The Internet Before DNS

In the beginning, before the Internet got big or was even called the Internet, each computer on it needed to know only about a handful of other computers on the Net and a list of all the important Internet hosts that was kept in one file called the hosts.txt file. This file was very similar to the /etc/hosts file on UNIX machines and the lmhosts file on Windows machines. In addition, it was maintained by a central authority, the InterNIC, from which computers would periodically transfer it. Site administrators would mail changes to the InterNIC, which would then incorporate the changes. Then, as time progressed, computers would pick up the latest version and get the updated information. Eventually, the change would be known all over the Net.

There is a problem with this simple scheme, though. It does not scale; as the file and Net grow, the maintenance and distribution of a single file from a single point become intractable. In 1983, long before the World Wide Web was even a glimmer in Tim Berens Lee's eye, the Internet was growing quite quickly, and the inadequacy of the hosts.txt file was obvious. It had grown large and cumbersome, and it was easy to see that one central file was not a future-proof solution.