Re: rationale for getUserMedia on MediaDevices

On 2014-06-25 18:45, Shijun Sun wrote:
> Resend to correct a typo ---
>
> In the latest version of the draft, there are two ways to call
> getUserMedia() within the NavigatorUserMedia.
>
> #1:	NavigatorUserMedia.getUserMedia();
> #2:	NavigatorUserMedia.MediaDevices.getUserMedia();
>
> Could someone share the rationale why we need the 2nd option?  I
> couldn't find any spec bug that led to the change.  It seems it also
> kind of limits future extension of getUserMedia () beyond the media
> devices.
>

Hi

The discussion about the MediaDevices object started around here [1]. We 
needed an object to be the EventTarget for the new event that tells the 
script that a new media device is connected, and we didn't want that 
object to be Navigator. Once we had an object that collected the 
functions related to local media devices it seemed to be a pretty 
natural place for getUserMedia() as well.

There wasn't a lot of people involved in this discussion so more input 
is welcome.

Regarding future extensions, the function is called getUser_Media_() so 
perhaps it shouldn't be extended beyond media devices anyhow. 
Additionally, getUserMedia() is pretty tightly coupled with Constraints 
and that might be unnecessary complexity for other types of devices.

[1] 
http://lists.w3.org/Archives/Public/public-media-capture/2014May/0243.html

Received on Friday, 27 June 2014 08:40:16 UTC