At 14:40 +1300 11/12/08, Robert O'Callahan wrote:
On Wed, Dec 10, 2008 at 5:56 PM, Dave Singer
<<mailto:[EMAIL PROTECTED]>[EMAIL PROTECTED]> wrote:
At 21:33 +1300 9/12/08, Robert O'Callahan wrote:
For what it's worth, loading an intermediate document of some new
type which references other streams to be loaded adds a lot of
complexity to the browser implementation. It creates new states that
the decoder can be in, and introduces new failure modes. It creates
new timing issues and possibly new security issues.
I'm not sure I agree; but if you believe that, we should address it
no matter which way this discussion goes. It should absolutely be
possible to reference a SMIL file, or an MP4 or MOV file with
external data (to give only two examples) from a <video> or <audio>
element, and have the DOM, events, states, and APis work correctly.
I agree it should be done eventually, it's just significantly more
complicated than what we have to deal with currently.
But if the state machine or other aspects are actually wrong for this
case, then we should fix it now. We have, for example, tried to keep
out these kinds of assumptions:
a) all media is downloaded (no, it might be streamed or even arriving
over non-IP, e.g. a TV broadcast)
b) all delivery methods are self-contained (no, they might reference
resources as well as contain them)
c) all delivery is sequential in play order (no, some file formats
decouple data timing and data ordering)
Rob
--
"He was pierced for our transgressions, he was crushed for our
iniquities; the punishment that brought us peace was upon him, and
by his wounds we are healed. We all, like sheep, have gone astray,
each of us has turned to his own way; and the LORD has laid on him
the iniquity of us all." [Isaiah 53:5-6]
--
David Singer
Multimedia Standards, Apple Inc.
