A view on the current webRTC API and next step...

Hi,

Based on our experience on developing JavaScript web apps using WebRTC and implementing WebRTC and in light of web developers feedback, we propose to start drafting on a "2.0" API already now, including a review of the use of SDP as well as O/A. The ideas of exposing lower layers to the web app, leaving much signaling to the JavaScript web app/JSL has much appeal (even though one must not forget security and browser interoperability challenges). These and a few, other selected enhancements should be explored carefully and with expedience by this WG and not later.

Having said that, leaving more to the JavaScript should still be done in a way that the most basic browser-to-browser use cases should be easy to do without dependency to JSL's.

One reason for start drafting the 2.0 is to seek a rough consensus in the community- especially including web developers views- on the design of the "2.0 API". Such a "straw man" can be used to safeguard that the current WebRTC API ("1.0") can indeed be done as a JSL on top of it and make sure the 2.0 is not limited by design of 1.0. Furthermore, 1.0 should be designed with an idea of 2.0 in mind, making a transition as smooth as possible.

We expect there are several 2.0-ish implementations around or being developed with a lower layer API with the WebRTC-like API implemented as a JSL, meaning it is reasonable to expect that the community can leverage on experience from running code in developing 2.0.

Regards
Göran Eriksson for the Ericsson webRTC team



Received on Tuesday, 23 July 2013 11:40:10 UTC