HOW IP ADDRESSES
GOT THEIR JOB.

01 / FIRST, THE NETWORK

Packets needed a common envelope.

Early packet-switched networks were built to connect unlike machines and unlike networks. The Internet Protocol supplied that common network-layer envelope: source address, destination address, and the information routers needed to move a datagram across an internetwork. RFC 791, published in 1981, defined IPv4 as a 32-bit addressing system.

Thirty-two bits produce 4,294,967,296 possible values. That looked immense in a research network era; it became constrained once consumer internet access, mobile devices, cloud infrastructure, and embedded systems arrived.

02 / SCARCITY CHANGED DESIGN

Addresses became a resource, not just a field.

Subnetting let organizations divide an allocated network into smaller operational networks. Later, classless routing made address allocation and route aggregation more flexible. Network address translation stretched IPv4’s lifespan by letting many private devices share a smaller set of public addresses—but it also made end-to-end connectivity less direct.

03 / THE 128-BIT ANSWER

IPv6 expanded the map.

IPv6 was designed as IPv4’s successor. Its 128-bit addresses dramatically expand the address space, while its header design and extension-header model modernize the protocol. IPv6 does not have a broadcast address; multicast is used where IPv4 once relied on broadcast. The change is not merely “more addresses”—it is a different operating model for autoconfiguration, addressing, and network design.

04 / THE CURRENT INTERNET

Coexistence is the practical story.

IPv4 has not disappeared, and IPv6 has not replaced it overnight. Most networks operate a transition period using dual-stack, translation, or tunneling techniques. For operators, the useful skill is not picking a side; it is being precise about which protocol, prefix, route, and policy is in front of you.