Re: [rtcweb] Transport -03, bundling question (Re: Comments on draft-ietf-rtcweb-transports-02)

Harald said:
"My memory was faulty; I had thought that the same conclusion had applied
for TCP ICE candidates as for TURN ICE candidates. Version -04 will have
this fixed."

[BA]  My memory was similarly faulty. But assuming the problem is with my
memory as opposed to the minutes, the decision to require support for ICE
TCP brings to mind other questions such as:

a. Is support for RFC 4571 (framing of RTP/RTCP over connection oriented
transport)  also mandated?  With respect to JSEP, what about RFC 4145
(TCP-based media transport in SDP )?
b. Are there any implications for transport of DTLS/SRTP key management?
 (e.g. if ICE TCP is needed because UDP transport is blocked, wouldn't that
also cause the DTLS/SRTP key exchange to fail?)
c. Are there any implications for the data channel?  (e.g. if UDP transport
is blocked, wouldn't SCTP/DTLS/UDP also not work?)

On Mon, Mar 31, 2014 at 8:08 AM, Harald Alvestrand <harald@alvestrand.no>wrote:

>  On 03/31/2014 03:54 PM, Rauschenbach, Uwe (NSN - DE/Munich) wrote:
>
> Hi Harald,
>
> The decision in London to make ICE-TCP a "MUST" has not been implemented in the -03 draft
> (I have noticed you have reflected the related TURN-TCP discussion already).
>
> What are your plans for addressing the ICE-TCP decision?
>
>
> Thanks!
>
> Checking the minutes, I find:
>
>
> Chairs called a hum between the alternatives:
>
> 1) TCP ICE Candidates are MUST implement
> 2) TCP ICE Candidates are SHOULD/MAY implement
> 3) TCP ICE Candidates will not be discussed in document.
>
> The hum indicated very strong support for 1).
>
>
> My memory was faulty; I had thought that the same conclusion had applied
> for TCP ICE candidates as for TURN ICE candidates. Version -04 will have
> this fixed.
>
>
>
> Kind regards,
> Uwe
>
>
>
>  -----Original Message-----
> From: rtcweb [mailto:rtcweb-bounces@ietf.org <rtcweb-bounces@ietf.org>] On Behalf Of ext Harald
> Alvestrand
> Sent: Monday, March 31, 2014 3:40 PM
> To: Magnus Westerlund; rtcweb@ietf.org; Harald Alvestrand
> Subject: [rtcweb] Transport -03, bundling question (Re: Comments on
> draft-ietf-rtcweb-transports-02)
>
> Version -03 is now published. I hope you like it!
>
> New outstanding item:
>
> I added requirements for implementations to be able to generate both
> fully bundled (one 5-tuple for everything) and fully unbundled (one
> 5-tuple for each flow) configurations, and for implementations to be
> able to tolerate being hit with any combination of bundling schemes.
>
> Is there a need to specify at MUST, SHOULD or MAY levels other
> combinations?
>
> Harald
>
> _______________________________________________
> rtcweb mailing listrtcweb@ietf.orghttps://www.ietf.org/mailman/listinfo/rtcweb
>
>
>
> --
> Surveillance is pervasive. Go Dark.
>
>
> _______________________________________________
> rtcweb mailing list
> rtcweb@ietf.org
> https://www.ietf.org/mailman/listinfo/rtcweb
>
>