Re: [AVT] RTP over TCP.

[Ram Kordale <[e-mail]>]

Philippe,

At 11:56 AM 5/15/2002 +0200, [e-mail] wrote:

>Ram, peter,
>
>We would also support work on this issue.

Thanks for your support.


>(I cannot make committement on any implementation at this stage;-)
>
>Maybe we could also consider the following option:
>
>What exactly would prevent to interleave RTP in RTSP on TCP for some 
>streams and use RTP/UDP for others ?

RFC 2326 recommends this option only when the option of sending data 
separately is not available. The reasons for this must be similar to the 
argument in section 5.2 in the ID for RTP.

Ram


>regards,
>
>Philippe Gentric
>Software Architect
>Philips MP4Net
>[e-mail]
>http://www.mpeg-4.philips.com
>
>
>"Peter Westerink" <[e-mail]>
>Sent by: [e-mail]
>
>05/14/2002 22:03
>
>         To:        Ram Kordale <[e-mail]>
>         cc:        [e-mail]
>[e-mail]
>Ross Finlayson <[e-mail]>
>[e-mail]
>(bcc: Philippe Gentric/MP4-SUR/CE/PHILIPS)
>         Subject:        Re: [AVT] RTP over TCP.
>
>         Classification:
>
>
>
>
>Ram,
>
>I believe such a profile existed in draft-ietf-avt-profile-new-09.txt, but
>it was removed in version 10 because "there were no 2 implementations".
>There were no technical reasons mentioned. The draft proposed prepending
>each RTP packet with a 2-byte length and that's all.
>
>IBM is interested in the same thing you're after, and I can assume there
>are more MPEG-4 companies that also would want to send their BIFS and OD
>streams reliably (albeit not guaranteed on-time) over TCP. With that we
>should be able to come up with 2 implementations and hence re-instate the
>profile?
>
>      Peter
>
>
>
>
> 
>
>                      Ram 
> Kordale 
>
>                      <kordale@kasenna.        To:       Ross Finlayson 
> <[e-mail]>
>                      com>                     cc:       <[e-mail]>, 
> [e-mail]
>                      Sent by:                 Subject:  Re: [AVT] RTP 
> over TCP.
>                      [e-mail] 
 >
>                      g 
 >
> 
>
> 
>
>                      05/14/2002 
> 03:50 
>
>                      PM 
 >
> 
>
> 
>
>
>
>
>Ross,
>
>Section 10.12 describes only the interleaving case. What about the case
>where RTP and RTCP data need to be transferred on separate TCP connections?
>
>This will need a separate profile. Does such a profile exist?
>
>Thanks.
>Ram
>
>At 12:17 PM 5/8/2002 -0700, Ross Finlayson wrote:
> >At 10:45 AM 5/8/02, Ethen Bommaiah wrote:
> >>I was curious to know if there were any standard profiles that define
> >>transferring RTP
> >>packets over TCP. In particular, we were interested in transferring RTP
>data
> >>over TCP
> >>while using RTSP as the control protocol.
> >
> >Yes - see section 10.12 of the RTSP specification (RFC 2326).  This
> >defines how RTP packets may be carried over the RTSP (TCP) stream.  (This
> >is actually not a new RTP 'profile', but instead, an optonal mechanism
> >defined within RTSP.)
> >
> >         Ross.
> >
> >
> >_______________________________________________
> >Audio/Video Transport Working Group
> >[e-mail]
> >https://www1.ietf.org/mailman/listinfo/avt
>
>
>_______________________________________________
>Audio/Video Transport Working Group
>[e-mail]
>https://www1.ietf.org/mailman/listinfo/avt
>
>
>
>
>
>
>_______________________________________________
>Audio/Video Transport Working Group
>[e-mail]
>https://www1.ietf.org/mailman/listinfo/avt
>