[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [AVT] NewVersion: draft-versteeg-avt-rapid-synchronization-for-rtp-02



Title: Re: [AVT] NewVersion: draft-versteeg-avt-rapid-synchronization-for-rtp-02
Roni-
 
The RS generates the RMS-I, and the semantics of the RMS-I are well defined from the standpoint of the RS. The Max Burst BitRate is the maximum rate at which the RS will send the burst. If the RS knows this value, it should populate the field.
 
The document does not currently specify what the RR does with the Max Burst Bitrate, because the RR could do a lot of different things with it. It may manage the de-jitter buffer. It may manage a decode buffer. It may manage a decrypt buffer. It may manage display buffer. It may display progress bars. It may present different screens to the user based on the anticipated duration of the stream acquisition process. Most of this is payload/implementation specific, and we are trying to stay away from specifying such behaviors.
 
So, in summary, we are trying to rigorously define the semantics of each component of each packet that flows on the network as seen by the sender of that message. We are currently not trying to define the device-specific actions that a given device takes when receiving a message.
 
Having said that, it sounds like we should put some guidance in the text regarding possible uses for the receiver. We do need to be careful not to make to many assumptions about application-specific behaviors, though.
 
bvs
 


From: Roni Even [mailto:ron.even.tlv at gmail.com]
Sent: Thursday, March 12, 2009 12:17 PM
To: Bill Ver Steeg (versteb); yekuiwang at huawei.com; Ali C. Begen (abegen); avt at ietf.org
Cc: Dave Oran (oran)
Subject: RE: [AVT] NewVersion: draft-versteeg-avt-rapid-synchronization-for-rtp-02

Hi,

I will have a separate email with my review comments.

I have one comment on this thread

 


  
   Section 7.2.

How does RR utilize the Max Burst Bitrate conveyed in the RMS-I message?

--- bvs -- This is not specified in the document, as it is out of scope.
However, the client can use this value to assist it in managing buffer
levels.

 

Roni: I think that if you have a parameter in a message its semantics should be specified. If it is not then how would an RS know if to put a value in this parameter or just have it as unspecified.