On Fri, Jun 07, 2002 at 02:42:06PM -0700, Ed Sawicki wrote: > I agree. However, in this case, we have a UDP packet with source > port 1745 and destination port 53. Ethereal incorrectly assumes port > 1745 is the real protocol when port 53 is more likely to be correct. > A well-known port (1-1023) should always trump a registered port (1024 > an up). This is far more likely to be correct more of the time than the > reverse. I checked in a change to have the TCP, UDP, and SCTP dissectors try the lower-numbered port first and the higher-numbered port second. > By the way, why does the IANA list call port 1745 remote-winsock? I don't maintain that list, so I have no idea.
Powered by MHonArc 2.6.10