Closed guidou closed 7 years ago
@alvestrand Any objection to removing this from the spec? I think the numerous issues make it clear that it is a bad idea to keep this as currently written in this spec.
Tagging @juberti @jan-ivar @aboba - I'm personally in favor of removing them, but since they're implemented, we might have customers depending on them. (I know they're in Chrome; not sure about the others)
Note that the predefined IDs are not implemented in Chrome. At least not the in the way currently described in the spec. Chrome exposes some abstract devices whose IDs are "default" and (on Windows) "communications", which represent the corresponding abstractions provided by the OS. These devices are shown by enumerateDevices() and can be used by getUserMedia(), setSinkId() and any other API that wants to deal with device IDs. The way I understand the spec is that special IDs "id-multimedia" and "id-communications" are specific to setSinkId only and are not exposed by enumerateDevices.
This addresses, or renders invalid the following issues: #18, #19, #20, #47 and #51.