Timed Text Working Group Teleconference

<trackbot> Date: 16 October 2014

<scribe> scribeNick: nigel

This meeting

Santa Clara F2F October (TPAC)

http://www.w3.org/wiki/TimedText/tpac2014

nigel: goes through agenda as currently drafted.

glenn: I may need to leave slightly early on the Monday to attend another event.

courtney: Maybe we should reserve a short period for WebVTT publication discussions

nigel: Lists the applicants to be observers.
... I intend to accept all the observer applications.
... I'll send the emails tomorrow
... It turns out that advanced registration is now closed, but I believe that registration is available on the day in person.

Action items

action-338?

<trackbot> action-338 -- Glenn Adams to Update change proposal 1 to reflect new approach and move to closed-implemented. -- due 2014-10-16 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/338

glenn: That's done.

close action-338

<trackbot> Closed action-338.

action-339?

<trackbot> action-339 -- Glenn Adams to Fill in the relevant sections of change proposal 12 and move it to closed-implemented. -- due 2014-10-16 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/339

glenn: That's done.

close action-339

<trackbot> Closed action-339.

action-340?

<trackbot> action-340 -- Glenn Adams to Update change proposal 24 and move to closed-implemented -- due 2014-10-16 -- PENDINGREVIEW

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/actions/340

glenn: That's done.

close action-340

<trackbot> Closed action-340.

Issues

issue-335?

<trackbot> issue-335 -- In order to handle offsets between start time in TTML docs and start time in video, allow negative times to be used in fragment begin times. -- pending review

<trackbot> http://www.w3.org/AudioVideo/TT/tracker/issues/335

nigel: We've had quite a bit of debate about this and decided to defer this until TPAC

glenn: I'll prepare a short presentation on this for TPAC.

Change Proposals

glenn: The only ones that require a significant amount of conceptual work and editing is the HTML5 mapping.
... The raw content is outlined already, so I hope not to have to do much conceptual work there.
... Before we go too far in the Rec process though we need some implementation support
... for the HTML mapping, so I need to do some work on that. I've already outlined some
... tasks on TTV to allow me to do some experiments with HTML mapping, which would be very useful.
... I'm also refining the ISD mapping process, which is the other area I need to spend
... some time on: the semantics and creation procedure. There's some material there
... already derived from TTML1. Implementation work is progressing on that, and it
... will help document the algorithm.

nigel: Is that for a Change Proposal?

glenn: It may come under Distribution and is also a pre-requisite for the HTML5 mapping too.

nigel: Do we need to liaise with the HTML WG on that?

glenn: I hope not!
... The only reason we might want to is if there's a specific issue we may want to
... discuss with them. I'm not aware of any. I'm not making any normative use of, e.g.
... the TextTrackCue APIs. THere's an open issue on HTML5 regarding the DataCue
... interface but that won't impede us here.

nigel: I was wondering if we were headed towards an HTMLCue.

glenn: I'm hoping we don't need to add that into TTML2. We have some pre-draft specs, for TTML1 and TTML2 APIs
... which will probably make reference to those. When we make further progress with
... those specs that issue will come up more.

nigel: We don't have those in our charter at the moment.

glenn: They're generally under TTML2 but not specifically marked as Rec Track deliverables.

jdsmith: TextTrackCue isn't really in a position now to be used for TTML now.
... Should we have a TTMLCue or something like that?

glenn: That's the question, it may not be the right time to discuss it now.

jdsmith: Should we track that as an issue in the TTWG? It's not in any of our specs now.

glenn: It's implicitly under the TTML API level 1 and 2 because they make use of TextTrackCue
... and define a TTMLCue interface. That's part of that discussion if we're going to
... move forward to Rec on that area.

<glenn> https://dvcs.w3.org/hg/ttml/raw-file/default/ttml1-api/Overview.html

Thierry: Apologies for joining late

<glenn> https://dvcs.w3.org/hg/ttml/raw-file/default/ttml2-api/Overview.html

MIME type processor profiles short codes parameter

nigel: We need to get to a message that we can send to MPEG.

dsinger: We have that: TTWG will define one or more parameters to indicate the dialect or sub-brand of TTML.
... We don't need to declare the format for that right now.
... For example application/ttml+xml;[subparameter]
... and the TTWG is responsible for defining [subparameter]

mike: We've agreed the syntax too, we just need to agree on the name of the parameter.
... I'll draft a message to send to MPEG.

nigel: We've already got time set aside at TPAC to look at the registry page. Are there any other actions?

Mike: We should go ahead and post the response.

dsinger: Let's write a contribution from me and Mike to submit at MPEG rather than a liaison that might take a while to get through.
... On the profile name, there are some proposals: procprofs, rprofiles, what else?

glenn: I vote for codecs

dsinger: That's used elsewhere for H-264 etc.

glenn: In this case it's codecs related to TTML.

dsinger: It's not a formal problem, just one of being informative.

nigel: We need to conclude that we're not re-using "profile".

dsinger: I'd rather not redefine it.

glenn: We shouldn't touch that.

nigel: Presumably a new parameter should have a reasonably short name.

dsinger: That's why I suggested procprofs.

nigel: Me too.

Mike: In TTML2 do we have processor profiles or content profiles?
... For interpreting the "profiles" parameter.

glenn: It is a processor profile.

mike: I would offer that procprofile is more confusing.

dsinger: What's the equivalent name inside the document?

glenn: It's ttp:processorProfiles but it doesn't take short codes.

Mike: profile has effectively been redefined already to be processor profile. We need a new label for the new syntax.

glenn: The original profile parameter in TTML1 referred to a single profile only, not multiple.
... It was a little vague on whether it meant content or processor profile. We've
... deprecated it in TTML2 and added processorProfiles and contentProfiles to
... distinguish those two uses.

Mike: How about short-profile?

nigel: How about short-profiles?

dsinger: Would "profiles" in addition to the original "profile" be too confusing.

nigel: I think so.

dsinger: I'm back to "procprofs" to allow room for "contprofs" later.

glenn: If you're doing that, then "processor-profiles" would be better.

dsinger: The length isn't going to be a problem.

glenn: It's less likely to be confused, and is a semantic equivalent.

dsinger: We need to call out that the codes are short names not full ones in the documentation.

RESOLUTION: We will adopt the parameter named processorProfiles
... We will host a registry page for this parameter including the long name that goes in the TTML2 document and the short name for the MIME type and a pointer to the ttp:processorProfiles parameter definition

<inserted> chair: dsinger

WebVTT publication

dsinger: Thierry pointed out that we haven't published a formal FPWD for WebVTT even though it's implemented in various places.
... To publish a FPWD, so we can call for wide review. We need a resolution from this WG to do that.
... nigel and I and Silvia have been working on this. We need a resolution to publish a FPWD. I was hoping to have a final
... draft of this to look at today. We're asking for formal feedback to come back to the TTWG and bugs to go to the CG tracker.

http://dev.w3.org/html5/webvtt/webvtt-staged-snapshot.html

nigel: This doc has a SOTD saying comments should go to public-tt and bugs go to the CG's bugzilla with some
... distinguishing information to show that it originated on the FPWD.

pal: So the document proposed to be moved to FPWD needs to be offered for review to the WG. Has that been done?

dsinger: That's what we're doing now. It's the same as the editor's draft that has been around for a while, with some
... boiler plate changes.

pal: I'd like some time to review this before we agree to move to FPWD.

dsinger: The FPWD doesn't imply any endorsement, but of course you can have some time to review it.

pal: I recommend this group gives itself at least a week to review it.

dsinger: OK, that's fine.
... regrets for next week from me. I'd like to have this resolved before TPAC so we can socialise it there.

nigel: Okay I'll put that on the agenda for next week and we'll resolve to publish if there are no objections.
... What about the short codes? The proposals are webvtt and webvtt1

glenn: I'd suggest going for webvtt1 because there's likely to be a webvtt2

dsinger: If the team can alias webvtt to the latest version that's fine too.

RESOLUTION: At the time we resolve to publish as an FPWD we will use the short name webvtt1

dsinger: My intention for demonstrating wide review to get to CR is to contact the W3C groups HTML, CSS etc, and at
... least MPEG and so on, and that it will get out in public lists etc.

nigel: You could use the same list as we used for IMSC1 as a basis, which was trawled from the Charter and the W3C liaisons page.

dsinger: I'll draft that email in anticipation of the resolution.

<Dsinger_> Apologies for running over

<inserted> chair: nigel

AOB

glenn: Quick announcement: I'm officially representing Skynav in this group now.

nigel: Thanks everyone, for a very productive meeting. [adjourns meeting]