Re: [xrblock] [AVTCORE] Open issue- identity info repetition indraft-ietf-avtcore-monarch

[Qin Wu <[e-mail]>]

Hi, Peter:
Thank for your comments. Please see my following question below.
  ----- Original Message ----- 
  From: Peter Musgrave 
  To: [e-mail] ; [e-mail] ; [e-mail] 
  Sent: Monday, June 06, 2011 3:01 AM
  Subject: Re: [AVTCORE] [xrblock] Open issue- identity info repetition indraft-ietf-avtcore-monarch


  Hi Qin, Colin, 


  I agree with Colin that the case for reducing identity inefficiency is not very compelling. These kinds of blocks are not sent as frequently as e.g. the RTP payload and in general a number of XR blocks can be stacked in one MTU with ease. 

  [Qin]: Yes, such block is not sent frequently. Suppose we separate identity block out from each metric block 
  and allow identity block only being sent once at the start of session, is it compelling to have such independent identity block?

  Hence I would suggest we not incur the protocol and implementation complexity of having identity blocks and references to them etc. 

  [Qin]: I agree that it is not compelling to let each small metric block to reference the independent identity block with identity information included if
  each metric block already have SSRC and support such reference. However it looks to me measure information defined in measurement identity 
  block of draft-ietf-avt-rtcp-xr-meas-identity-02 contain more than SSRC which could be used as the key to look up other application specific 
  information, e.g., diagonose information. So it will be good to still have identity block.

  Regards, 


  Peter Musgrave


  (I hope this threads properly - I have changed email address in the middle of this exchange. Apologies for not having the rest of the exchange below).