15:59:08 RRSAgent has joined #tt 15:59:08 logging to https://www.w3.org/2022/02/17-tt-irc 15:59:10 RRSAgent, make logs Public 15:59:11 Meeting: Timed Text Working Group Teleconference 16:00:29 atai has joined #tt 16:00:58 Agenda: https://github.com/w3c/ttwg/issues/210 16:01:07 Previous meeting: https://www.w3.org/2022/02/03-tt-minutes.html 16:03:09 s/will be late, still in previous meeting// 16:03:30 Present: Andreas, Gary, Nigel 16:03:35 Chair: Gary, Nigel 16:03:38 scribe: nigel 16:05:06 calvaris has joined #tt 16:05:09 Present+ Xabier 16:05:36 Topic: This meeting 16:05:47 Regrets: Pierre 16:06:06 Nigel: Today, quick update on IMSC HRM Wide Review 16:06:17 .. then Rechartering status update, and 16:06:23 Present+ Cyril 16:06:33 cyril has joined #tt 16:06:37 Nigel: Dubbing and AD requirements. 16:07:02 .. Is there any other business or topic to make sure we cover? 16:07:22 [no other business] 16:07:29 Hi, I'm Xabier Rodríguez (aka calvaris), from Igalia. I work on the WebKit GStreamer ports 16:07:46 Nigel: Xabier, welcome, I think this is your first call, great to have you here with us. 16:08:49 .. When you get your membership of the group confirmed, please send an email to member-tt@w3.org with 16:08:59 .. a quick introduction just so we can verify that everything is set up properly. 16:09:40 .. And welcome to Igalia too. 16:10:37 Topic: IMSC HRM Wide Review 16:10:50 Nigel: Quick update: we're happy with the wide review comms wording, 16:11:04 .. but since we received quite strong feedback about the way the spec is introduced 16:11:18 .. and the purpose of it seems to need better explanation, 16:11:28 (sorry for late cut in, but please tell your AC rep for approval.. Igalia itself is already in this TTWG as organization.) 16:11:46 .. I wasn't happy to send the WR comms out before we have made some editorial improvements 16:11:52 .. to help address those comments. 16:12:14 .. Otherwise, my prediction is that we will get similar mis-comprehension from others, and that would not be helpful. 16:12:25 .. That's why Pierre is working on updates rather than being here! 16:12:39 .. Any questions or thoughts on that? 16:12:46 Present+ Atsushi 16:12:56 Present+ Xabier 16:13:11 [no questions] 16:13:22 +1 for updates before sending wide review requests 16:14:28 Topic: Dubbing and AD requirements 16:14:42 Nigel: This is actually related to the Charter in that it affects our plan of work. 16:15:47 .. Cyril and I have been talking offline and agreed to work together on a single TTML2 profile for both AD and Dubbing. 16:16:00 .. I sent a message to the AD Community Group, which is at https://lists.w3.org/Archives/Public/public-audio-description/2022Feb/0000.html 16:16:16 .. Just to keep them in the loop and see if there were any comments or feedback. 16:17:14 Cyril: [shares screen] 16:17:27 .. This is a merge of the AD profile requirements and the Dubbing workflow requirements 16:17:35 .. from the work that Netflix has been doing. 16:18:05 .. The current title is Requirements for the Movie Translation Dubbing and Audio description profile of TTML2 (MTDA). 16:18:12 .. I had some feedback that it's too long! 16:18:18 Nigel: We can change it. 16:18:30 Cyril: It's capturing the requirements. 16:18:42 .. The first thing we did is come up with a combined workflow. 16:18:49 .. The AD CG came up with a workflow diagram. 16:19:01 .. The Netflix TTAL blog post had another diagram, 16:19:05 .. so we came up with a combined diagram. 16:19:22 .. It starts with AV content. 16:19:36 .. Then identify times, for dubbing, dialog times, for AD, the gaps where there are opportunities. 16:19:48 .. Step 2 is authoring of text, in one case describing the images for AD, 16:20:01 .. in the other, producing the dubbing text by transcribing, then translating, then adjusting. 16:20:15 .. Step 3 is audio rendering, by recording an actor or synthesising. 16:20:27 .. Step 4 is mixing the audio. 16:20:41 .. Step 5 is editing the script to match the performance. 16:20:48 .. What's in scope is the green boxes. 16:21:01 .. Then just like the AD profile, after the workflow there's a description of the steps and the requiremenst 16:21:07 .. associated with each step. 16:21:12 .. We followed the same approach. 16:21:17 .. Then we have the requiremenst. 16:21:40 .. Nigel and I are still going through that, then we will produce a document. Is it a Note? 16:21:47 Nigel: Yes, the requirements can be a WG Note. 16:21:51 Cyril: Any other question? 16:22:03 Andreas: It looks very good, and good to see the two activities align, as it should be. 16:22:17 .. Nigel, the AD work would then end up in this new specification or profile, right? 16:22:24 .. And there will not be a separate document? 16:22:31 Nigel: Yes, that's the intention. 16:23:08 .. I think this is a good move because it increases the target community of users, 16:23:21 .. and that I think makes it more likely to get more implementations. 16:23:52 .. They're so close that I feel it's obvious that they should be together. I may be proven wrong in the future! 16:24:04 Andreas: Do you have a rough feeling about what you want in terms of the timeline 16:24:16 .. for a FPWD of the specification document as opposed to the requirements. 16:24:48 Cyril: I expect the requirements to be done in 2-3 weeks, and I don't expect any issues. 16:24:52 .. We can always refine them afterwards. 16:25:13 .. For the spec, it's a good question. I think the FPWD could come pretty early, maybe within 2 months from now? 16:27:13 Nigel: Yes, I agree, mid-May would be my hopeful cut-off, since I have 3 weeks vacation then! 16:27:27 .. I think most of the substantive parts of the spec will be pulled from TTML2 features, so not too difficult. 16:27:49 .. The thing that will take some thought and consideration is where we source the metadata vocabulary 16:27:54 .. to support the workflow parts. 16:28:54 s/menst/ments 16:29:05 .. In terms of actions, I think we need a repo per document, 16:29:43 .. so one for the requirements and one for the spec. 16:30:44 Cyril: If we remove Movie Translation then we get DAPT for Dubbing and Audio description Profile of TTML2. 16:30:47 Nigel: Googling that... 16:31:06 .. Just Dual anti-platelet therapy. I think that's sufficiently different. 16:31:26 .. Happy to go with that. Any other thoughts about that? 16:31:42 PROPOSAL: Call this new thing DAPT for Dubbing and Audio description Profiles of TTML2 16:31:48 Nigel: Any objections? 16:31:57 Andreas: No, it's good 16:32:19 Cyril: Maybe we should open it up to others not on the call - we don't have to decide it right now. 16:32:29 Nigel: I'll highlight it in the minutes. 16:32:53 .. Atsushi can I ask you to make the repos when we know the name of it? 16:33:04 Atsushi: OK, yes. 16:33:15 Nigel: We'll leave ADPT as is and move the work to DAPT. 16:33:36 .. Any other questions or thoughts on this? 16:33:45 filed a tracker -> https://github.com/w3c/ttwg/issues/214 16:33:46 Gary: ADPT could link to DAPT when that exists. 16:34:04 Cyril: Yes we should add a link to ADPT saying the work is superseded by DAPT. 16:34:08 Gary: Exactly, yes. 16:34:13 Nigel: Good idea. 16:34:26 .. That will be one for me to do I suspect. 16:34:53 .. Thanks for that Cyril, everyone. 16:34:58 Topic: Rechartering status update 16:35:25 Atsushi: I will briefly show the update. 16:35:40 .. The draft charter is approved by W3M with two possible updates. 16:35:59 .. One is a choice because there's been some strong conversation about license within AC and W3M. 16:36:10 .. It might be better to switch to the Software and Document License. 16:36:30 .. The second is a request to add the ADPT work, or now, DAPT, as one potential item for profiles of TTML2. 16:36:55 Nigel: It is already there - do you need an actual deliverable? 16:37:07 Atsushi: There's no link there. It's just mentioned as an example application. 16:37:18 Nigel: OK, so the action is to add a link? 16:37:26 Atsushi: Yes, maybe DAPT is better? 16:37:29 Nigel: Yes 16:37:45 Atsushi: Let me add a link to DAPT for the newly created repository and maybe we should 16:38:01 .. prepare some basic content like a README before submitting to AC Review. 16:38:02 Nigel: Yes 16:38:13 Atsushi: Let me open a pull request. 16:38:17 Nigel: OK, thank you. 16:38:54 .. We haven't got everybody here for this but we should spend a bit of time thinking about the license question. 16:39:03 .. I've been doing a bit of digging here. 16:39:24 .. There are two W3C licences that can be used for Rec track documents. 16:39:37 .. The first, the Document License, is what we have used everywhere up until now. 16:39:55 .. It is less permissive for spec text reuse than the Software and Document License. 16:40:22 .. The concern as I understand it is that there's some possibility of W3C changing as an entity and 16:40:49 .. the desire is to be able to ensure that somehow work can definitely continue on specifications published 16:40:55 .. with copyright assigned to W3C as it is now. 16:42:33 .. My understanding is that in the past we chose the less permissive licence because 16:42:45 .. it is one tool for preventing spec forking, which could obviously be quite disruptive. 16:43:00 .. However Philippe has in the past assured us that other mechanisms exist that he would prefer to use. 16:43:06 .. I think that's where the debate lies. 16:43:31 .. It's also worth saying that the W3C Counsel, Wendy, has stated that she thinks the risk associated 16:43:37 .. with the Document license is very low. 16:44:20 .. I'm not in favour of change but happy to represent the WG's views. 16:45:16 Andreas: Did she say the risk is with the Document license not the software license? 16:45:20 Nigel: Yes, that's right 16:45:47 Andreas: And she said the risk is low? 16:46:03 Nigel: Yes, I'm not sure what W3M's view is, this view that people should change to the software license 16:46:13 .. was put forward by Mark Nottingham, who is not in W3M. 16:46:33 Andreas: And the spec forking concern: is that really a concern still, that someone would fork a TTML spec or one of the profiles? 16:46:46 Nigel: By the way, also WebVTT, as it is now, which surprised me. 16:47:03 Andreas: I thought it had the CG license still, Gary. 16:47:21 Gary: I'm not sure. 16:48:16 Nigel: It's not easy to find. The line above the Abstract tells you the "document use" restrictions, and when you click on that 16:48:24 .. in WebVTT you get to the Document License. 16:48:42 Gary: The GitHub version is the CG license, but the w3.org one is the Document License. 16:48:49 Nigel: Possibly there's a problem there! 16:48:59 Gary: I think that might be fine. 16:49:22 Nigel: I'm not clear on that. My expectation was it would be the Software and Document license, from my recollection. 16:49:36 Gary: What does it take to change it? Do we just have to say that we're changing it? 16:49:42 Nigel: Yes, we just make a resolution to do it. 16:50:00 .. We currently have a line in the Charter that says we can choose, per spec. 16:50:18 Andreas: I'm not sure of the views of others, but in general I think the more permissive the better. 16:50:27 .. I don't see a risk with the more permissive license. 16:50:41 .. I think the Software and Document License is close to the other licenses that are used, so 16:50:48 .. it is easier to understand the rules from the start. 16:51:01 Gary: Yes, it looks a lot like the MIT X11 license that is popular on the web. 16:52:02 Nigel: The question for us now is do we change the Charter to specify the Software and Document License, 16:52:14 .. or keep it as is where we can choose on a per-specification basis. 16:52:30 Andreas: Question about what Atsushi said - you said there have been discussions in W3M, 16:52:43 .. and that W3M thinks it is better to switch? Is it one person requesting that or more general? 16:53:03 Atsushi: I believe the status is that some AC Reps are raising issues [when the Document license is used] 16:53:17 .. so we may have some formal objections to the Charter if we continue with the current text. 16:53:31 I agree with Andreas, the more permissive, the better, in my humble opinion 16:53:40 Nigel: Even though it does allow us to use either? It's untested. 16:54:00 Atsushi: The discussion is to limit only to the Software and Document License, so if we point to other choices 16:54:07 .. then I suppose we may be subject to a Formal Objection. 16:55:05 Nigel: I think we have raised awareness today, and I would like to know more about 16:55:12 .. what the Formal Objectors think about our current wording. 16:55:24 Atsushi: Of course we can wait to hear what people will say. 16:56:12 Nigel: Going back to the top, is the W3M approval conditional on the changes, 16:56:26 .. or are these changes just ones they would prefer to see? 16:56:33 Atsushi: Closer to "conditional" I would think. 16:57:37 .. Of course there is also the point that we are asking about CR exit criteria, 16:57:41 .. which may also come up in AC review. 16:58:07 Nigel: So W3M decided to leave that wording as is, after our discussion last call. 16:58:19 Atsushi: Yes, W3M decided it can go ahead. 16:58:27 s/call./call? 16:58:42 .. But maybe there is a risk of some potential conversation in AC review. 16:59:02 Nigel: I think we welcome that because we know we're doing something different. 16:59:41 Nigel: Any last points? We're nearly at time for today. 17:00:01 Atsushi: I will prepare the repositories and raise a PR for adding to the draft charter, and then 17:00:19 .. prepare the AC review templates, possibly next Monday. 17:00:24 Nigel: OK, thank you. 17:00:38 Topic: Meeting close 17:00:50 Nigel: We're at time, let's adjourn for today. Thank you everyone. [adjourns meeting] 17:01:02 rrsagent, make minutes 17:01:02 I have made the request to generate https://www.w3.org/2022/02/17-tt-minutes.html nigel 17:09:09 s|s/will be late, still in previous meeting//|| 17:13:28 s/requiremenst/requirements 17:23:05 rrsagent, make minutes 17:23:05 I have made the request to generate https://www.w3.org/2022/02/17-tt-minutes.html nigel 17:29:55 scribeOptions: -final -noEmbedDiagnostics 17:29:59 rrsagent, make minutes 17:29:59 I have made the request to generate https://www.w3.org/2022/02/17-tt-minutes.html nigel 17:41:14 zakim, end meeting 17:41:15 As of this point the attendees have been Andreas, Gary, Nigel, Xabier, Cyril, Atsushi 17:41:15 RRSAgent, please draft minutes v2 17:41:15 I have made the request to generate https://www.w3.org/2022/02/17-tt-minutes.html Zakim 17:41:20 I am happy to have been of service, nigel; please remember to excuse RRSAgent. Goodbye 17:41:22 rrsagent, please excuse us 17:41:22 I see no action items