On Sat, Apr 4, 2009 at 11:45 AM, Joel M. Halpern <[email protected]> wrote: > William Herrin wrote: >> Identifier: Any element in the ongoing communication between a user >> and a network service he uses which is statically bound to the >> function of correlating the communication's data packets with the >> communication as a whole.
> That is a very specific kind of identifier. It sounds to me like the > definition of session identifier. Joel, Either I didn't describe it well enough or you read it too narrowly. It's any element at all that's incorporated into that correlation process with a static binding. It's the session ID that incorporates multiple transports into a session but its also the transport ID that incorporates the packets into each transport as well as the components (srcip, dstip, srcport, dstport) from which the transport ID is constructed. Not only that, it's the login that correlates multiple sessions into an account. If it's statically incorporated into that correlation process in any way then it's an ID. In the most general sense, an identity correlates a set of smaller entities into the single larger entity which contains them. If you think that's wrong, poke a hole in it. Give me one counter-example of an identifier which does not function as a correlator. If I can't either explain how it's used as a correlator or explain why it's not an identifier then I'll have to find a new theory. Regards, Bill Herrin -- William D. Herrin ................ [email protected] [email protected] 3005 Crane Dr. ...................... Web: <http://bill.herrin.us/> Falls Church, VA 22042-3004 _______________________________________________ rrg mailing list [email protected] http://www.irtf.org/mailman/listinfo/rrg
