[whatwg] Serving up Theora <video> in the real world

[Gregory Maxwell]

On Thu, Jul 9, 2009 at 10:53 PM, Robert O'Callahan<robert at ocallahan.org> wrote:
>> It works. Except where it doesn't. It's the "where it doesn't" that
>> counts. At the moment Safari has issues. Out of two widely used
>> production browsers with HTML5 support, one is broken. Not good odds,
>> but I'm hopeful for the future.
>
> It's unfortunate that Safari is broken out of the gate, but realistically
> hardly anyone is going to be using XiphQT right now. Hopefully Apple will
> have fixed the bug before too long.


The killer there is that it means you can't check if the user is using
Safari then ask them to install XiphQT, since the test will never
pass.

Wikimedia has been promoting XiphQT for a long time since an
object-tag with XiphQT installed was the best performing pre-videotag
solution for the relevant platforms. This also brings up another
problem with using canPlayType and browsers that use external codecs:
Most users with XiphQT have older versions which have problems in the
video tag, and there doesn't appear to be any obvious way to detect
this.


Perhaps there needs to be a way for a canPlayType like call that gives
you more information about the software used to decode the named type?
  Otherwise I foresee continued problems for every browser that hasn't
followed Mozilla's obviously wise lead of fully embedded codecs.