[AVT] RE: draft-ietf-avt-app-rtp-keepalive-01

Hi

Thank you for the feedback.

You are right there is a lack of details in section 4.8.
The keepalive packet is sent within the same RTP session as the actual media, so yes, the sequence number will be incremented by one for each packet.
For the timestamp, it seems good to fill the field with the same value a media packet would have at this time.
The marker bit is not significant for the keepalive packets, so it should be set to zero.

We'll add more text in a next revision. We'll also fix the typos, thanks.

Regards,
Aurelien 

-----Message d'origine-----
De : Alfred E. Heggestad [mailto:aeh@db.org] 
Envoyé : mardi 18 décembre 2007 11:05
À : MARJOU Xavier RD-CORE-LAN; SOLLAUD Aurelien RD-TECH-LAN; avt@ietf.org
Objet : draft-ietf-avt-app-rtp-keepalive-01

Hi

I have some comments on draft-ietf-avt-app-rtp-keepalive-01. I found the document very useful, many thanks for writing this!

Section 4.8 describes the RTP Packet with unknown payload type. It should be added some text describing how the RTP header fields should be handled.
For example marker bit, sequence number, timestamp. My guess is that the sequence number is incremented by one for each packet sent. What about timestamp? Should it be incremented for each packet sent? Since we are sending 0 bytes of payload, which value should it be increment by.

NITS:

6.  Additional considerations

    The agent SHOULD only send RTP keepalive when it does not send
    regular RTP paquets.
                ^^^^^^^

-> typo: should be "packets"

7.  Security Considerations

    The keepalive packets are sent on the same path as regular RTP media
    packets.  In addition, they do not convey any valuable information.
    So the mechanism described here doest not imply new security issues.
                                    ^^^^^

-> typo: should be "does"

/alfred

_______________________________________________
Audio/Video Transport Working Group
avt@ietf.org
https://www1.ietf.org/mailman/listinfo/avt