There is a problem confronting applications desiring to use a variety of APIs such as pointerlock, fullscreen, WebRTC, local storage and so on.
The problem is that each instance of attempting to use such an API leads to a new "Allow ..." popup the user has to click away. By some discussion on the www-dom ML the same misery is about to befall things like translating a Dom Event-Level 4 key code to a users locale. I would propose to centralize this and make it an up-front dialog remembered for a site such that: - The application can trigger the dialog before attempting to use any of the APIs - It can get the results back what the user allowed the page to do - For positive results the page will not have to double-check each APIs availability/success/failure individually but can assume it's enabled. - Unless the preference of the page changes in relation to the stored preference, the dialog (or any popup) for the desired functionality will not be pestering the user henceforth.
