Re: [AVTCORE] [xrblock] Fw: I-D Action: draft-ietf-avtcore-monarch-02.txt

I agree with Colin

Peter Musgrave
(as individual)

On Tue, Jun 14, 2011 at 6:27 AM, Colin Perkins <csp@csperkins.org> wrote:

> On 10 Jun 2011, at 09:59, Qin Wu wrote:
> > Hi,
> >> ----- Original Message -----
> >> From: "Colin Perkins" <csp@csperkins.org>
> >> To: "Qin Wu" <sunseawq@huawei.com>
> >> Cc: <avt@ietf.org>; <xrblock@ietf.org>; <magnus.westerlund@ericsson.com
> >
> >> Sent: Thursday, June 09, 2011 9:58 PM
> >> Subject: Re: [AVTCORE] Fw: I-D Action: draft-ietf-avtcore-monarch-02.txt
> >>
> >>
> >> On 27 May 2011, at 03:36, Qin Wu wrote:
> >>> This version addresses the open issues regarding identity information
> repetition and acknowledge section.
> >>> http://www.ietf.org/internet-drafts/draft-ietf-avtcore-monarch-02.txt
> >>>
> >>> The diff is:
> >>> http://tools.ietf.org/rfcdiff?url2=draft-ietf-avtcore-monarch-02
> >>>
> >>> Comments are welcome!
> >>
> >> This helps, but to my mind doesn't go far enough. We have the SSRC to
> identify participants. I don't see much benefit from having a separate
> identity tag in XR blocks.
> >>
> > [Qin]:  I think that we are talking about how to define future RTCP XR
> Block. It seems to me the rule we are following  to define the new RTCP XR
> BLock in the avtcore-monarch is a little different from the rule of we are
> doing to the existing RTCP XR described in RFC3611.
> >
> > As for the existing XR Block defined in RFC3611, each of the existing XR
> Block has allocated  a field for SSRC of source. However such identity
> information can be repeated several times if several metric blocks reporting
> one stream from one source  is carried in the same RTCP Packet.
> >
> > In order to reduce overhead to carry duplicated SSRC of source in each
> metric block,  in the new XR Block, we separate such SSRC of source out from
> each metric block.
> >
> > One way is to  allocate identity tag field in each metric block which can
> be used to tell which metric block report on stream from which source.
> >
> > Another way is to follow the existing rules defined in RFC3611. put back
> the SSRC of source field into each metric block which also can be used to
> tell which metric block report on stream from which source.
> > However this rule seems to contradict to the rule of reducing identity
> information repetition discussed in avtcore-monarch.
> >
> > Which way we should follow?
>
>
> I think you should follow RFC 3611.
>
> The savings from using the new identity tag don't seem worth the added
> complexity.
>
> --
> Colin Perkins
> http://csperkins.org/
>
>
>
> _______________________________________________
> Audio/Video Transport Core Maintenance
> avt@ietf.org
> https://www.ietf.org/mailman/listinfo/avt
>