RE: text track associations

Eric >> "Why? Like video, captions are a type of visual media." ... " It never even crossed my mind that a file with only audio samples shouldn't work in a <video> element"

My mistake - my testing was incorrect, it does seem that audio files are implemented across the board.  What I meant was that authors might not, as I did not, immediately grasp that to deliver an audio file with display captions requires using a video element; as captions aren't normally considered video and neither is audio. But I note that logical conclusion of your two statements above is that considering captions as timed visual media means one might expect them to work as a source on a video element, is that what you intended?

Eric >> " If we require the src to a <video> element to contain visual media, what should happen with a file that can change characteristics while it is playing, eg. an adaptive stream that switches to audio-only when the user's bandwidth changes?"

Nobody said anything about requiring video samples, I'm just concerned that the spec reflect reality.

Eric>> " What is the difference between a "video" and a "movie"? How about "A video element is used for playing video or audio files."

Works for me, I don't know why Ian included the term movie, unless he intended it to mean an audio visual resource, where video might be construed as having no audio.

Eric>> " I guess this could go either way, but I think it is more logical for an <audio> element to render only audio samples "

If the only difference between the video and audio element is that one can have a display rectangle and the other cannot, why don't we just say that? Better still get rid of the element and just make display-ability a toggle attribute on <video>.

Anyway, I'm OK with the solution suggested; provided the spec is clear that is allowed, and how its intended to be done.

Received on Wednesday, 13 April 2011 15:39:24 UTC