- From: Silvia Pfeiffer <silviapfeiffer1@gmail.com>
- Date: Tue, 25 Jun 2013 08:35:49 +1000
- To: Simon Pieters <simonp@opera.com>
- Cc: public-texttracks@w3.org
On Mon, Jun 24, 2013 at 4:38 PM, Simon Pieters <simonp@opera.com> wrote: > On 6/23/13 11:21 AM, Silvia Pfeiffer wrote: >> >> Hi all, >> >> I'd like to focus the discussion on the chapter issue for the moment. >> >> I'm really wondering how to write the "cues in isolation" section: >> http://dev.w3.org/html5/webvtt/#cues-in-isolation . >> >> Since the HTML spec refers to it for rendering chapters: >> "Note: For WebVTT, the rules for rendering the cue in isolation are >> the rules for interpreting WebVTT cue text. [WEBVTT]" >> it's bit of a dangling pointer that I want to close. >> >> I've wondered about these approaches: >> >>> The chapters could be represented through dots >>> on the timeline and a mouseover could render the thumbnails. Or the >>> chapters could be represented in a menu on the controls with the >>> thumbnail as an icon. >> >> There are more examples at >> >> http://wiki.whatwg.org/wiki/Use_cases_for_API-level_access_to_timed_tracks#Chapter_Markers >> >> So, chapters could be in a special menu dropping down from the top of >> the video or a list that is rendered off the video viewport, but >> navigates the video. >> >> Should the section say anything about rendering or just say that >> browser should somehow expose chapters in the video controls? >> What do people prefer? > > I think chapter selection falls into the user interface bucket where we > generally have no spec requirements whatsoever (having UI requirements in > specs either leads to the spec being ignored or implementations sucking). > >> Another related issues is that about thumbnails. You will have noticed >> that many chapter rendering approaches have an image thumbnail in >> them. >> >> I can think of two ways to add an image thumbnail to chapter cues: >> >> 1. as markup inside the cue text >> e.g. >> <img data-url-encoded-image> >> >> 2. as a cue setting >> e.g. >> thumb:data-url-encoded-image > > 3. The UA seeks to the start time of each chapter and uses that frame from > the video as the chapter thumbnail. (This might not have good enough > results, I don't know.) While it's a valid suggestion, it's not really feasible because of the performance hit. Chapters are displayed before the video starts playback. So, you'd have to go seeking to all the start times of the chapters, collect the images and display them. In the case of an autoplay video, this introduces extra delays. Experience with such a delay with theora files just for determining file length in Firefox show that the introduced delays are not acceptable to users. > 4. We leave this use case to custom controls to solve (for now). Right, that's why I asked about intention to implement. >> The first approach has the side effect that we can then also represent >> images in caption or subtitle cues. >> >> The second approach has the side effect that we can then also >> represent caption or subtitle cues with an image thumbnail. > > Are there use cases for these? Some subtitles have logos in them, so in-cue images are not such a bad idea (as long as it's not urls that need to be preloaded). I've never heard of a need for thumbnails for individual caption/subtitle cues, though it could help when seeking just like the YouTube thumbnails on the transport bar help seeking. Cheers, Silvia.
Received on Monday, 24 June 2013 22:36:37 UTC