On Sat, 29 Oct 2011, Robert O'Callahan wrote:
On Wed, Oct 19, 2011 at 11:57 PM, James Graham <[email protected]> wrote:
On 10/19/2011 06:40 AM, Anne van Kesteren wrote:
Is that an acceptable limitation? Alternatively we could postpone the
nested fullscreen scenario for now (i.e. make requestFullscreen fail if
already fullscreen).
I think punting on this makes sense. Pages can detect the failure and do
something sane (make the element take the whole viewport size). If the
feature becomes necessary we can add it in v2.
I don't think punting on nested fullscreen is a good idea. It's not some
edge case that most applications can't hit. For example, it will come up
with any content that can go full-screen and can contain an embedded
Youtube video. (It'll come up even more often if browser fullscreen UI is
integrated with DOM fullscreen, which we definitely plan to do in Firefox.)
If we don't support nested fullscreen well, then the user experience will
be either
-- making the video fullscreen while the containing content is already
fullscreen simply doesn't work, or
-- the video can go fullscreen, but when you exit fullscreen on the video,
the containing content also loses fullscreen
Both of these are clearly broken IMHO.
Presumably the embeded video could detect that it was already in a
fullscreen environment and deal with it accordingly. So in theory we could
wait and see if people just do that before deciding that we have to
implement the more complex thing. But that might be unnecessarily
difficult and easy to get wrong. So maybe we should just deal with this
now.
