Timed Text Working Group Teleconference

08 Oct 2015

See also: IRC log


nigel, pal, tmichel
Andreas, Frans, Mike


<trackbot> Date: 08 October 2015

<scribe> scribe: nigel

This Meeting

nigel: Most of this meeting will be dedicated to closing IMSC 1 pending review issues
... AOB?
... Mike has asked to postpone the incoming liaison from ATSC issue.

pal: There's also the topic of profile designators - it would be good to talk about it.

nigel: Great, that's added.

Action Items


<trackbot> Action-421 -- Nigel Megitt to Check with ttml player manufacturers if they're happy to be listed on our site. -- due 2015-09-24 -- OPEN

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/421

nigel: I did contact the manufacturer we discussed and they said they'd check but haven't yet confirmed.
... Do we have an implementations page?

tmichel: I don't think we have one yet but I'd like to start one. I think we should have a page
... per technology - one for TTML, one for WebVTT etc.

pal: We have a page for that - remember the contribution from the Web & TV IG?

nigel: Oh yes, I remember now.


https://www.w3.org/wiki/TimedText/EffortsAndSpecifications is a better URL to use

It's linked from https://www.w3.org/wiki/TimedText under Timed Text Efforts and Specifications

nigel: I guess we could add it to the group home page too - we probably should tidy up the
... group page vs wiki home page discrepancy. Having both isn't especially helpful.

tmichel: In former times we had the CVS hosted home page which was hard to edit. Now
... the wiki is simpler for people to add information to. Everyone should be able to add their
... own product if they want to.

nigel: Is it worth an action to redirect from the home page to the wiki?

tmichel: I'll look into that. There's some information like how to join on the home page.

nigel: Yes. we'd have to migrate the key information.

tmichel: I'll look into it and propose something.

<scribe> ACTION: tmichel Look into merging the group home page and the wiki home page and propose something [recorded in http://www.w3.org/2015/10/08-tt-minutes.html#action01]

<trackbot> Created ACTION-430 - Look into merging the group home page and the wiki home page and propose something [on Thierry Michel - due 2015-10-15].

TPAC 2015

nigel: Netflix has just joined the WG and would like to spend some agenda time on the Thursday
... presenting on their open source TTML2 implementation work.
... I'll add that to the topics/agenda page on the wiki.

tmichel: While we're talking about new participants, SImon from Opera joined today replacing Philip.
... Simon is the new editor of the WebVTT spec.

IMSC stuff...


<trackbot> action-425 -- Thierry Michel to Send a prompt/thank you to the potential test result contributors for imsc. -- due 2015-10-01 -- OPEN

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/425

pal: I've not heard back and am still chasing, so that's still open. No update despite best efforts.

Profiles registry

pal: Related to Cyril's email earlier today, which came out of this topic, I think from an
... implementation standpoint he wants users to be able to signal clearly in a TTML document
... what profiles a document conforms to. In IMSC 1 right now there's a 'MAY user ttp:profile'
... statement but that's it. Maybe we can add more to IMSC 1 to encourage people to label
... their documents. The challenge is, as we discussed before, EBU-TT-D forbids ttp:profile.
... And ttp:profile supports only one profile. So if a document conforms to both EBU-TT-D and
... IMSC 1 then you should not use ttp:profile but you should use ebuttm:conformsToStandard
... and list both. If the document is not EBU-TT-D conformant it probably isn't a bad idea to
... use ttp:profile and also include ebuttm:conformsToStandard so I'm thinking of adding
... a recommendation to IMSC 1 along those lines.

nigel: I can see the logic to that, yes. It makes sense I think.

pal: I'd like to be responsive to the request.

nigel: I agree - there's a use case we haven't spent a lot of time considering that is obviously
... important. I'd like Andreas's and Frans's input on this specific proposal too, but on the
... surface I think it makes a lot of sense.
... You can recommend using the URI designators.

pal: That would definitely be the recommendation!
... From a tools perspective it would certainly help tools to determine the document's profile.

nigel: We just need to be careful that ebuttm:conformsToStandard is indicating document
... conformance in other words the equivalent to TTML2's ttp:contentProfiles attribute.
... Whereas ttp:profile indicates a processor profile, so we shouldn't mix the messages there.

pal: I don't know why we hadn't collectively thought about this before. I think it works in exactly the right way.

nigel: I can see how this would work really well - if an IMSC 1 processor is needed then it's on
... the ttp:profile attribute and if not, because an EBU-TT-D processor is okay, then it would be
... omitted.

pal: I think we can file an issue for this, especially given Glenn's email just now proposing that
... IMSC 1 includes a 'should include ttp:profile attribute' statement, then implement and review
... next week.
... [filing issue right now]


<trackbot> issue-448 -- Add recommendation for ttp:profile and ebuttm:conformsToStandard -- raised

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/448

reopen issue-448

<trackbot> Re-opened issue-448.

pal: I'm going to reply to the thread too.

nigel: I think it's worth pinging back to EBU to check if they might reopen the discussion on
... whether or not to permit ttp:profile on EBU-TT-D documents.
... Of course if you have a stream of EBU-TT-D documents e.g. in DASH then this is likely
... to be redundantly repeated data.

pal: Of course the packager could strip it out if it knows the data won't be needed downstream.

nigel: We may be putting packagers in an uncomfortable situation of needing to modify the
... samples they're given to wrap up.

pal: True - we're talking about a few bytes though and it's tiny in comparison with e.g. HTTP headers.


nigel: Let's go through Pending Reviews...


<trackbot> issue-404 -- Where is #image feature defined? -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/404

pal: I think this is unambiguous, well specified and can be closed.

nigel: +1

close issue-404

<trackbot> Closed issue-404.


<trackbot> issue-405 -- why is #nested-span prohibited on image profile? -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/405

nigel: This isn't even a proposal to change the spec. It's a question with an answer.

pal: I think Glenn would have responded if he disagreed.

close issue-405

<trackbot> Closed issue-405.


<trackbot> issue-406 -- #lineBreak-uax14 is never 'used' by a document? -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/406

pal: uax14 is mandated for processors in the text profile and stated as not applicable to document instances.
... The image profile is silent about it.

nigel: Do we permit the profile features to be listed in the document?

pal: We do. The presence of the profile designator in the document requires that the processor
... implements uax14. What the spec currently says is the document constraints. The uax14
... feature is not applicable to constraints on the document.

nigel: There's a niche issue for the text profile: if the profile features are included then it would
... be an error to state that the #linebreak-uax14 feature is a 'shall not use'.
... That would create a situation where an IMSC 1 text profile conformant processor should
... reject an IMSC 1 text profile conformant document that includes this profile feature as
... 'shall not use'.

pal: This is a strange and theoretical case, that applies only to the uax14 feature.

nigel: We could add a statement that if profile features are included within a document then
... they shall not conflict with the processor conformance rules for IMSC 1.

<scribe> ACTION: nigel File issue re profile feature inclusion in IMSC 1 documents [recorded in http://www.w3.org/2015/10/08-tt-minutes.html#action02]

<trackbot> Created ACTION-431 - File issue re profile feature inclusion in imsc 1 documents [on Nigel Megitt - due 2015-10-15].

pal: We could add that to the profile section for document conformance.
... Going back to issue-406, Glenn suggested to require the feature. Shall we start the 2 week review clock on this today?

nigel: yes, let's do that.


<trackbot> issue-408 -- Initial value of color -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/408

close issue-408

<trackbot> Closed issue-408.


<trackbot> issue-409 -- Font family of default means monospaceSerif -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/409

pal: IMSC 1 defines what it means for "default" fontFamily, i.e. what a processor shall do when
... the fontFamily is set to "default". Glenn points out that this isn't always the right thing to do.
... My response to that is that authors should not rely on default. TTML1 doesn't prohibit
... implementations from mapping the default font to a known one.

nigel: I'm a bit nervous here - Glenn hasn't proposed a solution, and I don't know why.
... For example we could include a table mapping language to default font map or reference
... one in e.g. CLDR.

pal: Document authors can make their choice though if default won't work for them.

nigel: My nervousness here is that we're making a language market-specific value for default,
... which could make use of default less useful.
... However Glenn knows most about this from those here and he hasn't picked it up.

pal: We should close this on the basis of no response in over 1 month. It can be reopened if someone is unhappy.

close issue-409

<trackbot> Closed issue-409.


<trackbot> issue-410 -- Constraints on #linePadding and #multiRowAlign -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/410

nigel: The problem with SHALL NOT here suggests a syntactic conformance requirement whereas
... what you mean is that if the attributes are present then they shall be ignored by implementations.

pal: We don't have Glenn's view on this so it would be uncool if he doesn't agree.
... Don't forget the problem arose in EBU-TT-D.

nigel: The new EBU-TT Part 1 v1.1 might get us out of that problem - it defines the features
... and states that they may be used in EBU-TT-D, and fixes the gaps from the original definition.

<scribe> ACTION: nigel Nigel, pal, andreas and frans to hook up re changing IMSC 1 EBU-TT feature references to point to tech3350 v1.1 [recorded in http://www.w3.org/2015/10/08-tt-minutes.html#action03]

<trackbot> Created ACTION-432 - Nigel, pal, andreas and frans to hook up re changing imsc 1 ebu-tt feature references to point to tech3350 v1.1 [on Nigel Megitt - due 2015-10-15].

issue-410: [meeting 2015-10-08] See also Action-432 - it may be possible to avoid the problem by referencing tech3350 v1.1.

<trackbot> Notes added to issue-410 Constraints on #linePadding and #multiRowAlign.

nigel: We're out of time so I'll adjourn now.

pal: Maybe we could do 2 hours next week to get through the remaining issues ahead of TPAC?

nigel: That's fine for me.

pal: I'd like to have gone through all of them - there's one that may need TPAC discussion.
... Glenn brought up an excellent question about the HRM, in relation to Arabic, where the
... mapping between code point and glyph is in the majority of cases is not 1:1 and the shape
... depends on the context. That really messes up the glyph buffer. I'm researching it.
... The glyph buffer was put in to support e.g. 608, 708 and Teletext. Hopefully we'll have closed all the others.

nigel: Thanks both. [meeting adjourned]

Summary of Action Items

[NEW] ACTION: nigel File issue re profile feature inclusion in IMSC 1 documents [recorded in http://www.w3.org/2015/10/08-tt-minutes.html#action02]
[NEW] ACTION: nigel Nigel, pal, andreas and frans to hook up re changing IMSC 1 EBU-TT feature references to point to tech3350 v1.1 [recorded in http://www.w3.org/2015/10/08-tt-minutes.html#action03]
[NEW] ACTION: tmichel Look into merging the group home page and the wiki home page and propose something [recorded in http://www.w3.org/2015/10/08-tt-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.140 (CVS log)
$Date: 2015/10/08 15:20:27 $