packet
/PAK-it/ · noun · Development · Origin: 1960
Definitions
A formatted unit of data carried across a network, made of a header holding delivery information (source and destination addresses, length, protocol, time-to-live) and a payload holding the data itself. Packet switching, developed independently by Paul Baran and Donald Davies in the 1960s (Davies coined the word), breaks messages into small units that are routed independently and reassembled at the destination, so links are shared efficiently and there is no single path to fail. Pedantically the name belongs to layer 3: Ethernet carries frames, IP carries packets, TCP carries segments, and UDP carries datagrams, though everyday usage calls all of them packets. Size is bounded by the path's maximum transmission unit, typically 1,500 bytes on Ethernet, and a packet too large for a link must be fragmented or dropped. Tools such as tcpdump and Wireshark capture and decode packets for debugging.
In plain English: A small chunk of data sent across a network — the internet works by breaking everything into packets, sending them separately, and reassembling them at the destination.
Example: The packet capture showed that requests were leaving our server fine but the response packets were being dropped by the intermediate firewall.
Etymology
- 1960
- Paul Baran at RAND and Donald Davies at NPL independently conceive of breaking data into small, routable blocks.
- 1969
- ARPANET sends its first packets between UCLA and SRI, proving that packet-switched networking works.
- 1974
- Vint Cerf and Bob Kahn publish the TCP specification, defining how packets are reliably transmitted across networks.
- 2020s
- Packets remain the fundamental unit of internet communication. Per-packet inspection and quality-of-service controls are standard in modern networks.
Origin Story
The Tiny Envelope That Carries the Internet
A packet is a small, formatted unit of data transmitted across a network. The concept was independently invented by two researchers in the early 1960s. Paul Baran at the RAND Corporation in the United States developed 'distributed adaptive message block switching' (later called packet switching) as a way to build military communication networks that could survive nuclear attacks. Meanwhile, Donald Davies at the National Physical Laboratory (NPL) in the United Kingdom independently conceived the same idea and coined the word 'packet,' inspired by the idea of small parcels being routed through a postal network. Davies presented his concept in 1965 and built a working packet-switched network at NPL in 1969. The ARPANET, the precursor to the internet, adopted packet switching in 1969, and Baran's and Davies' ideas merged into the foundation of modern networking. In a packet-switched network, data is broken into small packets, each carrying the destination address. Packets may travel different routes through the network and arrive out of order; the receiving system reassembles them. This design was revolutionary compared to circuit switching (used in telephone networks), where a dedicated connection must be maintained for the entire duration of a call, even during silences.
Coined by: Donald Davies (coined the word); Paul Baran (independent co-inventor of the concept)
Context: National Physical Laboratory (UK) and RAND Corporation (US), early 1960s
Fun fact: Paul Baran's original proposal for packet switching was rejected three times by AT&T, whose engineers insisted that their existing circuit-switched telephone network was superior. The internet proved them wrong.