[Live-devel] testMPEG1or2AudioVideoToDarwin source code doubt

[Ross Finlayson]

At 01:13 PM 7/29/04, you wrote:
>the last RTSP command issued by testMPEG1or2AudioVideoToDarwin seems to be
>different from a regular broadcaster...wat i mean to say is after the
>ANNOUNCE,SETUP commands QuickTime Broadcaster issues a RECORD command
>where as testMPEG1or2AudioVideoToDarwin issues a PLAY command...why is
>this being done any specific reason...

It used to be the case that "QuickTime Broadcaster" issued a "PLAY" to talk 
to Darwin.  The current "DarwinInjector" code just duplicates that 
protocol, to make it more likely that it will work with all DSS versions.

>also in the SDP file we are putting zero's in variable 'm' instead of
>telling at which ports the Audio or Video is being broadcasted...is this
>done only because we are using same TCP port to send the RTP stream...

No, this is standard for all unicast RTSP/RTP streams - not just those 
streamed with/through Darwin.  It tells the media player client that *it* 
is to choose the port number(s) that it wishes to use to receive the 
stream.  It then tells the server these port numbers in the subsequent RTSP 
"SETUP" (and "PLAY") commands.


	Ross Finlayson
	LIVE.COM
	<http://www.live.com/>