15:00:36 RRSAgent has joined #tt 15:00:36 logging to http://www.w3.org/2016/11/24-tt-irc 15:00:38 RRSAgent, make logs public 15:00:38 Zakim has joined #tt 15:00:40 Zakim, this will be TTML 15:00:40 ok, trackbot 15:00:41 Meeting: Timed Text Working Group Teleconference 15:00:41 Date: 24 November 2016 15:03:20 Present: Nigel, Pierre 15:03:47 Regrets: Glenn, Philippe, Andreas, Thierry 15:03:51 scribe: nigel 15:05:27 Present+ Rohit 15:11:05 rpuri has joined #tt 15:13:29 pierre: [describes implementation experience with imscjs] 15:17:44 -> https://github.com/sandflow/imscJS 15:20:02 Topic: This meeting 15:20:15 nigel: There are just 3 of us, so I would propose to describe and minute last night's 15:20:35 ... TAG meeting where they reviewed TTML2, also look at F2F agenda items, and 15:20:39 ... cover the liasions. 15:22:08 s/liasons/liaisons 15:22:25 Topic: F2F agenda 15:22:42 nigel: Agenda items for the F2F? 15:24:37 nigel: My goal for the meeting is to review the ED of TTML2 as it is at that time and try 15:24:59 ... to get consensus on any remaining issues so that we can publish a "last call" style 15:25:18 ... WD for wide review, i.e. to be the basis of the Candidate Recommendation. 15:26:02 nigel: I would also like to review any IMSC 1.1 feedback. 15:26:12 Pierre: We may need 3-4 hours for that. 15:26:27 ... I can make sure we get to that meeting with a good baseline IMSC 1.1 edit that 15:26:39 ... addresses those issues that have been assigned to IMSC 1.1 and proposals to defer 15:26:43 ... some issues to IMSC2 etc. 15:27:17 nigel: As far as I know we will not be covering WebVTT feedback. 15:27:32 nigel: Also we may have an observer from the TAG. 15:27:41 ... I'd like to describe the TAG meeting's TTML2 review briefly. 15:28:36 ... They focused on profiles, and alignment with CSS, mainly. Also if it is intended for 15:28:55 ... browsers and how to support changes in display of HTML/CSS at defined times 15:29:03 ... in a better way than the existing VTTCue implementations. 15:29:19 ... On profiles, TAG isn't sure they're a good idea, mainly because they have a view that 15:29:43 ... we should specify stuff that everyone implements/can implement. 15:30:26 nigel: My view on that is that profiles in TTML are being used and that they form 15:30:36 ... rungs on the ladder, i.e. labelled points of stability that people can rely on. 15:32:54 ... On CSS, there was some good analysis that suggested that we may be making the 15:33:13 ... alignment with HTML and CSS worse than we need to. They would like to help 15:33:40 ... people understand how to translate TTML to HTML/CSS to reuse the rendering engines 15:34:00 ... in user agents. In particular there are two areas of concern: 1. that we have some 15:34:18 ... style attributes with names that sound like CSS properties but are actually different, 15:34:45 ... such as fontVariant. 2. that we could align better the value space for specific style 15:35:22 ... attributes, perhaps by not omitting values that have use, like "auto" in some cases. 15:38:53 ... Finally they did consider the HTMLCue idea and were generally positive that something 15:39:16 ... like that would be a good idea, but not confident about their ability to influence 15:39:19 ... all browser makers. 15:39:51 Rohit: What do you think the TTWG should do? 15:40:14 nigel: In my view we should do one more pass on alignment with CSS to check 15:40:26 ... style attribute names and value spaces, and also to consider what the fallback 15:40:45 ... scenarios are for cases where our semantics are not (yet) supported in CSS. 15:41:08 Rohit: Would it be worth considering other serialisations than XML such as JSON? 15:41:32 Pierre: It's definitely possible to create a 1:1 mapping. 15:41:35 Rohit: It's easy. 15:41:50 Pierre: The question of what is the right mapping would depend on the exact use case. 15:42:03 ... I mean what is the interoperability point for that exchange? Are we looking for a 15:42:44 s/ Are we looking for a/ 15:43:00 ... If we know the use case then we can figure out the approach. 15:43:03 Rohit: Fair enough! 15:43:40 nigel: I'm reminded that Glenn told us that any serialisation that supports namespaces 15:43:53 ... could be used. JSON does not support them out of the box of course, but we could 15:43:56 ... define something. 15:44:22 Pierre: There are some solutions to that, such as specifying a qualifying name in the JSON structure. 15:44:36 ... How we solve that problem depends on the use case and the interoperability point. 15:47:24 nigel: The last point on the TAG review relating to the F2F is that one of the TAG members 15:47:41 ... who is based in London may attend part of our meeting as an observer, which I think 15:47:44 ... would be great. 15:47:54 Topic: IMSC 15:48:07 nigel: I just want to note that the liaison text is done, and unless there are any late 15:48:20 ... changes to the text I sent to the reflector then it's just a matter of me finding time 15:48:33 ... to turn the handle and send them out, which I plan to do very soon (e.g. tomorrow) 15:48:55 ... That should get us the responses we are requesting in time for the F2F. 15:49:07 Topic: Profiles registry 15:49:18 nigel: We are not quorate for this right now, deferring until next week. 15:51:47 Topic: Unicode liaison 15:51:57 nigel: As on the agenda, there's an update to this - they're thinking about it! 15:52:12 Pierre: It's great that they are looking at it. 15:53:53 nigel: I think we're done for the agenda for today, so thank you guys very much for joining on thanksgiving and enjoy the rest of your day! [Adjourns meeting] 15:54:08 rrsagent, make minutes 15:54:08 I have made the request to generate http://www.w3.org/2016/11/24-tt-minutes.html nigel 15:56:15 Chair: Nigel 15:56:42 s/Topic: F2F agenda/Topic: F2F agenda and TAG review 15:57:01 s|s/liasons/liaisons|| 15:57:10 s/liasions/liaisons 15:58:40 rrsagent, make minutes 15:58:40 I have made the request to generate http://www.w3.org/2016/11/24-tt-minutes.html nigel 15:59:17 ScribeOptions: -final -noEmbedDiagnostics 15:59:22 rrsagent, make minutes 15:59:22 I have made the request to generate http://www.w3.org/2016/11/24-tt-minutes.html nigel 17:42:29 Zakim has left #tt