Ethereal-dev: Re: [Ethereal-dev] [PATCH] T.38 dissector improvements

Note: This archive is from the project's previous web site, ethereal.com. This list is no longer active.

From: Guy Harris <guy@xxxxxxxxxxxx>
Date: Mon, 5 Jan 2004 19:47:06 -0800

On Jan 5, 2004, at 9:12 AM, Martin Regner wrote:

I also needed to do some small changes in epan/conversation.c and
epan/packet.h so that Ethereal will continue with other dissectors if the
conversation dissector rejects the packet.

In what cases would a conversation dissector reject a packet? Conversation dissectors are usually set up either for

1) earlier packets indicate that a future conversation will be using a particular protocol - e.g., RTSP or SDP specifies a future RTP/RTCP conversation

or

2) a dissector needs to ensure that future packets that might not otherwise be handed to that dissector be handed to it (some heuristic dissectors might do to keep subsequent packets from being taken by other dissectors, and the TFTP dissector uses it to ensure that packets after the first one get dissected as TFTP)

and in both of those cases the packets shouldn't be inappropriate.

Is the problem that you have, in a capture, more than one conversation between two endpoints, the first of which is RTP and the second of which is T.38?