Re: [MMUSIC] DTLS-over-SCTP, anyone?

Hello

I think the fact that the draft doesn't support DTLS-over-SCTP has been 
known for some time. No one has been pushing for its inclusion. I'd be 
happy if it was left out to be able to progress the draft.

Alternatively we can do what I think Albrecht is suggesting. Reserve 
proto values for DTLS/SCTP/IP and SCTP/UDP/IP in the draft and indicate 
their use by SDP should be defined in a future specification. I don't 
think its worth solving that problem now.

Regards, Christian

On 9/02/2016 7:45 AM, Christer Holmberg wrote:
>
> Hi,
>
> > The problem with DTLS-over-SCTP (RFC 6083) is that it is not fully 
> compatible with draft-ietf-mmusic-dtls-sdp.
>
> > In particular, RFC 6083 does not allow DTLS association span across 
> multiple SCTP associations, but
>
> > draft-ietf-mmusic-dtls-sdp allows to preserve DTLS association 
> through a transport change. The most likely
>
> > solution to this is updating RFC 6083, but, if no one is using 
> DTLS-over-SCTP, it would be easier just to say
>
> > that DTLS-over-SCTP is not supported.
>
> **If** we decide to keep DTLS-over-SCTP in the draft, my suggestion 
> would be to say that “DTLS preservation over transport change” does 
> not apply to DTLS-over-SCTP, with a reference to RFC 6083.
>
> Then, if someone at some point updates RFC 6083, the SCTP-SDP spec 
> also has to be updated.
>
> I do NOT want to delay draft-sctp-sdp until a possible RFC 6083 update 
> has been done (assuming people would agree to do it in the first place).
>
> Regards,
>
> Christer
>
> _____________
> Roman Shpount
>
> On Mon, Feb 8, 2016 at 10:12 AM, Schwarz, Albrecht (Nokia - DE) 
> <albrecht.schwarz@nokia.com <mailto:albrecht.schwarz@nokia.com>> wrote:
>
>     Hello Christer,
>
>     the titel of the (future) RFC will not exclude any SCTP transport
>     modes.
>
>     And a comprehensive and future safe protocol specification should
>     cover all existing ones, which is “DTLS/SCTP/IP” as well as
>     “SCTP/UDP/IP” (RFC 6951).
>
>     These two SCTP transport modes should be indicated as well,
>     independent of potential intentions to be used soon.
>
>     Codepoint space could be reserved, placeholder sections tagged as
>     “not yet supported” of “for further studies”, etc, but the worst
>     case would be any kind of interaction issues in future.
>
>     My view,
>
>     Albrecht
>
>     *From:*mmusic [mailto:mmusic-bounces@ietf.org
>     <mailto:mmusic-bounces@ietf.org>] *On Behalf Of *EXT Christer Holmberg
>     *Sent:* Montag, 8. Februar 2016 15:33
>     *To:* mmusic <mmusic@ietf.org <mailto:mmusic@ietf.org>>
>     *Subject:* [MMUSIC] DTLS-over-SCTP, anyone?
>
>     Hi,
>
>     draft-ietf-mmusic-sctp-sdp currently defines the SDP O/A
>     procedures for SCTP, SCTP-over-DTLS, and DTLS-over-SCTP.
>
>     As we know, SCTP-over-DTLS is used for the WebRTC data channel.
>
>     My question is: does anyone intend to use DTLS-over-SCTP?
>
>     Regards,
>
>     Christer
>
>
>     _______________________________________________
>     mmusic mailing list
>     mmusic@ietf.org <mailto:mmusic@ietf.org>
>     https://www.ietf.org/mailman/listinfo/mmusic
>
>
>
> _______________________________________________
> mmusic mailing list
> mmusic@ietf.org
> https://www.ietf.org/mailman/listinfo/mmusic