The story
In November 1983, Paul Mockapetris of the University of Southern California's Information Sciences Institute published the design of the Domain Name System, or DNS, in two Internet standards documents, RFC 882 and RFC 883. The system translates names people can remember, such as example.com, into the numeric addresses that computers use. It is a distributed database organized as a tree, with a root at the top, top-level domains such as .com and .org below, and servers around the world responsible for their own branches. The first .com name, symbolics.com, was registered on March 15, 1985.
Before DNS, every computer on the early internet found others using a single file called HOSTS.TXT, kept by the Network Information Center at SRI and downloaded by everyone. That worked for a few hundred machines on ARPANET, but by the early 1980s the growth of TCP/IP networks made it impractical, since each change had to be pushed to every site. DNS replaced it with delegation: each organization manages its own names and answers queries about them, and servers cache answers to save work. The design was revised in 1987 as RFCs 1034 and 1035, which are still the basis of the system.
Why it mattered
- DNS let the internet grow from hundreds to billions of named devices without a central list.
- Domain names became a commercial asset, giving rise to the domain registration business and later disputes over trademarks.
- The web depends on DNS, since every browser request starts with a name lookup, which also makes DNS a target for attacks and a tool for filtering and censorship.
Sources
- And the DNS was Born USC Information Sciences Institute
- A Brief History of the DNS and BIND Internet Systems Consortium
- Development of the Domain Name System ACM SIGCOMM (Mockapetris and Dunlap)
Drafted by AI (Claude) from the sources above and checked against them by AI; no historian has reviewed this page. Spot a mistake? Suggest a correction. How pages are made
Community notes
Loading notes…