- From: Nigel Megitt <nigel.megitt@bbc.co.uk>
- Date: Thu, 4 Jun 2020 16:41:34 +0000
- To: "public-tt@w3.org" <public-tt@w3.org>
- Message-ID: <17CCB51D-EE84-45FF-9BC6-20EDD4687A2B@bbc.co.uk>
Thanks all for attending today's TTWG meeting. Minutes can be found in HTML format at https://www.w3.org/2020/06/04-tt-minutes.html
In text format:
[1]W3C
[1] https://www.w3.org/
Timed Text Working Group Teleconference
04 June 2020
[2]Previous meeting. [3]Agenda. [4]IRC log.
[2] https://www.w3.org/2020/05/28-tt-minutes.html
[3] https://github.com/w3c/ttwg/issues/118
[4] https://www.w3.org/2020/06/04-tt-irc
Attendees
Present
Andreas, Atsushi, Cyril, Glenn, Nigel, Pierre
Regrets
Gary
Chair
Nigel
Scribe
nigel
Contents
1. [5]This meeting
2. [6]Add consideration for font fingerprinting w3c/ttml2#1203
3. [7]TTML2 Implementation Report
4. [8]IMSC 1.2 CfC to request transition to PR
5. [9]IMSC/ARIB issues
6. [10][WR/ARIB] Compatibility with ARIB-TTML / 1. Image
handling imsc#546
7. [11][WR/ARIB] Compatibility with ARIB-TTML / 2. Font
handling imsc#547
8. [12][WR/ARIB] Compatibility with ARIB-TTML / 3. Animation
imsc#548
9. [13][WR/ARIB] Compatibility with ARIB-TTML / 4. Audio
presentation imsc#549
10. [14]Meeting close
11. [15]Summary of resolutions
Meeting minutes
This meeting
Nigel: I'd like to wrap up the TTML2 font fingerprinting issue,
and there's a placeholder
… for the TTML2 IR too.
… Then we can continue to iterate through the IMSC issues from
ARIB.
… I think that's it. AOB, or points to make sure we cover?
Atsushi: We could discuss the IMSC branch protection rules?
Nigel: OK if we have time at the end
Add consideration for font fingerprinting w3c/ttml2#1203
github: [16]https://github.com/w3c/ttml2/pull/1203
[16] https://github.com/w3c/ttml2/pull/1203
Nigel: 2 things outstanding.
… 1. Change "Font Matching" title to "Font Detection"
… Any views against calling the section "Font Detection"?
Glenn: Fine with me
Pierre: No objection, and I'm happy for me or someone else to
make that change.
… I can do it right now.
Resolution: Rename "Font Matching" to "Font Detection".
Nigel: The next one is about other detections using
`condition`.
… First question: any counter-views to my recent conclusion
that condition does not
… provide any further ability to detect fonts than what is
already written in this pull request?
Glenn: Note we can conditionalise on user's preference for
language, which may not apply
… to, say, CSS media queries.
Pierre: Yes, you could conditionalise an image based on
someone's user preference so
… you could determine user preference based on the request for
an image.
Glenn: It would have to be heuristic based and it is possible
to fail a variety of modes.
Pierre: Yes, and it doesn't change my overall perspective that
trying to solve this complex
… topic that potentially requires coordination, normatively, at
the last minute, would lead to
… a mistake.
… I would keep it simple.
Nigel: I am not suggesting a normative change.
Pierre: I don't object to adding condition but someone would
have to propose the text.
Glenn: My preference would be to defer treating condition,
which is something Nigel
… raised. If we try to agree on it now then we need language,
and agreement on it, and
… that will open up the discussion further. This feeds into
Pierre's point that it is a wider
… broader issue and we're likely to get it wrong if we address
it at the last minute.
Pierre: We could make a generic statement. As soon as a
document allows access to external
… resources the fingerprinting risk increases.
Glenn: That depends on the delivery mechanism, a prepackaged
carousel vs on demand.
… It depends on the context to know if it is an issue.
Pierre: Yes and on top of that the particular implementation
provides heuristics.
Glenn: That's why I think it's in the application environment,
outside the scope of TTML.
Nigel: Focusing tightly on this pull request, is there some
additional fingerprinting risk
… for font matching derived from condition?
Glenn: I pointed out that user preference language can be used,
which can be used to make
… decisions about font defaulting.
… Let's say the user's default is Mongolian, for example. An
implementation might look at
… the font list and throw out any fonts that don't support
Mongolian, for example, and never
… do fetches on them.
Nigel: And that would be caused by the condition attribute and
not by the font element?
Glenn: Let's say you conditionalised a style element or
declaration based on user language
… preference and then that style happens to be the one that
specifies a fontFamily attribute.
… You might have different values for fontFamily depending on
the condition.
Nigel: And that's independent of font, so you could do the same
thing based on fetching
… any external resource even not a font, also based on the
user's language preference.
Glenn: Correct.
… Resource fetching is a known mechanism for communicating
usage back to a server.
… Non-resource-fetching semantics such as presentation without
fetching would have
… no route back to the server.
Nigel: My conclusion is that there may be a need for a new
issue or pull request related
… to condition but it should not block this pull request. Is
that correct?
Glenn: No objections from me.
Pierre: I agree too.
Resolution: Detections based on condition attribute need not be
factored into this pull request.
Nigel: Then thank you Pierre for volunteering to update the
pull request for the section
… title, when that's done I think we can go ahead and merge.
TTML2 Implementation Report
Glenn: I just did a pull request last night on ttml2-tests to
address a message that
… Cyril sent a week or so ago. Also there was an open issue
#238 that Cyril had opened.
… I think I dealt with all of the questions in #238 as well as
the new message. The answer
… to Cyril's question as to whether they belong in the manifest
is Yes, and they were
… missing by admin error on my part. In a couple of cases there
were some bad names.
… I believe I've got that all fixed up, and if you find any
other discrepancies let me know
… and I'll fix those.
Cyril: Thank you, I will update the IR to make sure these tests
are in there.
Glenn: Yes, keep in mind a couple of test names changed too.
Nigel: They're pretty easy to spot in the pull request.
… Thanks for that Glenn, assuming that all looks good I think
we can merge it quickly.
Glenn: The only other thing is moving forward with updating the
IR to list an implementation
… for all the validation tests.
Nigel: Thanks for that. I've already put our implementation on
against the audio presentation test, which passes.
[Glenn leaves]
IMSC 1.2 CfC to request transition to PR
Nigel: Any queries or anything to raise on the CfC?
… The CfC period ends tomorrow.
group: [nothing to raise]
Nigel: Thanks, Atsushi, it looks as though we'll be going ahead
with that transition request.
Pierre: Thank you everybody.
Nigel: Thank you!
IMSC/ARIB issues
Pierre: It would be helpful to have a roadmap, otherwise we
have a risk of going through
… the issues without doing any of the work, and then the
relevance would be lost before
… we do get around to it. I would like us to have a plan for
IMSC, do we want a new version
… in the next year, or change to moving to a working draft
model. I'm interested in everybody's
… input on that.
Nigel: I'm interested in that too, but I think the first thing
we have to do is understand the
… scope of our response to ARIB.
Pierre: Ah, that would help, if we limit our discussion to the
scope of the response to ARIB
… that would be helpful to know.
[WR/ARIB] Compatibility with ARIB-TTML / 1. Image handling imsc#546
github: [17]https://github.com/w3c/imsc/issues/546
[17] https://github.com/w3c/imsc/issues/546
Nigel: I want to know more about the use case for this.
Pierre: I agree this is the essential question.
… Thinking about going back to ARIB, asking for sample
documents or exemplar renders
… would be really helpful.
Cyril: I remember that when we discussed mixing text and image
profile we said we would
… not do it because there are other ways to do it, and that
there's a complexity side to it too.
Pierre: That's my recollection too. My question is do they have
use cases that cannot be
… solved with the current approach.
… Image and Text can mean many different things. It would be
good to understand what they mean.
Cyril: I wonder if we should ask about the complexity of
implementation too.
Pierre: Yes I think that would be great if they could point to
a reference implementation so
… we can understand complexity.
Nigel: Two things for me to add to the list for w3c/ttwg#116:
1. Request more info about use cases; 2. Ask about complexity.
SUMMARY: TTWG would like to know more about the use cases for
this feature, and complexity of implementation.
[WR/ARIB] Compatibility with ARIB-TTML / 2. Font handling imsc#547
github: [18]https://github.com/w3c/imsc/issues/547
[18] https://github.com/w3c/imsc/issues/547
Nigel: This relates a bit to the conversation we had about
putting images into text via fonts.
Cyril: It is not even a feature of IMSC 1.2. Even in IMSC 1.1
or earlier, if you have the right
… font, then you can do it.
Nigel: I've always thought you can only do that if you have a
contract effectively that
… links the document to the font.
Cyril: You can have a closed environment out of band of the
document that allows this to be done.
Nigel: Ah yes, understood.
Pierre: I think this is exactly what they have in mind, an
environment with specified font
… requirements.
Cyril: Even out of band, broadcasting a font alongside the IMSC
document. You could get
… it for free that way, whilst still being outside the scope of
IMSC.
Nigel: I think they are saying that they have an
arib-tt:font-face element that references
… a font delivered externally somehow.
Cyril: They seem to link the ability to do IVS and Gaiji with
this arib-tt:font-face element.
… Maybe in our response we should say that in a closed
environment you don't need
… a font-face element to use IVS and Gaiji.
… We can ask them if they have considered that.
Nigel: It seems that in ARIB-TT and IMSC 1.2 we have both
arrived at a similar solution.
Pierre: Yes, I agree that it looks like the same capabilities
are present and we can ask for
… samples of how they use it.
Atsushi: +1
Nigel: Does anyone have a sense of whether we should add
something to IMSC vNext,
… informatively, noting the similar functionality?
… More widely, is it useful in IMSC to talk about functional
overlap with other profiles of TTML?
Cyril: We already do, with SDP-US or EBU-TT-D.
… Speaking of scope, yes maybe it could be in scope in the next
version to extend the list
… of standards it matches or overlaps.
Pierre: As soon as we have the details to get comfortable with
that, absolutely.
SUMMARY: TTWG is interested to know more about the usage of
these features, and would like to consider noting the
functional overlap in a future version.
[WR/ARIB] Compatibility with ARIB-TTML / 3. Animation imsc#548
github: [19]https://github.com/w3c/imsc/issues/548
[19] https://github.com/w3c/imsc/issues/548
Cyril: Nigel noted it matches #animation-version-2.
… I think our answer is similar to #546 that we would like to
have use cases and examples.
Nigel: And I assume complexity is an issue here too.
Cyril: Yes
Nigel: Looking at it from a spec development perspective it
would be easy to add, but
… a lot of work for implementers.
Pierre: That's an understatement. We need to know more details
about what support is
… needed, especially from an interop perspectivee.
SUMMARY: TTWG is interested to know more about the usage of
these features, and complexity of implementation.
[WR/ARIB] Compatibility with ARIB-TTML / 4. Audio presentation
imsc#549
github: [20]https://github.com/w3c/imsc/issues/549
[20] https://github.com/w3c/imsc/issues/549
Cyril: I agree with @nigelmegitt's comment and would like to
know more about the use cases.
Nigel: Behind my question is "what is the use case specifically
for playing audio back as
… part of a subtitle document?"
… I obviously have been working on an AD profile, but my
assumption has always been
… that it can be an independent resource from the subtitle
document.
… As an example, I'd be interested to know if they want to send
audio renderings of the
… subtitle text as something like a pre-recorded version of
speech to text, or if they are using
… it for audio description, or some other thing.
SUMMARY: TTWG is interested to know more about the usage of
these features.
Meeting close
Nigel: Thanks everyone, we're out of time! See you next week.
[adjourns meeting]
Summary of resolutions
1. [23]Rename "Font Matching" to "Font Detection".
2. [24]Detections based on condition attribute need not be
factored into this pull request.
Minutes manually created (not a transcript), formatted by
[25]scribe.perl version 117 (Tue Apr 28 12:46:31 2020 UTC).
[25] https://w3c.github.io/scribe2/scribedoc.html
Received on Thursday, 4 June 2020 16:41:54 UTC