See also: IRC log
<scribe> scribe: nigel
Nigel: Today we have an agenda
item on WebVTT that Dave will take at the top of the
meeting,
... then IMSC and TTML. We may be about to request publication
of three specifications
... all at once, at different levels of transition.
... We should also cover TPAC briefly if time allows.
... There's been a bit of progress on HDR in PNG too. Any other
business?
group: [silence]
Dave: We haven't talked about the
WebVTT spec for a long time on a call. I wanted to
... bring you up to speed. Firstly an apology we haven't been
publishing WDs, as I didn't
<dsinger_> https://w3c.github.io/webvtt/
Dave: realise we don't need
permission to publish. Unless anyone wants to object then
... we would like to publish the above as a new WD.
Nigel: For wide review?
Dave: Let's come to that. Initially just a heartbeat publication.
<dsinger_> https://www.w3.org/wiki/WebVTT_Wide_Review
Dave: 2 years ago we requested
wide review, and received a load of comments. We finally
... reached resolution of all the issues raised. I asked the CG
to check once more on the
... disposition of all these bugs. Thierry has helped to build
the table of all these issues,
... including links to the issues and the dispositions.
... The WG status says @@ on every issue. At the bottom we see
the key for WG status
... to be filled in. I would like to see status 3.1 CG
disposition agreed by WG.
... I would like the WG to indicate in the next couple of weeks
if the disposition is okay
... or if more needs to be done, and let me know.
<dsinger_> By mid-June I would like to know how the WG feels on the disposition of each of these
<dsinger_> Thierry and I will be verifying the commenters' reactions to each of them
<dsinger_> The test suite was recently improved
Dave: I hope that by the middle
of the month we have a new WD, agreed dispositions
... and a test suite report, and then to publish the CR. I'd
like to be at CR by TPAC.
... Then we will discuss if we will have to remove features to
move to Rec. Now with the
... test suite and the spec dealt with... Oh one more request,
look at the GitHub issues
... that were not wide review and say if there are any there
that need to be fixed. I didn't
... think myself that any urgently need to be addressed but you
may disagree.
Nigel: Is there a requirements doc for WebVTT that we can review against?
Dave: I don't think there was
ever a formal requirements document. We took it on as a
... working document from WHATWG, historically.
Thierry: I'd like to understand a
bit more, because it's not clear to me exactly what are
... the milestones. We are planning to publish on /TR a new WD,
that I understood.
... Then in parallel we are asking the TTWG to review the
comments that were sent and
... the proposals. I don't see what the TTWG could say now if
it's been approved by the CG
... and by the commenter, what could the WG bring? Then are we
going to trigger a new
... Wide Review before going to CR? The WR was 2 years ago and
there have been
... substantive changes since then.
<dsinger_> Push new WD now; review the disposition of the last round of wide review comments [2 weeks]; then we'll do another (briefer) wide review;
Dave: Yes push a WD now, review
the disposition at WG level over a couple of weeks
... if that is enough, then another WR request, and hopefully
fewer comments will arrive,
... and by then we will have a test report and I'm hoping we
can deal with any comments
... and then move to CR pretty rapidly.
Thierry: We don't need tests to get into CR. It's good if we do but it's not a requirement.
Dave: I don't want to enter CR without the test suite being reasonably complete and working.
Nigel: Have all the dispositions gone back to the commenter?
Dave: They have all been worked
through in dialog with the commenter.
... Nonetheless someone here might notice something we haven't
noticed. Formally the
... WG needs the opportunity to say if the disposition is not
good enough. I don't want
... to short-circuit the WG here.
Nigel: There was a question about how long we need to review. Is 2 weeks enough?
<dsinger_> https://www.w3.org/wiki/WebVTT_Wide_Review
<dsinger_> https://w3c.github.io/webvtt/
Thierry: Another issue: why are
we publishing a WD in a few days and then waiting
... 2 weeks or so for the WG to respond? Why not directly
publish a WR document?
Dave: We could get WR on this WD if you like, and do them in parallel.
Thierry: I don't see the difference actually.
Dave: Yes let's parallelise it, if you can help publish the WR requests Thierry.
Thierry: Okay, if the TTWG
publishes this as a WR we need to agree a review period
by
... the public.
Dave: Yes, what's typical?
Thierry: I think at least 4 weeks, given it's summer.
Dave: I'd be happy with end of
July assuming we publish the WD in the next day or two.
... OK?
... Thierry if you could help with this that would be
helpful.
Thierry: I tried yesterday but it
uses Bikeshed and I don't know how that works - I spent
... 4 hours yesterday trying to understand it and I still could
not achieve it. I need to go
... through the usual former publication process by the
WebMaster unless Philippe can
... give me some help.
Dave: Silvia might be able to help.
Thierry: Bikeshed is typically used in CSS WG and not many other groups.
<dsinger_> Silvia and/or Philippe (or the CSS folks) can probably help. Bikeshed is widely used, I think
Nigel: So it's going to be 4 weeks post-publication?
Thierry: That's the minimum especially in summer.
Nigel: Do you want longer?
Thierry: I'm saying it's the minimum.
Dave: It's the second round and
all the changes have been as a result of the previous
... round of wide review.
Thierry: I think what we should
do is restrict the wide review to the new features that
... were added, so I don't know if we have a list of
substantive changes that were done.
Dave: That's a good point.
Thierry: You can ask in the review to review only the delta.
Dave: Yes, I will work with the
Editors on working out what that list is.
... If you can help me work out the differences Thierry I will
do that.
Thierry: It says that on the wiki page.
<dsinger_> previous Wide Review was https://www.w3.org/TR/2014/WD-webvtt1-20141113/
Dave: I'll do a review and
indicate the significant changes since then.
... Sounds like a plan - we'll push the WD, analyse the
differences, get the WG dispositions agreed.
Nigel: We need the deltas for the WD for WR, which normally go in the SOTD.
Thierry: So we do a standard WD
publication in the next few days, then a WR WD when
... we have that ready.
Dave: Ok
Thierry: I will work with
Philippe to publish the current version as the new WD and if
I
... can't do that I will ask the WebMaster to do it on
Tuesday.
Dave: Ok, thanks.
... Thanks for getting the details of exactly what we need to
do. Thierry please help
... to make sure we don't drop any of these.
... Thanks everyone, [drops off]
action-498?
<trackbot> action-498 -- Nigel Megitt to Invite i18n to discuss imsc 1.0.1 issues -- due 2017-06-15 -- PENDINGREVIEW
<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/498
Nigel: Huge apologies for this
dropping off my radar, but I entirely failed to do it.
... I mean to ping them again after last week's meeting. The
original invitation was sent.
... Obviously Richard and Addison were pinged on the GitHub
issues in any case.
... For now I want to close the action since they did not
respond.
close action-498
<trackbot> Closed action-498.
Nigel: In terms of the timeline
Pierre, you have closed the issues where we said we
... would do so after a deadline.
Pierre: There are two ARIB WR issues that can't be closed until tomorrow unfortunately.
Nigel: We haven't seen anything from ARIB - Thierry?
Thierry: No that's correct. If we don't get anything by tomorrow I propose to close the issue.
Pierre: The issues that remain are:
Add diff from IMSC 1.0.0 and update substantive-changes-summary.txt
Pierre: I can't do that until #227 and #228 are closed tomorrow, so I will have a draft ready for publication on Monday.
Nigel: Right now I think it's
quite unlikely we will get a response in time, but it's
possible.
... Is it worth preparing #244 in anticipation of no more
changes?
Pierre: I was going to do that
over the weekend so it would be available to Thierry on
Monday.
... The only outstanding point without an issue is what we do
with the schemas.
... We have dealt with it for IMSC1 but we need to agree how we
are going to publish
... XSDs going forward.
Nigel: It's good to raise that. I put forward one view, but it may be too much change for some people.
Pierre: The goal is to reference
the XSDs from the spec in a way that can be referenced
... and updated by group Consensus, right?
Thierry: I agree.
Nigel: I agree too.
Thierry: If the schemas were
normative we would have no choice, but since they are
... not normative we can publish them anywhere. There are no
restrictions or guidelines.
... In the past we published them on the W3C site at some URI
but now people are using
... GitHub. I'm fine with either proposal. I don't think GitHub
is very friendly. I prefer a
... page that directly shows up in my browser but that's really
personal. I think w3.org
... is a more stable URI because we don't know what will happen
with GitHub one day.
Nigel: You can download that on GitHub directly.
Thierry: Yes but its hard to find.
Pierre: My main concern with
GitHub is that anyone can change the tags, so I would
rather
... have a more formal publication step on the W3C site.
Nigel: I don't mind either
option. Would we publish a wrapper page or would the URL
... just point to a directory?
Thierry: It would be like IMSC 1 with a wrapper.
Nigel: If its a wrapper then we
can do both, since it will always be on GitHub anyway.
... Then people who find GitHub easier can use that, or people
who prefer the direct download
... can get it from a W3C resource.
Pierre: I would be having them in a GitHub repo with a tagged release anyway.
Thierry: Would you prefer separate downloads or a ZIP to be downloadable?
Pierre: Right now we have a
wrapper page? I would do exactly the same and what I
would
... need from Thierry is that W3C page so I can put the XSD in
the doc.
<tmichel> I propose to host those schemas at
<tmichel> https://www.w3.org/AudioVideo/TT/IMSC/ttml-imsc1.0.1/xml-schemas/
<pal> the same as https://www.w3.org/TR/2017/WD-ttml-imsc1.0.1-20170322/xml-schemas/
Nigel: I thought that was a
directory listing but I see it's actually a wrapper page.
... So we can add to that wrapper page a link to the GitHub
page with the latest release tag?
Pierre: If they want to go to GitHub then they can just go there.
Nigel: How would they know to do that?
Pierre: I'm reluctant to point to
a specific release tag or path on GitHub because that
... can change. We can point to the repo, or if you think a ZIP
is better we can just put a
... ZIP there. I'm trying to avoid duplication.
Thierry: Me too, I'd like to avoid that.
Nigel: I don't think it is
duplication - or any duplication is from the GitHub repo to
the
... w3 site.
Pierre: By design or mistake it's really easy to change the release tags.
Nigel: I'd like to check maybe
with Philippe if we can manage the rights to create or
modify
... release tags, and if we can add it then do so.
Thierry: Anyhow we can modify the page anytime to add new links if we want.
<pal> https://www.w3.org/AudioVideo/TT/IMSC/ttml-imsc1.0.1/xml-schemas.zip
Pierre: The link now will be slightly different because it will have a URL like the above?
Thierry: I thought we would have
a wrapper page and then in that wrapper page we can
... add a link to GitHub if we want. And we can even add a link
to a ZIP file from that
... wrapper page.
Pierre: Okay that sounds fine to
me.
... I'll add a new issue for moving the XSDs to outside of
/TR.
Thierry: You know when we changed
the XSD in place I put a comment in the XSD to
... say that it was changed inplace on June 23 so I propose to
remove that for 1.0.1.
... It was just to explain the change on the Rec home page.
Pierre: Yes, anyway, it's not in the GitHub version.
Thierry: Should I take the GitHub version?
Pierre: Absolutely.
Thierry: OK I'll start from that.
Nigel: Okay I'd like to propose
CR publication based on no changes coming from ARIB.
... Did we set a review deadline?
Pierre: Yes we said August 6.
Thierry: Okay I have prepared a
transition request document which I will send to the
... Director because we don't have any formal objection anymore
so I hope we don't need a call.
Nigel: Oh thanks for the reminder
- Glenn told me he will not object formally to CR
... publication as v 1.0.1.
PROPOSAL: Publish IMSC 1.0.1 based on the current ED with request for review by August 6 assuming there are no changes due to comments by ARIB tomorrow and the differences issue is resolved.
Pierre: CR publication for July 6 and end of CR to August 6?
Thierry: Yes.
Pierre: Add diff from IMSC 1 and
add substantive changes summary and update the
... XSD link.
... These are all captured in #244 and #248 in GitHub.
Nigel: Is there any more work to do on the Disposition of comments for the CR transition request?
Thierry: There are some issues I'm not sure if they are closed or not.
<tmichel> https://www.w3.org/wiki/IMSC1.0.1_Comments_tracker
Thierry: What remains is the Call
for Review to W3C groups - only i18n responded.
... Nigel if you can provide me with the exact list of external
bodies that you sent the
... calls for review to?
Nigel: They are all in the member-tt archives, if you could get them from there.
Thierry: I will do that.
... Now the comments themselves: we have only 1 comment from
ARIB and we are...
Nigel: Sorry to interrupt, didn't we get a positive comment from DVB?
Thierry: Maybe - if you find it send me the link.
Nigel: Will do.
Thierry: Now we have 3 issues: #236 with no answer from Richard?
Pierre: True, no further communications since 2 weeks ago.
Thierry: Should we close that and say it's done?
Nigel: It's already closed.
Thierry: The next from Richard is
done and the last from Addison also closed due to no
answer.
... Then there are 9 comments from WG members. I'm not sure the
latest status on some of them.
... 5 are marked as pending - Pierre, what is the latest status
for those?
Pierre: They're all on GitHub right?
Thierry: Yes but it's not clear to me if it was agreed.
Pierre: #221 the commenter never
agreed.
... I think it resolves the comment.
Thierry: So there's no response from the commenter?
Pierre: Correct, nobody disagreed.
Thierry: okay that one is "no
response, closed".
... Then we have #223.
Pierre: This was just a question.
Nigel: And the person who asked it closed the issue.
Thierry: Ok that's closed.
... Now #228?
Pierre: That's one from ARIB.
Thierry: OK I understand now.
I've been tracking it in the archives only.
... Then #232 and #242.
Pierre: #232 was my comment and I'm happy with the resolution.
Thierry: Thank you.
... Then we have #234.
Pierre: Andreas created it and it is part of the solution to #233.
Nigel: Andreas proposed the pull
request to resolve #233. #234 was merged.
... Since the original issue was #233 that was raised by
Andreas I think we can conclude
... that he was happy with it.
Thierry: Ok, the next is #238.
Pierre: That was closed by the commenter.
Thierry: Ok, the last one is #240.
Pierre: that was mine, and I'm happy with it.
Thierry: Okay I will finish the
disposition of comments today so that we are ready and
... then I'll send the transition request on Monday because I
need to have the CR version
... document. Okay?
Nigel: Please could you send me the updated draft transition request before you raise it?
Thierry: I'll send it to you on
Monday because I need the final details, and I have to
remove
... the placeholder for the formal objection.
Nigel: My diary is okay on Monday so I should be able to scan it and return it fairly quickly.
RESOLUTION: Publish IMSC 1.0.1 based on the current ED with request for review by August 6 assuming there are no changes due to comments by ARIB tomorrow and the differences issue is resolved.
Nigel: Thank you all, is there anything else on IMSC?
group: [silence]
Nigel: Our goal here is to
publish the current ED as a WD for Wide Review.
... Current status: We have been discussing ttp:mediaOffset
offline, and there is one comment on the open pull request, to
complete #195.
... #396 is the audio related features pull request and I just
asked for the features
... to be added to profiles where they are currently omitted,
for completeness.
... I don't know why Glenn hasn't responded to that comment, or
done it - anyway
... I can do it.
... On the mediaOffset issue #323 I don't want to hold back WR
publication - I guess
... we can still remove between WR and CR if there are
supporting negative comments.
Pierre: I agree with you Nigel
that there's no use case for that feature and it will
actually
... cause harm so I would like to remove it (see the
issue!).
... However given the desire to publish the WD for WR I would
not hold up that publication.
... For the record Movielabs would not object to the
publication of the WD for WR as long
... as the review period is no less than 3 months to give
adequate time for review including
... by groups with long review cycles.
PROPOSAL: Publish a WD for WR of TTML2 based on the current ED, with a review period ending at end of September.
Nigel: I think we can probably
add the audio feature designators after WR publication though I
would prefer to include them before.
... In that case the only action will be on Glenn to update the
SOTD for the review period
... and then publish with echidna.
Dae: Can I close the open i18n issues where I haven't had a response yet?
Nigel: No they can stay open until we need to move to CR.
Pierre: You can definitely not close those - we haven't even given a deadline for feedback yet.
Dae: I want to close off the
issues that are related to printed publication not to
... subtitles.
Nigel: What we need to do is go
through each issue, for you to propose that disposition,
... for the group to agree it, and then to go back to the
commenter and check they are okay with it.
Dae: Okay sure.
Nigel: For the proposal, I'm hearing no objections?
Dae: Yes, and that keeps us on track for end of 2017.
Nigel: It does if we have implementations so we are confident of moving out of CR quickly.
Dae: Yes.
RESOLUTION: Publish a WD for WR of TTML2 based on the current ED, with a review period ending at end of September, preferably though not necessarily with pull request #396 merged.
Nigel: Anything else to cover now on TTML?
group: [silence]
action-497?
<trackbot> action-497 -- Nigel Megitt to Invite csswg to joint meeting at tpac 2017, with list of topics. -- due 2017-06-15 -- OPEN
<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/497
Nigel: Apologies this has slipped another week - I will get around to this!
Pierre: Some progress here on the
PQ HDR in PNG. Adobe has officially released the
... sample ICC profile for publication by W3C.
... I never heard back from Chris Lilley so I opened a pull
request to try to address his concerns.
... Also I've heard concerns that this PQ in PNG approach,
because PNG is limited to 8 bit
... or 16 bit, and 8 is not sufficient for general HDR use and
16 is not efficient, so PNG
... is not suitable for this use case. I've encouraged the
concerned party to raise a comment
... on GitHub as an issue.
... I think it's a legitimate concern. It means that the scope
of the document goes from
... being general to specifically how do I do PQ in PNG until
there's a better solution.
... The scope of the document will be limited to be just for
subtitles and captions.
Nigel: It's a general purpose format so how can you tell people what they can or cannot use it for?
Pierre: The document would say it
is appropriate for that use but may not be appropriate
... for other applications.
Nigel: I see.
Pierre: My plan is, if we get
that comment, to address it and maybe in a couple of
weeks
... have something for the group to consider.
Nigel: Thank you!
Pierre: I've also learned
yesterday that this is actually in use today - these kinds of
PNGs
... are being exchanged now to address those markets that
accept HDR content.
... Documenting that is a good idea.
Nigel: Are they doing that in the absence of any referrable document?
Pierre: Yes they'd really like a referrable document!
Nigel: Thanks, I think we've covered everything on our agenda. [adjourns meeting]