NAT
/nat/ · noun · Development · Origin: 1994
Definitions
Network Address Translation — a technique where a router modifies IP addresses in packet headers to map between private (internal) and public (internet-facing) addresses. NAT is why millions of home networks can all use 192.168.1.x internally while sharing a single public IP. It extended IPv4's life by decades but complicates peer-to-peer communication.
In plain English: The thing that lets all devices in your home share one internet address — your router translates between your private network and the public internet.
Example: The WebRTC call kept failing because both users were behind NAT and the STUN server couldn't punch through the symmetric NAT on the corporate firewall.
Origin Story
The clever hack that kept IPv4 alive decades past its expiration date
**Network Address Translation** was formalized in RFC 1631 (1994) by Kjeld Borch Egevang and Paul Francis. The internet was running out of IPv4 addresses (only 4.3 billion available), and NAT offered a stopgap: let multiple devices share a single public IP address.
NAT works by rewriting IP addresses in packet headers as they pass through a router. Your home network might have dozens of devices, all using private addresses (192.168.x.x), but they appear to the outside internet as a single public IP.
NAT was supposed to be temporary until IPv6 rolled out. Three decades later, IPv6 adoption is still incomplete, and NAT remains everywhere. It accidentally became a security feature too -- devices behind NAT are harder to reach from the internet, providing a basic firewall effect.
Coined by: Kjeld Borch Egevang, Paul Francis
Context: RFC 1631, 1994
Fun fact: NAT broke the internet's original end-to-end design principle, where any device could directly communicate with any other. This is why peer-to-peer applications, VoIP, and online gaming need NAT traversal hacks like STUN, TURN, and ICE.