Re: [rtcweb] Let's define the purpose of WebRTC

[Matthew Kaufman <[e-mail]>]

On 11/7/11 11:58 AM, Roman Shpount wrote:
>
> First of all, this is not about interop. We have a fairly significant 
> experience regarding real-time communications, but for many reasons 
> most of this experience is related to SIP. Based on this, experience 
> we know that there are issues, such as glare, media forking, media 
> negotiation, support of future codecs, that need to be addressed.

Or not.

> If we are asking about scenarios which have not been relevant to your 
> particular "innovative" application, it does not mean that other users 
> of WebRTC will not encounter it. This is all about creating the best 
> possible future proof system we can.

The best possible future-proof system would be one that exposes 
low-level access to the primitives needed to construct anything. That is 
not, unfortunately, what is currently proposed.

>
> Second, the reason I raised my comment about SRTP was exactly in order 
> to support future applications that we are not aware about now. I do 
> not see a why we should mandate something without a good reason. The 
> more control we will give to developer, the more applications it would 
> be possible to build. I would strongly support required to implement, 
> but not required to use model for SRTP. This way SRTP is available to 
> the application developer but not something they must use. There are 
> places where communications are illegal, unless they can be 
> intercepted by the "big brother". Things like Skype are illegal there 
> for this exact reason. I would not argue if this situation is 
> appropriate, legal, or moral, but it exists. It would be better that 
> WebRTC would be able to operate in environments like this.

Do you want *your* browser to be able to switch to plain RTP for the 
call you're making from a place with an unsecured wi-fi network (as 
essentially all are at this point, given the ease of cracking the 
typical schemes used for wireless security these days) ?

I don't. Even though I do agree that there might be external reasons why 
service providers might want the browser to have that ability.

Matthew Kaufman