[rtcweb] Need for Default signaling protocol for RTCWeb [was RE: draft-ibc-rtcweb-sip-vs-websocket]

Hi Tim,

I agree with you that it is possible to implement using plugins and it is highlighted in RTCWeb charter (http://tools.ietf.org/wg/rtcweb/charters) as "These are not interoperable, as they require non-standard extensions or plugins to  work.". And also, the aim of this WG is to "There is a desire to standardize the basis for such communication so that interoperable communication can be established between any compatible browsers". Now, I don't the reason for implementing signaling in JavaScript or plugin or applet.

Also, I don't know about native plugin but I guess that those plugin are installed as part of the browser installation. Please correct me if I'm wrong.

Thanks
Partha

>-----Original Message-----
>From: Tim Panton [mailto:tim@phonefromhere.com]
>Sent: Thursday, September 15, 2011 3:19 PM
>To: Ravindran Parthasarathi
>Cc: Iñaki Baz Castillo; rtcweb@ietf.org
>Subject: Re: [rtcweb] draft-ibc-rtcweb-sip-vs-websocket
>
>It is worth pointing out that even with a signalling  protocol agnostic
>webRTC you
>wouldn't _have_ to implement your signalling protocol in javascript.
>
>Browsers support other plugin languages - specifically it would be
>possible to implement
>a classic full UDP SIP stack as a java applet and delegate the media
>handling to the webRTC layer.
>
>Chrome is moving towards support for native plugins, but it isn't clear
>that they
>would be suitable for this use.
>
>Tim. (speaking for himself)
>
>
>On 14 Sep 2011, at 11:03, Ravindran Parthasarathi wrote:
>
>> Hi Inaki,
>>
>> <snip>
>>
>> The fact that there are other alternatives for signaling in the web
>does not mean that using SIP is invalid.
>> If I want to build a SIP phone in a web, why should I use libjingle
>rather than SIP protocol? Why should I code a complex server behaving as
>a gateway between Jingle and SIP protocols?
>>
>> Any protocol conversion (i.e. from Jingle to SIP) means loss of
>features. Our draft proposes the contrary: no protocol conversion (just
>SIP), and just transport protocol conversion (as already exists in SIP
>when bridging UDP/TCP/TLS-TCP/SCTP...).
>> </snip>
>>
>> I agree with your problem statement. I have raised the same concern in
>http://www.ietf.org/mail-archive/web/rtcweb/current/msg00845.html. IMO,
>your solution is a workaround and we will end-up with your solution in
>case signaling protocol is not standardized as part of RTCWeb.
>>
>> Thanks
>> Partha
>>
>> _______________________________________________
>> rtcweb mailing list
>> rtcweb@ietf.org
>> https://www.ietf.org/mailman/listinfo/rtcweb