Re: Roll-up captions in WebVTT

On Mon, Feb 27, 2012 at 1:36 PM, David Singer <singer@apple.com> wrote:

> > I don't find explicitly grouping cues together objectionable.  What I
> don't like is the idea of markup that says "these cues should be rendered
> as roll-up captions".
>
> we might be all in agreement here.  I want the VTT language to tell me
> what text is on screen, and where it goes, and something else (I suggest
> CSS transitions) to handle smoothness of motion


It's not immediately obvious to me how transitioning the position would
interact with the "Adjust the positions of boxes..." positioning algorithm.

It may be useful for use cases like scrolling credits and captions that
follow signs across the screen (though I wouldn't prioritize it until
implementations catch up somewhat).

I think it would be a bad idea to recommend using transitions to implement
roll-up captions by hand, though, if you're suggesting that.  That seems
brittle, since it would depend on the exact positioning chosen by the
positioning and layout algorithms (to know when and how far things have to
move), and it would make it hard for users to toggle roll-ups off if they
don't want them.  I think if enough users really want roll-ups, UAs should
simply be allowed to switch rendering modes as a user preference, with no
change to markup.  (I'm still doubtful that people really want that,
though, and I don't know if there would be any unwanted side-effects from
permitting UAs that level of rendering freedom.)

-- 
Glenn Maynard