Re: [vibration] How to handle iframes?

Intuitively, I would expect the vibration api to work like playing audio,
which is additive. So as long as any pattern in any frame says the device
must vibrate, it vibrates. I haven't fully gone through the implementation
implications, but it seems doable to merge multiple patterns into one.

We would have to change the spec to say that navigator.vibrate(0) only
cancels the pattern in that frame. Patterns played from other frames are
unaffected.

I've gone over the Geolocation API spec a number of times recently, and I
don't remember it saying anything about restricting to the top level frame.

What do you think?

Michael




On Wed, Sep 4, 2013 at 4:35 PM, Kostiainen, Anssi <
anssi.kostiainen@intel.com> wrote:

> On Sep 4, 2013, at 6:18 PM, Michael van Ouwerkerk <
> mvanouwerkerk@google.com>
>  wrote:
>
> > Is there any precedent for restricting to top level browsing context? It
> would make the API less useful, so I don't find it an attractive solution.
>
> W3C's Privacy Interest Group did a privacy review of the Proximity Events
> and Ambient Light Events specs and suggested we limit the events to the
> top-level browsing context.
>
> I also recall there was some discussion on improving the privacy of
> Geolocation API similarly, but I'm not sure if that ever landed to the spec.
>
> What would be your preferred solution re iframes?
>
> -Anssi