Re: WebVTT spec

On Tue, Mar 26, 2013 at 6:55 AM, Ian Hickson <ian@hixie.ch> wrote:

> On Sat, 23 Mar 2013, Silvia Pfeiffer wrote:
> > On Tue, Mar 19, 2013 at 9:13 AM, Ian Hickson <ian@hixie.ch> wrote:
> > > On Tue, 19 Mar 2013, Silvia Pfeiffer wrote:
> > > > >
> > > > > I moved them because I'm not sure all text track languages will
> > > > > want to define this. We can always move it back later if it turns
> > > > > out other formats do always want it.
> > > >
> > > > I don't think the format matters. What matters is what the web devs
> > > > need and these are the two core functions a web dev would use IMHO.
> > >
> > > The format clearly matters, since if it's not defined for a language,
> > > there's no way for the developers to use the API...
> >
> > .text provides the uninterpreted content of a cue - every format will
> > have such.
>
> Not necessarily. DVD subtitles wouldn't have one, they'd likely have
> something that returned an ImageBitmap or some such. TTML probably
> wouldn't have one, since their format is structured DOM and might never
> have had a textual form (think script-created DOMs) -- they'd just have a
> way to get a pointer to the relevant node in the actual DOM tree, I'd
> guess. Other formats might have other needs, e.g. EBU might want to just
> expose the Text Field as a raw 112 byte ArrayBuffer; I don't know how
> you'd really expose an unparsed DOMString for EBU.
>

OK, fair enough.


On Sat, 23 Mar 2013, Silvia Pfeiffer wrote:
> > > >
> > > > So how does the browser get informed about the format if the cues?
> > > > Which rules it has to use to interpret then?
> > >
> > > This is already defined in the spec, no? I'm confused about what is
> > > ambiguous here. Why would it be unclear?
> >
> > In the spec where addTextTrack(kind, label, language) is defined [1] it
> > says:
> >
> > "Initially, the text track list of cues is not associated with any rules
> > for updating the text track rendering. When a text track cue is added to
> > it, the text track list of cues has its rules permanently set
> > accordingly."
> >
> > So I go to the definition of TextTrackCue() and it contains no
> > information on the "rules for updating the text track rendering".
>
> Not sure why the constructor (assuming that's what you mean by
> "TextTrackCue()") would be where that would be defined. These objects can
> be created in many ways; most commonly, they're created by the WebVTT
> parser directly, without the constructor.
>

If you're using <track>, yes. I am assuming a JavaScript-only use of
TextTrack() constructor and TextTrackCue(). And yes, you are right: the
format is much better suited to being defined in the TextTrack()
constructor rather than in a TextTrackCue().


>
> The definition of the WebVTT data model says of WebVTT cues:
>
> # The associated rules for updating the text track rendering of WebVTT
> # text track cues are the rules for updating the display of WebVTT text
> # tracks.
>

When you are doing it all in script, there is no WebVTT file and no means
to specify that the cue that you are providing is in the format of a WebVTT
cue.


> So I go and check the addCue() function [2], which says:
> >
> > "If the text track list of cues does not yet have any associated rules
> > for updating the text track rendering, then associate the text track
> > list of cues with the rules for updating the text track rendering
> > appropriate to cue."
> >
> > However, what does "appropriate to cue" mean? How is the browser
> > supposed to distinguish between a cue in a WebVTT format or a cue in a
> > TTML format or anything else?
>
> A WebVTT cue isn't a TTML cue. Why would distinguishing them be difficult?
>


Again: assuming you do it all in script: you create a text track using
TextTrack(), then you add a cue using TextTrackCue() and addCue(). The
browser has no idea what format that cue is in - it could be in WebVTT Cue
format or TTML cue format or anything else for that matter.


> Is it supposed to guess from the tags used in the cue?
>
> Why would it need to guess? I don't understand.
>

Hopefully I've been clearer this time around.


On Sun, 24 Mar 2013, Silvia Pfeiffer wrote:
> >
> > I've copied the spec text to a new HTML file and uploaded it to the W3C
> > Mercurial repository - it's now staged at:
> >
> > https://dvcs.w3.org/hg/text-tracks/raw-file/default/webvtt/webvtt.html
>
> Do you have a diff of this spec to the existing one? I'd like to make sure
> that the diff is zero before we actually hand things over.
>

Are you after a diff on the html level or something else? I just copied all
the text from the whatwg source file that was between <!--START webvtt -->
and <!--END webvtt --> markers, then applied the ReSpec boilerplate.

Silvia.