Re: [rtcweb] ALPN question - other labels?

[Martin Thomson <[e-mail]>]

On 8 August 2014 09:12, Paul Kyzivat <[e-mail]> wrote:
> Are you saying that the offering and acceptance of ALPN values is controlled
> totally by the browser, without any influence by the JS app?

Well, for the confidential media use case, it has to be.

>>  > And, for interop reasons, will there be a way to get a browser to
>> offer another label and still support the JSEP APIs if the other end
>> accepts it?
>>
>> That is an interesting question. You would need to teach the browser
>> that other protocol, I guess.
>
> IMO this would only be interesting if other values can also be used in the
> server role.

Do you mean the server can pick something that the client didn't offer?

> For now I think it is more important that you be very clear what you intend.
>
> I think it means that CLUE implementations will always need to use "webrtc"
> as the ALPN for the CLUE data channel, so that it will be right when
> interoperating with WebRTC. That is possible, though it seems weird when not
> interoperating.

If the intent is to interoperate with WebRTC, by effectively sitting
on top of the WebRTC stack, then I don't find that to be particularly
strange.  That doesn't mean that you can't define a "clue" token,
which may or may not be identical or at least very similar to
"webrtc".