48-bit absolute internet and Ethernet host numbers

Abstract

Xerox internets and Ethernet local computer networks use 48-bit absolute host numbers. This is a radical departure from practices currently in use in internetwork systems and local networks. This paper describes how the host numbering scheme was designed in the context of an overall internetwork and distributed systems architecture.

On the Naming and Binding of Network Destinations

“…Although the ideas outlined so far seem fairly straightforward, it is surprisingly easy to find real-world examples that pose a challenge in interpretation. In the Xerox/DEC/Intel Ethernet [5,6], for example, the concept of a network attachment point is elusive, because it collapses into the node name. A node can physical attach to an Ethernet anywhere along it; the node brings with it a 48-bit unique identifier that its interfaces watches for in packets passing by.…”

“…For another example, the ARPANET NCP protocol provides character string names that appear, from their mnemonics, to be node names or service names, but in fact they are the names of network attachment points [6]. Thus the character string name RADC-Multics is the name of the network attachment point at ARPANET IMP 18, port 0, so reattaching the node (a Honeywell 68/80 computer) to another network attachment point requires either that the users learn a new name for the service or else a change of tables in all other nodes.…”