is useful and gives a good idea of how
   you could use TTML and
   … some of the basic patterns you can apply.

   Nigel: Moving on due to time, that's really positive, thank
   you. The action is with me
   … to do another editorial pass on the first paragraph first
   sentence.

  AOB TPAC Planning

   Nigel: I think this is for Chairs now, and we have good input
   from members on this from
   … previous discussions. I may also have trouble getting funding
   for travel now, even if
   … there were no potential health implications.

   Pierre: As another data point, in another group I chair, we
   have had long discussions about
   … this and concluded that we should try to arrange the best
   possible virtual meeting and
   … if somehow a physical meeting becomes possible, deal with
   that then.
   … Let's make the best out of it.

   Nigel: Agreed.

   Pierre: Especially with CSS-WG we should try to make an effort
   to meet virtually together
   … soon, regardless of TPAC, to coordinate our efforts on Ruby,
   text emphasis etc.
   … For example ruby overhang, if they want to discuss that.
   … tate chu yoko was another one.

   <atsushi_> (just these reminded me that I should work on JLReq
   more...)

   Nigel: Action for me and Gary to work out how to fill the form
   in.

   Gary: If you want to, feel free to fill it out, otherwise I
   guess if you want my feedback you
   … could put it in a doc and I can look it over and add comments
   or something.

   Nigel: Okay, I'll fill it in, get what I wrote emailed back to
   me, share that with you, and if
   … there are any changes to make we can go back in and change
   it.

   Gary: That works.

  Meeting close

   Nigel: Thanks everyone, we're 2 minutes over so I'll adjourn.
   See you next time. [adjourns meeting]


    Minutes manually created (not a transcript), formatted by
    [15]scribe.perl version 114 (Tue Mar 17 13:45:45 2020 UTC).

     [15] https://w3c.github.io/scribe2/scribedoc.html