Re: A new proposal for how to deal with text track cues

On Wed, Jun 12, 2013 at 7:44 PM, Glenn Adams <glenn@skynav.com> wrote:
>
> What is clear to me is that the current design was strongly influenced by
> the (arguably) premature conclusion that WebVTT could or should be developed
> and promoted as a singular timed text content format. Such decisions were
> clearly taken without general agreement among timed text authors and the
> existing commercial content community, and without agreement within the W3C
> itself. So saying that "we" made this decision and so forth is rather
> presumptuous.

The design of the Text Track API and WebVTT had people involved from
all corners of the caption universe, so this is the outcome of work of
multiple people. If you don't feel you should be included in that
"we", you are free to interpret it as "they".


> Secondly, it is most certainly true that the current design of the WebVTTCue
> interface has been done without taking into account pre-existing and
> deployed timed text formats. While proposing a new, more general purpose cue
> interface may aid in interoperation with different formats, such a new
> design would need to be created in close cooperation and agreement with the
> Timed Text WG and other industry participants (e.g., from EBU, SMPTE, etc).
> The HTML WG is not a good forum for discussing or accomplishing this work.

I am grateful that you have forwarded the proposal to the TTWG. I was
simply waiting to hear feedback from other fora first, which include
the HTML WG, the WHATWG and the TTCG. Seeing as this design currently
influences the WebVTT spec and the HTML spec, I found these three fora
had to be my first point of call.


> Third, I have to wonder about your parenthetical "or in fact (gasp!) through
> a TTML file"? Here, you clearly represent a partisan position in writing
> this, and denigrate an accepted W3C REC. Are you trying to perpetuate
> controversy and conflict or are you trying to work with the TTML community?

I apologize if you take offence at that. It was written to point out
that both TTML and WebVTT are essentially serialisation formats for
cues and that they should each be able to deal with each other's cue
formats. This was not meant to create more controversy, but instead
bridge that controversy.

Might I ask you to take your time to read the proposal properly and
provide me with technical feedback. I would highly appreciated it.

Thanks,
Silvia.

Received on Wednesday, 12 June 2013 12:42:26 UTC