aom
aom copied to clipboard
Assistive tech : screenreader speech events?
I am very excited by this initiative. A lot of good work has gone in here, and I hope it makes its way into a full blown standard.
When I saw the reference to Asssitive Tech events, the first thing I hunted for, but could not find, was a set of events which indicates when a screenreader starts and stops 'reading'.
The products that we develop feature a lot of pre-recorded audio (mostly speech), and we have no reliable solution for preventing that audio from 'colliding' with (i.e. playing on top of) the screenreader voice. Getting events to indicate when a screenreader is actually reading would be a great help.
I am thinking (at the very least) something similar to the WebAudio API's onended
, but it would be useful to have a corresponding event for when the screenreader starts 'talking'.
Has this been considered and rejected? If so I am very curious to read any discussions on that topic, and/or to any discussions about how to handle such cases today.
This seems like it would be useful to authors, but also a privacy concern as in https://github.com/WICG/aom/issues/116#issuecomment-387887203.
The privacy arguments are compelling, but the use case remains.
Perhaps a more general use case could help here - if a sound started or stopped playing from anywhere except the current document (could be totally unrelated to AT - audio from another tab, or from another application, an incoming call etc.), this could generate an event. I can see various non-AT related uses for this.
A request for implementing such an event would need to be made elsewhere, but where?