Re: Should <video> buffer control be tri-state?

On Sat, Jan 9, 2010 at 11:57 PM, Robert O'Callahan <robert@ocallahan.org> wrote:
> I think it would be reasonable to offer preload="no", preload="metadata",
> preload="playthrough" and preload="all" with the understanding that due to
> resource or policy limits any value (other than "no") is only a hint and the
> user agent may not be able to honor it.

This seems like the best proposal so far, to me.  The name "preload"
is clear enough that I'm not *too* worried about authors misusing it,
at least if they look at the HTML they're using (although I'm sure
some still will).  I would really like the spec to say that in the
absence of a preload attribute, UAs *should* try to guess what videos
users will play, and behave somewhere between preload="metadata" and
preload="playthrough" depending on how likely they think it is.  As an
author, it would be nice not to have to specify this explicitly in
most cases -- one less thing to worry about.

Received on Sunday, 10 January 2010 13:05:36 UTC