Re: I-D ACTION:draft-zimmermann-avt-zrtp-03.txt

[Colin Perkins <csp@xxxxxxxxxxxxx>]

On 9 Mar 2007, at 18:37, Alan Johnston wrote:
> Richard Barnes wrote:
...
> > > 3. Best Effort SRTP.  Only ZRTP supports best effort SRTP  
> > > natively.  DTLS-SRTP relies on the still being developed MMUSIC   
> > > SDP extensions for this to work.  ZRTP can be, and is being,  
> > > deployed today as a result.
> >
> > I think this is another instance where there's not a technical  
> > barrier, it's just something that hasn't been done.  All the ZRTP  
> > discovery mechanism entails is sending a ZRTP message to an  
> > endpoint with unknown support.  This doesn't seem incompatible  
> > with DTLS -- just do the same thing a ClientHello message.
> Perhaps.  In order to do this you must know that you will do no  
> harm, i.e. not cause UAs to crash and burn.  We have carefully  
> developed this approach in ZRTP and tested it at SIPits to ensure  
> that it works.

Do you have statistics on the number of implementations which cope  
with packets containing an RTP header extension, yet fail the  
simplest packet validity check?

Colin