- From: Anssi Kostiainen <anssi.kostiainen@nokia.com>
- Date: Tue, 29 Jan 2013 15:33:10 +0200
- To: ext Anne van Kesteren <annevk@annevk.nl>
- Cc: Doug Turner <dougt@mozilla.com>, Frederick Hirsch <Frederick.Hirsch@nokia.com>, "public-device-apis public-device-apis@w3.org" <public-device-apis@w3.org>
On 28.1.2013, at 18.39, ext Anne van Kesteren wrote: > On Mon, Jan 28, 2013 at 2:08 PM, Anssi Kostiainen > <anssi.kostiainen@nokia.com> wrote: >> Anne - the side effects concerns aside, what is your preference? > > Well, as far as I can tell you need to dispatch in each Window that > has a listener. Since the event model is such that you cannot know > there's a listener, you'll have to dispatch in all of them. If this is the intended behavior (it seems so, but pending confirmation from Doug), I think we can keep the current concise language: [[ When the current XXX state changes, the user agent must queue a task to fire a XXX event at the Window object. ]] Or would "at all the Window objects" be better? Sounds a bit strange though, as plurality is implied. On a related note, here's an idea how this could be altered to improve privacy: Dispatch these events only at the top-level browsing context's Window object, and if loosened a bit from there, at any nested browsing context's Window object whose Document object has the same origin as the top-level browsing context's Document. This would allow these events to be dispatched only within iframes and frames in a frameset that have the same origin as the top-level browsing context's Document. Lose some use cases, gain some privacy protection. WDYT? -Anssi
Received on Tuesday, 29 January 2013 13:34:27 UTC