Re: Why does JSEP exist yet?

On 06/21/2016 10:30 AM, Iñaki Baz Castillo wrote:
> To summarize: If I want to know about the pc.setLocalDescription()
> method, should I check the JSEP spec [1] or the WebRTC W3C spec [2]?
>
> [1] https://tools.ietf.org/html/draft-ietf-rtcweb-jsep-14#section-4.1.5
> [2] https://www.w3.org/TR/webrtc/#dom-rtcpeerconnection-setlocaldescription
>
>
> Why is JSEP draft still needed at this stage?
> And why should a IETF draft describe&document a W3C (JavaScript) API
> rather than the corresponding W3C spec itself?
>
> I don't think that such a duplicity of similar and related information
> across documents belonging to different WG's is good for newcomers.
>
>
> The W3C spec mentions JSEP in many places, for example:
>
>> 4.3 RTCPeerConnection Interface
>> The general operation of the RTCPeerConnection is described in [JSEP].
> Why? why doesn't the W3C spec describe the general operation of the
> API that it defines? why does it refer to another (IETF) document?

The two documents are edited by different editors with different skillsets.

This is more of a pragmatic division of labor than a
purity-of-architecture question.

>
>
> We may take a look to the WebSocket spec(s). They are split into a
> IETF spec defining the WebSocket wire protocol [3] and a W3C spec [4]
> defining an API for browsers to create and manage WebSocket instances.
> Period. They don't not collide, and the W3C spec does not reference
> the IETF one when it comes to describe a *JavaScript* method.
>
> [3] https://tools.ietf.org/html/rfc6455
> [4] https://www.w3.org/TR/websockets/
>
>
> So, if there is a good rationale for keeping both WebRTC related specs
> (other than the fact that both were born several years ago) I think
> they should target different audiences. Said that, it's not clear to
> me who should read JSEP and who should read the W3C spec.

All implementors should read both.

Received on Thursday, 23 June 2016 11:40:19 UTC