Re: [AVTCORE] Comments on draft-ietf-avtcore-rfc5285bis-03

["Roni Even" <[e-mail]>]

Hi Magnus,
So can you try to summarize what text you propose to add
Thanks
Roni

> -----Original Message-----
> From: Magnus Westerlund [mailto:[e-mail]]
> Sent: Tuesday, October 18, 2016 4:52 PM
> To: David Singer
> Cc: Roni Even; [e-mail]; IETF AVTCore WG
> Subject: Re: [AVTCORE] Comments on draft-ietf-avtcore-rfc5285bis-03
> 
> Den 2016-10-18 kl. 12:07, skrev David Singer:
> >
> >> On Oct 18, 2016, at 16:53 , Magnus Westerlund
> >> <[e-mail]> wrote:
> >>
> >> I think there is a negotiation mechanism, in that if one include
> >> extmap mappings for values above 15, then you indicate that you
> >> support the two-byte format, and if you accept you can clearly
> >> indicate this. Maybe this should be made more explicit, as a method
> >> of determining the operations mode. To me we do now end up with four
> >> different cases:
> >>
> >> A. One-byte only: only ID's in range 1-14. The state of
> >> a=extmap-allow-mixed has no meaning, only indication if one want to
> >> renegotiate.
> >>
> >> B. two-byte only: only ID's above 15. The state of
> >> a=extmap-allow-mixed has no meaning, only indication if one want to
> >> renegotiate.
> >
> > sorry, I don’t get this.  Two-byte headers get you full-byte length
> > fields and thus longer payloads. You could use these with small local
> > IDs.  Why say you have to use IDs outside 1-14???
> 
> Yes, you are correct. I made a mistake in my thoughts, and forgot about
> these possibilities. I would like to correct myself to say that only at
least one
> ID above 15 implies and no a=extmap-allow-mixed that you can use two-
> byte headers and may do that for some RTP streams.
> 
> But the other implication of your comment is that one can't draw the
> conclusion that IDs only in 1-14 range implies only one-byte headers.
> 
> So to make this more complete, the additional criteria that can be used to
> determine at signaling time if one will use one or two-byte headers is the
> inclusion and acceptance of an RTP header extension that requires a length
> of data that is more than 16 bytes. But, for extension that doesn't have
that
> given, like the CNAME that could be of quite different lengths both above
> and below the range one can't make that determination.
> 
> I was trying to arrive here that one could determine support for two-byte
> headers based on what was signalled, even without the a=extmap-allow-
> mixed SDP attribute.
> 
> So a=extmap-allow-mixed provides not only indication that you can switch,
> but also that one supports both one and two-byte headers.
> 
> But for the legacy case, one could negotiate using low IDs (1-14) range
only,
> and then realize at transmission start time that one have to use the
two-byte
> headers. Then one could end up in an interoperability failure as the old
> receiver is not required to support the two-bytes header.
> 
> So based on the above implications, one could write text that gives
> recommendations that for any signaling case where one knows that one will
> have to use two-byte to assign that extension an ID above 15 to indicate
this.
> But, I guess for the one where one don't know one simply have to hope that
> it works, and indicate that this issue is resolved if one follows this
> specification and indicates a=extmap-allow-mixed.
> 
> Cheers
> 
> Magnus Westerlund
> 
> ----------------------------------------------------------------------
> Services, Media and Network features, Ericsson Research EAB/TXM
> ----------------------------------------------------------------------
> Ericsson AB                 | Phone  [phone]
> Färögatan 6                 | Mobile [phone]
> SE-164 80 Stockholm, Sweden | mailto: [e-mail]
> ----------------------------------------------------------------------