Manoj,
I'm sure that was an oversight when the tunneling stuff was added. There are plenty of examples of that sort of thing in H.323. Let's see, how shall we rationalize this... :-) Okay, how about this? When talking about tunneling, H.323 always uses the term, "separate H.245 connection" or "separate H.245 channel." The important qualifier here is "separate," so when H.323 simply talks about the H.245 Control Channel, for example, it is referring to the logical channel irrespective of the transport mechanism, i..e., separate or tunneled. Therefore an EP is required to transmit endSessionCommand regardless of whether the channel is separate or tunneled. How does that sound?
Paul Long ipDialog, Inc.
---------- Original Message ---------------------------------- From: "Paul, Manoj" mpaul@TRILLIUM.COM Reply-To: "Paul, Manoj" mpaul@TRILLIUM.COM Date: Thu, 12 Apr 2001 19:07:02 -0700
Hi,
Does an endpoint using Tunnelling to exchange H.245 messages with peer, need to send EndSessionCommand message during call termination? There are two different texts relating to this:
H.245 Specification(V7), Section B.13.4: "This command indicates the end of the h245 session. After tx EndSessionCommand, the terminal shallnot send any more of the messages defined in this Recommendation".
H.323 Specification (V4), Section 8.5: Phase E-Call Termination "..It shall transmit H.245 endSessionCommand message in the H.245 Control Channel..."
First statement seems to indicate that notwithstanding how H.245 messages are exchanged, an EndSessionCOmmand should be sent to stop any further exchange. Second statement seems to mean that EndSessionCommand is sent only when H.245 Control Channel exists (tunnelling not used).
Could someone clarify? thanks Manoj Paul. Trillium Digital Systems.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ For help on this mail list, send "HELP ITU-SG16" in a message to listserv@mailbag.intel.com