See also: IRC log
<scribe> scribe: nigel
Nigel: Today we have any TPAC
scheduling/agenda points, IMSC planning and requirements,
... and any other things Thierry?
Thierry: David mentioned that he
is unable to join today, so I don't know if we will cover
... WebVTT now that the review has ended.
... We can discuss the TTML2 HR process too.
Nigel: Okay, thanks. Any other points to raise today?
group: [no more agenda items]
Nigel: Pierre, you contacted me
in the week and requested a schedule change, which you've
now
... added to the wiki page - thank you.
Nigel: The IMSC vNext work
session is now on the Thursday morning.
... Any other schedule requests or agenda topic changes?
... Seems like no. Please let me know if you want to dial in,
or if you're attending in person
... add your name to the participants list on the wiki
page.
... Pierre, I'm conscious that I owe you an arrangement to
discuss the demo stuff!
action-506?
<trackbot> action-506 -- Thierry Michel to Draft a wiki page explaining our review and disposition steps and labels -- due 2017-09-21 -- OPEN
<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/506
Thierry: I have updated the
previous email and labels, as per the email sent
yesterday.
... The updated WR process page is:
Thierry: Some new comments have
been sent, by Pierre, so my next stage is to add the
... labels to the repo and apply them to the relevant issues.
I'll do that next.
action-506: [Meeting 2017-09-28] Thierry plans to add the GitHub labels shortly.
<trackbot> Notes added to action-506 Draft a wiki page explaining our review and disposition steps and labels.
Thierry: The last thing is that
there are comments from commenters who have reviewed
... TTML2 and said they were fine with it - APA did that and I
think someone from an external
... body said something. They're not raised as issues so
they're not in GitHub. Should I put
... them in with a link to the mailing list and say that
they're resolved and agreed by the
... commenter, so we can track them in the same place to show
to the Director.
Nigel: I think that makes sense -
a "no action" issue.
... Right, so last week's feedback on the process and labels
has been processed, the
... next step is to make it happen on GitHub.
... I see that there are 2 days of review period left for
TTML2.
... We did receive some feedback from organisations saying that
they will respond later. I
... guess we will process feedback whenever it arrives.
... The TAG has been meeting and asked for an update on their
feedback to TTML2, so
... I added a comment to their tracking issue:
Pierre: A bit of background:
IMSC1 is one of the timed text formats supported by CMAF,
... which is an MPEG specification 23000-19, which is being
adopted by Microsoft, Apple, Netflix and others.
... They are planning a new version of CMAF and they would
ideally like to include a version
... of IMSC. The main point they want to cover is the lack of
ruby support. So there's interest
... in moving as quickly as possible with IMSCv2, perhaps
meeting the timeline of MPEG, which
... meets in October. Something we should plan as an action is
to communicate our timeline
... to MPEG if we come up with one, ahead of their October
meeting.
Nigel: It'd be helpful if they told us formally - it's a bit strange as an order of events to communicate with them in advance!
Pierre: Yes, one way we could do
this is to tell them and other organisations about the
work
... that we're doing, and they could respond with their
thoughts.
Nigel: Seems like a good idea to me.
Pierre: So with that interest in
IMSCvNext, I've started the process by pushing to a new
... repo to document the requirements.
<pal> https://w3c.github.io/imsc-vnext-reqs/
IMSC v Next requirements document GitHub repository
Pierre: This is to get us
started. I have some questions, so I've filed two issues. It is
based
... on what people have told me informally they are interested
in. It is a first pass so input is
... really welcome on those requirements, especially if we can
link to detail use cases, requests
... from other organisations etc. I don't expect the scope and
overall goal for compatibility
... with IMSC 1.0.1 to change.
... The reason I've gone ahead and published this is I'm also
making a first stab at the spec
... itself based on those requirements, which I hope to have
ready later today. This is to make
... it less abstract. In the process I think it will also
highlight potential issues with TTML2
... and other things we ought to fix sooner rather than
later.
Cyril: One thing about CMAF -
CMAF is always creating amendments, so there will be more
... later for sure. I'm not aware of anyone submitting the idea
of having a new version with
... IMSC2 yet. I don't think there's an urgency for the October
meeting as opposed to the
... January or February meeting.
Nigel: I see there's nothing
about the HRM - does that need to be updated to take into
... account the new features?
Pierre: That's on my to-do list - it has to be addressed just as a matter of consistency in the spec.
Nigel: I think from the
requirements perspective we may just want a statement that
a
... hypothetical render model is required.
Pierre: That's already included
in the scope given IMSC 1.0.1 compatibility. I expect
there
... to be an HRM in IMSC v.next.
Nigel: The other thing is we
probably need to explain what we actually need rather
than
... just referencing TTML2 feature designators.
Pierre: We have to strike a
middle ground here.
... Ideally we'd be able to point to external documentation
explaining the requirement source.
... For instance I notice in TTML2 there are multiple levels of
ruby support, including nested
... rubies. One of my obvious questions is do we really need
nested rubies - hopefully we
... will be able to get answers to those questions.
... Japanese language support is overwhelmingly a desired
feature for IMSC v.next.
Nigel: Okay, what are the
actions?
... 1. Propose some liaisons.
... 2. Review the requirements documents.
Pierre: After I've published the
first draft spec today I'll be happy to propose a roadmap
and
... start working on liaisons to other groups.
Nigel: Thank you.
... Just to be clear, I believe the goal for the IMSC v Next
requirements is that we publish
... it as a WG Note.
Pierre: Yes, that's what it says in the SOTD.
<scribe> ACTION: nigel Add IMSC vNext repo to agenda, board, github-bot etc [recorded in http://www.w3.org/2017/09/28-tt-minutes.html#action01]
<trackbot> Created ACTION-507 - Add imsc vnext repo to agenda, board, github-bot etc [on Nigel Megitt - due 2017-10-05].
Nigel: Thierry, I might ask you for some help with that.
Thierry: OK
Pierre: And I've decided the
least evil option is to put IMSC v.next into a sub-folder of
the
... existing IMSC repo, because it makes it easy to share
issues and move them across versions.
... I expect the versions will share common tools, files
etc.
Nigel: That makes sense to
me.
... Would you use labels to identify target specs?
Pierre: Yes. All the old issues
deferred from IMSC v1 have a v2 milestone against them.
... Might be okay to use labels instead or as well.
Nigel: Yes, we just need a clear model - milestones are for events in time, labels for what they are.
Pierre: I'll take care of that,
I'll add a label to the v2 milestone issues and then we can
move
... them to new IMSCvnext milestones as needed.
... I'll do it right now.
Nigel: Thank you.
github: https://github.com/w3c/ttml2/issues/429
Andreas: Thanks for raising this,
I support it, and wanted to check it myself.
... As itts:fillLineGap is a complicated attribute, I see
problems implementing it, and in
... compatibility, so it makes sense to keep the namespace.
Pierre: We might do something
subtler like reference IMSC 1.0.1 from TTML2, and then
... remove it from IMSCv.next. What's really bad is today the
semantics don't
... match.
... It's hard to understand if the TTML2 semantics are the same
as the IMSC 1.0.1 semantics
... because the wording is so different.
... Either the wording is insufficient in IMSC 1.0.1 and should
be fixed there or it should be
... made the same in TTML2. Implementers need to know that the
behaviour is the same.
Nigel: Okay, alignment is the
issue here - we should define something with the same
... semantics in the same way.
Pierre: I'd extend that to the namespace too.
Andreas: I mentioned it before -
there has been a bad experience changing namespaces
... going from the ttaf namespace to the ttml namespace, which
broke things unnecessarily.
Nigel: I can see the issue about
namespaces - ideally the profile should be a pure profile
... of the spec, and not have this zigzagging up between the
profile and the base spec.
Pierre: Yes, in the future, maybe
we just have the definition in TTML2 and remove it from
... IMSC, just referencing it by feature designator.
SUMMARY: Majority view in favour of aligning semantics and syntax between IMSC 1.0.1 and TTML2, partly constrained by existing implementation work and adoption.
github: https://github.com/w3c/ttml2/issues/428
Pierre: Identical comment to that on ttml2#429
Nigel: I thought there was
already an agreement to adopt the offset and extent
syntax
... in TTML2 - if there's no issue we should bring that in
here.
SUMMARY: Majority view in favour of aligning semantics and syntax between IMSC 1.0.1 and TTML2, partly constrained by existing implementation work and adoption.
Thierry: I have invited David
Singer to Chair this part of the meeting but unfortunately
he
... had a conflict and could not attend today. Thank you Silvia
for joining. We've been through
... 2 reviews for WebVTT, one some years ago and a recent one
that ended a few days ago.
... From the first review, most of the comments were handled
and the spec was updated
... but what was missing from the review itself to show a
Disposition of Comments document
... to the Director, an agreement for the commenter. I'm not
sure we have that for each
... comment. The second review that ended a week ago, we have a
fair number of comments
... and some discussion has started in the CG. Silvia, do you
have an idea how you plan to
... handle those comments, if that was discussed already?
Silvia: The biggest problem we
have right now is that the Editor has just stepped down,
... who would be the key person to address those comments. I've
offered to take on a little
... bit of effort there however I don't have enough time to do
the spec justice because my
... full time job is not related to WebVTT anymore. So I'm
going to meet David this afternoon
... to see how we can address that. Next week at FOMS we may
find someone interested who
... can step up as an Editor and help us address the issues.
Most of the issues have come
... from Nigel from the BBC and Andreas from IRT, both of whom
are present here.
... Would you be able to give a quick summary, and the
priorities? That might really help
... us identify who can best help in resolving those
issues.
Andreas: Thanks for your response
and answering to the previous comments, around 2 years
ago.
... The new issues I filed have been part of the syntax
analysis of WebVTT as written in the
... email and of course some of them are more generic and not
always concerned with
... changes since the last review. But I also see that most of
the issues are not so critical,
... it is more to give a hint that there may be a problem, and
to get an answer to fix or not
... to fix. That would be okay in most of the cases. I think
also from the discussions on
... GitHub issues, the missing support for text color if there
is a non-CSS conformance class
... is not acceptable for accessibility so that's definitely a
big concern that really needs to
... be fixed.
Silvia: Okay, there's a couple of
things that both of you mentioned that have been
discussed
... for a long time. We'd much rather have browser vendors
implement support for the basic
... functionality we have right now than implement new features
rather than delay the spec.
Andreas: I agree, and partly don't agree, especially for text colour.
Silvia: That's why I want to know which things really need to be fixed.
Andreas: For the discussion, I
made it clear that this is a really important feature for
... accessibility in most of the countries in Europe for
example, so this really needs to be done.
Nigel: +1
Andreas: An issue I had also was
the decision to be based on browser support. If the
... browsers decide not to support it then the browsers can be
closed for the time being.
... If there are other stakeholders, that should not be the
deciding factor.
Silvia: I was with you on that.
The two previous Editors had more of a focus on browsers
... than anyone else involved in the process. Maybe we can find
the next Editor, and I'm
... more open to fixing those things right now. Certainly all
of those comments that have
... been filed will help fix those things.
... Nigel - I know you have a lot of things so thank you for
the effort.
Nigel: Ok, no problem. [fails to
scribe while also talking] I think from a UK audience
perspective I second
... the need to support colours, but I would also highlight the
positioning features and
... the comprehensibility of the spec. I know you pushed back
on my comment about
... algorithmic specification, but maybe there's a middle
ground of refactoring the big
... monolithic algorithm into smaller units each of which makes
sense by itself.
... In particular the interplay between the two positioning
mechanisms is very hard to understand.
Andreas: I made similar comments
before and I agree about positioning and how easy to
... understand. I know Philip and Simon mentioned that the spec
doesn't have to be an
... explainer or user guide, but maybe that can be done
separately.
Silvia: A couple of things to be
aware of with the WebVTT spec. It strongly follows the
way
... the HTML spec was written. The algorithmic part that is
complicated to read is really not
... the part that authors need to read. They need to understand
the syntax and a user guide,
... which is something that other people can write to explain
in simple words and maybe
... for restricted use cases how to do best practice for users.
What the spec is trying to do,
... which is why the algorithmic approach is so important, is
to give implementers a unified
... approach to implementing it so everyone implements an
interoperable version. That's
... the goal, so its what we are trying to optimise in the
spec.
... Maybe with the positioning what we can do is make it much
clearer in the earlier sections
... in the spec what the cue model and the region model is. I
thought we had improved that
... but I admit I haven't had the time to read in detail all of
your comments Nigel. As an
... outcome of all the feedback I think we can improve the
readability. It's great to hear that
... positioning and colour are the things to focus on. I take
on board the idea of breaking
... up the algorithms some more. I'm going to look for someone
who can take that on as
... the Editor. That's really useful, thank you.
... I really wanted this opportunity have this conversation
with both of you. They are the
... key feedback we got so I wanted to address that. What I
need to know now from Thierry
... is what are the next steps so that we can progress this to
a proper standard, which is
... what the community has been asking for for a while.
Thierry: The way we're working
with a WG and a CG is pretty specific, so we have to
invent
... our own ways of working together. What we have done is for
the first review for the CG
... to discuss the comments and once resolved approve them in
the WG. I think we should
... follow the same procedure here. I think we will still use
GitHub and we have a wiki page
... explaining the status of each comment. For example also for
TTML2 we have designed
... a labelling system in order to mark the comments. We could
use something similar for
... WebVTT if you'd like.
Silvia: okay
Thiery: When the comments are
finally resolved and responded to then we have to produce
... a final disposition of comments including both the
reviews.
... Then it's for the WG to agree to transition to CR with some
discussion of the CR exit
... criteria, and then we will request transition and maybe
discuss with the Director. Then
... during CR we need a test suite and demonstrate that we
fulfil the exit criteria.
Silvia: We have a test suite, so
I think that can go ahead. I assume the WG will want to
... review the comments. Is there anyone else planning to
provide comments?
Thierry: The review is closed
since September 22. If we do have a comment it will be
... difficult to refuse it but I don't think we will get that
many comments in.
Silvia: So it's a matter of addressing the comments we have and agreeing the resolutions.
Nigel: I don't think I've seen any horizontal review comments at all - is that correct?
Thierry: Yes, we requested comments but didn't get any responses.
Silvia: For the first review we
got heaps of feedback e.g. on i18n and resolved all their
... issues so I think that may be the reason why there's no new
feedback.
Thierry: I'd agree with that, and
also, we did trigger a Horizontal Review to the
appropriate
... groups and I also reminded the groups about a week before
the closure of the review
... period so I think we're safe.
Silvia: We did have feedback on
one accessibility feature that hasn't made its way into
the
... GitHub repo yet. That was for the alt text on the image. We
need to make a change to that.
Thierry: That's really
editorial.
... Yes, they said they'd reviewed the spec and didn't find
anything else wrong.
Silvia: Cool, so David and I will
meet and figure out the next steps on the Editor and how
to
... address all the comments and get back to you, and I'm also
going to try to get the CG
... involved. So I guess that's about as much as we can do
right now.
Thierry: We should allow some
time to discuss WebVTT during TPAC and I hope David
... will be there and be able to Chair and have other people
from the CG.
Silvia: I can't make it though I
could maybe join remotely if necessary.
... There's 8 hours between California and Australia. If you
can make it at a time I can join
... I'd be happy to join.
Nigel: I've made the request for
a speaker phone on the AV equipment survey. If you can
... add your name to the TTWG TPAC 2017 wiki page and any
constraints on timing that
... would be very helpful.
Silvia: Okay, I will.
Cyril: Do we have any idea when
we want the comments to be addressed, in terms of
timeline?
... Maybe it's too early to say.
Silvia: When is TPAC?
Cyril: 6th November.
<inserted> TPAC is 6-10 Nov, TTWG F2F is 9-10 Nov
Silvia: I'm sure David will want
to move fast. Let's see what I can get done with the WG
and
... the conference next week. I'm going to target 6th November
with progress even if not
... all of the comments. It's a good point, thanks Cyril -
without timelines it will never happen.
Nigel: Thanks everyone, we've completed all our agenda items for today. [adjourns meeting]