Re: Filtering of enumerateDevices() results

Den 20. mai 2015 18:00, skrev Joe Berkovitz:
> 
>     > 1. For audio devices, channel count and a binaural (essentially,
>     > speaker-vs-headphone) flag are important attributes.
> 
>     Query - are these distinct attributes or points on the same continuum?
> 
>     ie will a monaural device always have channel count 1? Or is the
>     "binaural" flag specifically for the headphone case, with "each ear of
>     the listener hears one and only one channel"?
> 
> 
> These are distinct attributes.  "Binaural" means the device delivers a
> signal to each ear of the listener separately, as in headphones.
> 
> Although it seems this might be restricted to the 2-channel case,
> perhaps it's better not to legislate that. The species might evolve :-)
> 
>     > 2. How would the WG feel about including more "filtered" information in
>     > the MediaDeviceInfos returned by enumerateDevices(), other than the
>     > device label -- information that the application can use to restrict or
>     > augment the list of devices displayed to a user (since constraints are
>     > not accepted by enumerateDevices())? Ideally channel count, binaural,
>     > sample rate and other attributes such as width and height could be
>     > exposed here.
> 
>     Query - do you think of these things as singular attributes or as a
>     range of possible values? IE some devices are capable of being
>     configured into multiple sample rates - what info would you want?
> 
> 
> Hmmm, I hadn't considered. If the information is returned in each
> MediaDeviceInfo object then perhaps a range would be necessary. In which
> case the approach of submitting constraints to enumerateDevices() (to be
> ignored if filtering is in effect) might be superior.
> 
> 
>     Also - what do you think of when you say "width" and "height" here? Are
>     you thinking of video sources, or is there something new here?
> 
> 
> I was thinking of video sources.
>  
> 
> 
>     If we don't care about fingerprinting, exposing the result of calling
>     getCapabilities() on a device might be OK. Or not....
> 
> 
> I think that may not  work for this purpose because one would have to
> call getUserMedia() for every device in the enumerated list... with lots
> of permission grant interactions with the user.

I was thinking in terms of returning in the result of enumerateDevices()
data structures for each device containing what would have been the
result if you called getCapabilities() on a track connected to that
device. Fingerprints can't get much better than that.


> 
> .            .       .    .  . ...Joe

Received on Wednesday, 20 May 2015 21:12:15 UTC