- From: Adam Bergkvist <adam.bergkvist@ericsson.com>
- Date: Tue, 17 Apr 2012 10:26:20 +0200
- To: William Lin <chaer.lin@gmail.com>
- CC: "public-media-capture@w3.org" <public-media-capture@w3.org>
On 2012-04-17 02:47, William Lin wrote: > > On Fri, Apr 13, 2012 at 9:47 PM, Harald Alvestrand <harald@alvestrand.no > <mailto:harald@alvestrand.no>> wrote: > > __ > On 04/12/2012 04:53 AM, William Lin wrote: >> Hi, >> >> in these days, Constraints and Capabilities API for getUserMedia >> were discussed while i found one scenario doesn't be covered: >> "users may want to initiate capture from multiple devices at one >> time in multiple media streams" (refer to >> http://www.w3.org/TR/capture-scenarios/#multiple-active-devices). >> is it possible adding the options to set the number of active >> devices? and the callback function needs to change: > Given that the normal case will be one device, what's the cost of > this relative to "just" calling getUserMedia multiple times? > > > Seems it is one mean to get 'all' devices. but suppose getUserMedia > will be applied to 'monitor system', it is more sensible just calling > one time. I think we need to support this in some form. The home-security monitoring use case may not be the most common, but consider a mobile device with a front and a back camera, it's likely that people would like to toggle between the cameras. In that particular use case I see having one MediaStream with two video tracks as the most natural alternative. > And how do you see the user interaction dialog looking? > > > in my opinion, it is depended on UA implementation: a list/checkgroup is > ok (If native platform application can support user interaction in the > multiple devices case, web runtime application will do). If getUserMedia() is supposed to return several MediaStream objects then the UI will be responsible for grouping the different devices into MediaStreams which will add extra complexity to the UI. Therefore I'm more in favor of getUserMedia() providing one MediaStream object with several tracks (which we currently support). It would then be up to the application to create new MediaStream objects from the tracks of the LocalMediaStream if needed. This doesn't solve all of Williams problems, we still have the question about the app requesting, e.g., a certain number of video capture devices. /Adam
Received on Tuesday, 17 April 2012 08:26:49 UTC