See also: IRC log
<scribe> scribe: nigel
Nigel: Hi everyone!
... Today is our last meeting for a couple of weeks. Our next
after today is on 30th August.
... For today, we need to start focusing on tests since there's
no spec work to do urgently,
... just editorial things.
... I think we will cover TTML1, TTML2, IMSC 1.1, CSS stuff. I
don't think we have anything
... to cover with profile registry yet.
... Any particular items to cover, or any other business?
Pierre: I think it's already on
the agenda, but TTML1 3rd Ed tests, and also two
particular
... TTML2 tests for luminanceGain and disparity.
Nigel: OK, thanks.
group: [no other points to raise]
Nigel: Pierre, you wanted to cover TTML1 3rd Ed Tests, but before that...
Thierry: TTML1 3rd Ed CR2 is published, and the call for exclusion has been sent.
Nigel: Thank you!
Thierry: We are on track and schedule.
Nigel: Okay, Pierre, tests?
github: https://github.com/w3c/ttml1/pull/361/files
Pierre: I just pushed it a couple
of hours ago so I don't expect anyone to have had a
... thorough look. I've gone through all the substantive tests
since 2ed and created tests
... for everything that can be tested and highlighted what I
think cannot be tested.
... The tests are inspired by the IMSC tests so they will seem
familiar.
... I'd like a review especially on the things marked as
non-testable.
... From a practical perspective, if Glenn could try to render
or check them using TTPE and
... TTX that would be great. Last time we spoke the plan was
for TTPE and IMSC.js to be the
... two implementations for the TTML1 3ed tests.
Cyril: How many of these tests are relevant for IMSC 1.1?
Pierre: First, there may be zero
IMSC 1.1 tests because everything in IMSC 1.1 is already
... in TTML2 or IMSC 1.0.1, so there will be no additional
tests needed to meet the IMSC 1.1
... exit criteria.
Cyril: I understand that, but how
many of the tests for TTML1 3ed are relevant for IMSC 1.1
... features?
Pierre: All but one are covered
by IMSC 1.1. But I think the purpose of those tests if I
recall
... correctly was specifically to convince the Director that
the substantive changes were in
... fact implemented.
Nigel: That's right.
Pierre: I created those tests
specifically to demonstrate that. All but one are already
... exercised by IMSC1 tests and TTML1 tests in fact.
Cyril: Which is the exception?
Pierre: Anamorphic fonts. There's one test that is triggered by anamorphic fonts.
Cyril: 2 value relative font size?
Pierre: Exactly. That one is not
part of IMSC1 or 1.1 and I'm not even sure it was part of
... TTML1 test suite either.
Cyril: Ok, thank you.
Glenn: Q: what did we change in the spec that that particular test is used to demonstrate?
Pierre: If you recall, we added a
bunch of text that discussed inheritance.
... Example, p fontSize = "1c", then child span fontSize="1em".
The font size is calculated
... to be 1c, that's boring.
... There are two examples in the TTML1 text that describe
this.
Nigel: The examples are in Notes?
Pierre: Yes
... The 3rd and 4th note in that section.
... "1em 1em" can result in an anamorphic font size.
... The relative font size is relative to the computed parent
font size which can be anamorphic.
Glenn: What normative text changed that drove adding that test?
Pierre: The entire text of that section was heavily changed.
Glenn: I'm just wondering if we
went too far in creating that test or if the original
test
... suite was under-represented.
Pierre: The 5th paragraph, "When
a single relative <length> value is specified, ..."
... If that is in TTML1 2nd Ed then I agree there's no need for
tests, but I'm thinking it was not.
Glenn: I'll check.
Pierre: The two that can't be
tested or don't need to be tested are:
... 1. 'should' regarding the tts:lineHeight.
... 2. application defaults for frame rate and sub-frame
rate.
... I don't think that's testable.
Nigel: Presumably we could
include the application settings to apply for testing frame
rate
... and sub-frame rate, for example in text outside the TTML
document instance, then the
... same instance would have a different evaluation in some
way?
Pierre: Sure, but that in itself,
the decision to apply an application default, has no
requirement
... so I don't think it's testable.
Nigel: In other words we have
merely made explicit the already existing option for an
... implementation to do its own thing?
Pierre: Yes, that's right.
Nigel: That seems like a
reasonable argument to me. What about lineHeight.
... Why not be able to test that?
Pierre: It's a should.
Nigel: But the semantic is still testable.
Pierre: Yes but applications can be conformant without doing it.
Nigel: Yes but we can still test the semantic.
Pierre: The syntax is unchanged, and it's a should.
Nigel: Yes but the test needs to
demonstrate implementability, so there needs I think to
be
... some test that shows the should behaviour can be
implemented.
Glenn: If normal already appears
in any of the TTML1 tests then we don't need a new test
... for this.
Nigel: Why not?
Glenn: If normal is already there
then that test for how normal is used, with no normative
... or exemplar images...
Pierre: I don't understand - say
tts:lineHeight="normal" and an implementation returns to
... you something with a line height that is double the font
size, what do you conclude?
Nigel: Depending on the algorithm
in the spec and the font resources, on balance most
... likely that implementation is not demonstrating that the
spec can be implemented.
Pierre: It's only a recommendation.
Nigel: If the spec said "should
go back in time by 10 minutes" then the Director would
want
... a test to show that, but of course no time machine exists,
so I think that text would
... lead to trouble.
Pierre: Good news, there already is a test for lineHeight="normal".
Glenn: As I said.
Pierre: I wanted to agree the scope first.
Glenn: I wonder if it would be
consistent if we add an exemplar to a small subset of
TTML1
... tests.
Nigel: I don't think we need to worry about that.
Pierre: My proposal was just to
check in the test files without exemplars, but file the
outputs
... generated from our implementations under the implementation
report.
Glenn: Ok. When is the last date for this?
Nigel: Thanks to Thierry,
Nigel: Shows that implementation report needs to be finalised by 24th September.
Pierre: Can we give ourselves
until August 30th to review those tests and make changes,
... then freeze them then create the implementation report, and
then start submitting results.
... Just a suggestion.
Glenn: TTPE implements normal as
described as well as anamorphic fonts so I think we're
... good there but I'd need to run the tests to verify that. I
don't think there's any problem
... in using TTPE as one of the implementations to verify
those.
Nigel: I like Pierre's suggested plan, any issues with that?
group: [silence]
Nigel: Okay, then agenda+ for
30th August to confirm the TTML1 3ed test suite so we can
... begin to create the implementation report.
... [discusses tests, with comments on the pull request]
Glenn: I just double checked the
anamorphic font text, and it turns out that it is
basically
... present in 2nd Ed in the 4th note in ยง8.2.9, in the last
sentence. So this is basically a
... paragraph of text in a note already, now made normative
whereas it was more an
... explanation of an implication in 2nd Ed.
Pierre: So do we need the tests still?
Glenn: In my opinion the semantic
was already there and we're not demonstrating a new
... semantic.
Pierre: I'm not excited by it.
Glenn: I'd be willing to have no test and point to the 2nd Ed text for that.
github: https://github.com/w3c/ttml1/pull/361
Nigel: I think the Director was
asking for tests to demonstrate the substantive changes,
... and this counts as one because normative text clarifies
what may have been ambiguous
... before.
... We need to look more carefully at this to see if an
existing TTML1 test for two value fontSize
... can be reused or already demonstrates this.
Glenn: I added a comment under this pull request.
Nigel: This needs further
investigation - I see that the diff tool isn't helping
us.
... in the direction test, shouldn't the direction on the first
p be direction="ltr"?
Pierre: That is a subtle point
that 3ed clarifies - without bidiOverride the "natural"
direction
... of the script is not overridden.
Glenn: Where that comes into
effect is resolving the directionality of weakly
directional
... or neutral directional characters at the boundaries of the
paragraph, like the period that
... ends the paragraph is neutral directionality. If the
paragraph embedding level is ltr then
... a period at the end of a hebrew or arabic sentence takes on
the direction of the previous
... character, but there are scenarios where it doesn't work.
On this point, previously in
... 2ed and prior we didn't call out this semantic but probably
many implementations
... implemented it as we have now clarified. Since we did not
say either way, some
... implementations may have read between the lines and applied
it to p which would not
... have been non-conformant since we had no test suite
examplars to follow. It could have
... fallen through the cracks.
<tmichel> FYI you may use the following diff tool
Glenn: The XSL spec says to apply it this.
Cyril: 2 comments about the test
suite.
... Some tests have a copyright.
Pierre: Yes, I thought I'd removed them and didn't. Please add a comment.
Cyril: Secondly, in the pull
request you've done the work to link the issues to the
tests,
... but there's no link backwards. I wonder if we should put
metadata in the tests to point
... to the issues or the spec sections that it is trying to
test.
Pierre: Yes that'd be awesome.
Nigel: Is that information available?
Pierre: Yes it's in the pull
request. Most of the tests have a metadata section so it's a
matter
... of copying and pasting the pull request info into the
metadata header. If someone wants
... to spend 45 minutes doing it that'd be awesome.
Cyril: Going a bit further,
referencing the spec itself?
... I suppose the issue goes to the pull request goes to the
section of the text.
Pierre: Exactly. Copying the bullet point is all we need to do.
Nigel: Any other points on this test suite?
SUMMARY: Test suite to be finalised August 30th, review to continue until then.
Glenn: I would note that August
30th is prior to when it is actually needed so we have
... some room to slip that if necessary.
Nigel: Yes but there has to be time for the implementations to respond to the tests.
Glenn: Yes
Pierre: The goal is to freeze the tests so implementers can work on them.
Glenn: It's a good goal, just not an absolute hard deadline.
Nigel: It's a target, and as
Cyril mentioned there is other work to do at the same
time.
... If we can freeze the tests on 30th then that gives 3 weeks
for implementations, which
... seems reasonable.
Pierre: If you can run the validator that would be good, Glenn.
Glenn: Sure I can do that, and check what TTPE does as well.
action-443?
<trackbot> action-443 -- Glenn Adams to Prepare a document showing mapping arib ruby extension features to ttml2 for use as a liaison document to arib. -- due 2018-08-09 -- OPEN
<trackbot> https://www.w3.org/AudioVideo/TT/tracker/actions/443
Glenn: Put that off please.
Nigel: Okay, you asked for this
and asked me to put it off until today.
... We made the transition request as per the plan, and if all
goes well we should have
... TTML2 CR3 published by 14th August, with a deadline for
comments of 11th September.
... We need to leave reasonable time for any response.
Thierry: I don't expect a
Director meeting because this is straightforward. I will ping
them
... (Philippe and Ralph) but it should be approved
tomorrow.
Glenn: The documentation is already in the right place, so this is just a formality?
Thierry: Yes, I'm waiting for the normal process to go through.
Nigel: Raising this as it's the next thing to do, and I see that Glenn has been submitting tests.
Glenn: I've now blessed 249
unique tests that are all in the validation category.
... They are either validity or invalidity tests, documented in
the spreadsheet shared
... previously.
<glenn> https://docs.google.com/spreadsheets/d/1nunAY0F0bxL0nZDng2VxmWHv7i2RyAIvKQwDiiVzooc/edit#gid=0
Glenn: If you don't have share
access to that then request it.
... I'm almost finished, I expect to have 100% coverage of the
features for validation tests
... by Monday, hopefully, if not earlier. That's going very
well.
... Then I will begin populating presentation tests, starting
from about 100+ tests that are
... in TTPE already and use those to populate presentation
tests in this spreadsheet, and
... also I'll upload to the ttml2-tests repository.
... As I'm going through these I'm also updating either TTV or
TTPE implementations to
... verify that they work and produce desirable results. There
will be some points where
... I have presentation tests with no TTPE rendering code so I
will be relying on 3rd parties
... to demonstrate those. For example disparity and
luminanceGain.
... They are included in IMSC 1.1 that will be depending on
others to satisfy.
... I would like to hear if anyone has implemented or plans to
implement those.
Pierre: I just submitted a couple
of hours ago two proposed tests, one for disparity and
... one for luminanceGain, on the imsc-tests repo.
<pal> https://github.com/w3c/imsc-tests/pull/69
Pierre: That's the pull request.
Glenn do you want to look at those, and tell me if those
... would fit in TTML2 presentation test suite, and if so we
can put them in there?
Glenn: My thinking is that I will
also draw from other tests as well as the TTPE ones. I
took
... a quick look at those this morning and they look fine on
first glance. I will check they
... validate, but the presentation part modulo Nigel's comments
seem okay. Do we have
... anyone signed up for implementing them?
Pierre: I will be looking into
this, but before that I wanted to look at concrete tests
and
... check if those tests are going to be sufficient before I
try to figure out test results.
Cyril: A clarification question
for Glenn. You said you will have validation tests
finalised
... maybe by end of this week, but it is unclear to me if you
foresee gaps in what you will
... be submitting in tests. Do you need any help to provide the
tests?
Glenn: I will provide 100%
coverage of validation tests.
... The presentation tests though I mentioned my procedure. I
will draw from TTPE then
... other sources including what Pierre has done in IMSC 1.1
and will probably be soliciting
... others to add to that. The only gaps I see probably are
going to be audio related.
... We don't have any plans to do audio presentation work right
now.
Nigel: I will do that, I created
the adhere column in the implementation report.
... I have begun to produce audio style attribute presentation
tests and plan to contribute
... those after my vacation.
Glenn: On that point I noticed
that you did not check the box for #audio-description,
... however I think the features already cover all the features
so you may already satisfy
... that one as well.
Nigel: The difficulty with that
is that the Web Speech API audio output is not available
as
... an input to the Web Audio API, so if an implementation is
required to do text to speech and apply pan and gain that is
difficult to do.
Glenn: Okay we'll have to come back to that.
Nigel: But that raises a really
important point.
... We are currently dependent on the Web Audio API Spec, and I
had been assured by one
... of the Chairs of the Web Audio WG that it would be in CR
actually before TTML2, but
... that hasn't happened, even though the spec is stable and
implemented in more than
... one browser. I believe the hold-up is with the team support
for the Web Audio WG, whose
... staff member has not been progressing this, I don't know
why. Thierry, please could you
... ask Chris Lilley if he can progress it?
Thierry: Yes I can.
Nigel: Thank you.
Glenn: Just doing a quick glance
at the feature definition, audio-description is a subset
... of the audio-speech features, so if you have an
implementation that can support that
... then it already supports audio-description.
Nigel: OK
Glenn: [details of the features]
Nigel: Thanks, I'll go and look
again and double-check that.
... Is there any requirement for continuous animation in
there?
Glenn: Not in those feature definitions.
Nigel: Okay, we're mid-implementation of continuous animation of pan and gain, for information.
Glenn: For the sake of mentioning
it, we don't have any standards in this WG or in W3C for
... how many tests have to be in the test suite for some
feature.
Nigel: Okay, that's noted.
Glenn: I just wanted to remind us of that fact.
Cyril: One question about the
time wallclock feature. I'm wondering who is going to
support
... it, if anyone would be supporting it.
Nigel: Remind us?
Glenn: We added a new clock time
function to the time expression syntax. TTV will
implement
... validation of that, but I'm not sure who will be
implementing presentation.
Nigel: Oh yes, the ability to add
date to clock times.
... What would that do?
Glenn: The time would be synchronised to the wallclock date and time in a presentation.
Nigel: I foresee that being used more for recording live events historically than for triggering events to occur.
Glenn: It does have presentation semantics.
Nigel: Right, so the CR exit criteria require a presentation processor for it.
Glenn: We don't say if a
presentation processor is preceded by multiple
transformation
... processes. I don't know if that would be applicable here.
If it is thought that this would
... mostly be used by an authoring process then you might be
able to claim that processing
... up the chain has presentation semantics eventually.
Nigel: Thank you for the
reminder, I'll have to look to see how we can support
implementation of this.
... Are there any other features like this?
Glenn: I think Cyril has mentioned all of them now. Thanks for adding that Cyril.
Pierre: Glenn and others, when do
you think we can have a generic yea or nay on the
... disparity and luminanceGain tests?
Glenn: By tomorrow.
Pierre: Awesome, thank you that would really help.
Nigel: Our timeline says
implementation report finalised 27th September, and we need
to
... give time for people to provide implementations against
tests, so that means we need
... the tests to be finalised sufficiently far in advance. We
said just over 3 weeks for TTML1.
... There are a lot more tests for TTML2.
... Three weeks before 27th Sep is 6th Sep. Is there any hope
of finalising the test suite
... by then?
Glenn: We can have tests by then, whether they are finalised or not is a different question.
Nigel: As a goal, I mean, not necessarily locked down.
Glenn: 6th Sep leaves about 3-4 weeks from today. It's possible, it's aggressive.
Nigel: We have to be aggressive to meet our timeline.
Glenn: Absolutely!
Nigel: Implementers may come back
with issues on the tests, but we need a line in the sand
... to go to potential implementers and say here's the test
suite.
Glenn: It depends who will
provide implementations. Skynav and Netflix will provide a
lot
... of coverage, and BBC and maybe IRT will provide some
implementations too.
Andreas: IRT will implement IMSC 1.1 validation.
Glenn: Thank you, Andreas. There
are 25 features out of 135 new features that are in
IMSC1.1.
... That leaves 110 features not included in IMSC 1.1.
... If the audience for our testing effort is primarily those
of us here I'm presuming that
... we're keeping track of what's being proposed for tests and
being able to be responsive.
Nigel: You may be right, but we have to offer the test suite to others also.
Glenn: Absolutely. That's why Sep
6 is okay with me for a "beta" version, which gives
... people something to work with, right?
Nigel: Yes.
Cyril: I counted 141 new features not 135. Is there an issue here?
Glenn: Let me double check that.
There may be some where all of the tests are a subset
... of something else that is already tested. I think the 6 are
primarily new feature labels
... for things already in TTML1, like #set, which is an alias
for #animation in TTML1. I think
... the remaining 6 come under that category.
Cyril: Thank you.
... Another question on the implementation report. Do we expect
implementers to provide
... renderings?
Glenn: It's not required.
Nigel: It's not required, it's nice to have.
Glenn: It's not part of the
implementation report review process. The Director can't
require
... there to be images, because it is not a requirement.
... We operate on an honour system for implementation reports,
since no public demonstration
... is needed.
Nigel: That's right.
... Of course it is nice when people do provide one. Also
remember that implementations
... can be anonymous, as long as the team can have a reasonable
belief that it does exist
... and does what it says.
Glenn: Testing is different from
HTML where the tests are targeted at browsers. We have
... to allow for implementation reports from walled garden
contexts where the report may
... be anonymised, right.
Nigel: Yes.
Glenn: There were a couple of
TTML2 issues, #945 and #950 that were substantive, where
... I have implemented the deferral procedure. I just wanted to
note that.
... For #950 the group agreed it as a resolution.
... For #945 the commenter agreed to defer it in an email to
the list, which I used as the
... basis for the deferral.
... I wanted to note it for the minutes.
Nigel: Thanks for reminding me of that.
Glenn: At the moment there are 0
pull requests on TTML2 and 12 open issues, all of them
... marked for PR.
Nigel: And editorial.
Glenn: Yes, and editorial. I will
be implementing pull requests for those over the next
... few weeks as we go towards PR.
... You also put #973 on the agenda Nigel.
Nigel: Thank you for reminding me.
github: https://github.com/w3c/ttml2/issues/973
Nigel: This might affect tests.
It is not obvious. It also could potentially impact IMSC,
where
... some constraints may be needed on where rw and rh units can
be used, so they are not
... on every length expression.
... If for example a document instance requires processor
support for rw and rh but that
... processor doesn't actually support it in every context,
does that make the processor
... unsuitable or non-conformant?
Glenn: We don't qualify the
feature definitions, we don't say "anywhere" or
"everywhere"
... in the text. I don't think we qualify its context anywhere,
and we don't say that if there
... is no constraint on context then it applies everywhere.
Nigel: That might be counter-productive!
Glenn: I'm okay with not adding
such a note. It was the intention of how we authored them
... that they would apply everywhere.
... We do have language about unsupported values for most
properties to say what the
... correct behaviour should be.
... How does the language of nearest supported value apply to
the definitions of features?
... That's not a question we've entertained before I think.
Pierre: I think we do what we
have done in other places, say that this particular
feature
... is permitted in IMSC 1.1 but add constraints like "it shall
not be used in these circumstances".
Glenn: We have a few of those in
TTML2 like extent on the tt element, which must be in px
units.
... There are a couple of places where the units must be the
same if there are multiple values.
... That does not abregate the semantics of the original
feature.
Pierre: My current plan, we added
#length-root-container-relative to IMSC 1.1 at the last
... minute and marked it as at risk, I will be submitting as an
issue a set of proposed
... constraints and if folks have thoughts about what is really
important vs less important
... then please forward me your data and I will try to propose
a strawman.
Nigel: Okay, thank you.
... Okay, so coming back to this issue, a claim of support for
#length-root-container-relative
... implies support on all length expressions in the absence of
any other caveats.
Glenn: That's correct.
Pierre: That's the only logical interpretation.
Nigel: That's fine, do we need some editorial text that explains that, or is it unnecessary?
Glenn: I don't think we need
anything absolutely in TTML2. I would not object to adding
a
... note somewhere or eventually making it a normative general
rule.
... I don't think we have to do anything in TTML2 1st edition
if we don't want to.
... I would oppose saying anything specific to one feature, I
wouldn't mind a general
... note.
Nigel: Yes, that makes
sense.
... Any other views?
group: [none]
SUMMARY: WG generally accepting of a general note explaining non-normatively that features like this apply universally or not at all, in the absence of any other caveat.
Nigel: Is there anything to discuss here?
Pierre: I don't think so. I may
have found a bug in the IMSC test suite - please review
it
... if you are interested.
... Other than that, modulo luminanceGain and disparity, and
root container length, the only other at risk
... feature is lineShear.
... Other than that, nothing to report. I'm nearly done with
the IMSC 1.1 test suite, not
... intended for exit criteria, just to be useful for the
industry.
Glenn: TTPE will include TTML2
support for lineShear presentation semantics, so there's
... a potential that we could be able to claim support for IMSC
1.1 lineShear.
Pierre: Thanks for that. I expect
we would formally be able to include lineShear in IMSC
1.1.
... We'll meet the exit criteria. My question is more if we
want to do that. I'm still trying to
... see what is the right thing. There's a thread on the CSS WG
issue. Hopefully we can gather
... the right data to make the decision there.
Pierre: I have some editorial work to do there.
Nigel: Thank you.
... The CfC for publishing a WG Note ends today, but until the
issues are resolved that
... decision cannot be enacted.
... That's pretty straightforward.
Nigel: Note that we have a new relevant CSS issue as Pierre mentioned a moment ago.
#2983 Support for shearing of lines and inline elements
Nigel: Also, yesterday, the CSS
WG agreed a name for the equivalent property to
fillLineGap,
... which from memory I think is inline-box-sizing or something
like that, with property values
... "normal" and "stretch".
Nigel: The question has been
raised with EBU, and I hope there will be some feedback
... on new profiles to add to the registry. We haven't had time
to discuss the approach
... regarding processor profiles and content profiles yet.
Something to look forward to,
... maybe for the agenda for TPAC?
Nigel: Thanks everyone. Reminder
next meeting 30th August. Enjoy the break (from
meetings!).
... [adjourns meeting]