The Gandi Community

Tech Fundamentals: IANA

In March 1972, Jon Postel along with Vint Cerf called for the creation of a catalog of socket numbers. The purpose was to create, essentially, a list of all the existing codes and numbering systems allowing computers to network with each other.

Postel was editor of the Request for Comment (RFC) document series in which this call and the resultant catalog were published. In it he dubbed himself the “czar of the socket numbers,” in December 1972 when the catalog was complete.

This role as “czar” became a permanent function. He took it with him when he earned his Ph.D. from UCLA and moved to the University of Southern California’s Informational Sciences Institute (USC/ISI). There, he brought on-board Joyce Reynolds, a graduate student.  The two of them became the “Internet Assigned Numbers Authority,” or IANA.

IANA then collaborated with Elizabeth Feinler from Stanford and the rest of Stanford Research Institute’s Network Information Center (SRI-NIC) to maintain, more or less, a directory for the Internet.

When a new host joined the network, they would email Feinler. Feinler would then add them to a HOSTS.TXT file sent out and installed manually on every networked machine. When domain names were invented, the task shifted to adding strings to the root name server.

In 1990, NIC’s function was contractually shifted to Network Solutions, a private company. In 1995, they received authorization from the National Science Foundation to charge a fifty-dollar fee per year on domain name registration. This move caused widespread dissatisfaction with Network Solution’s concentration of power and money. The researchers who had been contracted to create the Internet were especially critical.

It wasn’t until January 28, 1998 that this conflict came to a head. On that day, Postel made history. He emailed eight of the twelve operators of the Internet’s regional root nameservers and told them to change their root zone servers from Network Solutions’s A.ROOT-SERVERS.NET to IANA’s DNSROOT.IANA.ORG. The operators made the switch. This “test” demonstrated that the ultimate control of the root zone belonged to IANA and not Network Solutions.

This led directly to the Clinton administration’s creation of ICANN to take control of the IANA function. Despite being told by presidential science advisor Ira Magaziner that he would “never work on the Internet again,” Postel was set to become ICANN’s first chair.

Sadly, Postel died before ICANN was realized, and it was Joyce Reynolds who managed the transition of the IANA function from USC/ISI to ICANN.  Ostel’s old colleague Steve Crocker who took the role of first ICANN chair. Postel is memorialized in RFC 2468: “I remember IANA.”

Reynolds passed away more recently, just this December, but hers and Postel’s legacy lives on. Their contribution can be seen not only in the numerous technical documents they authored and co-authored. They are also visible in principles like “rough consensus” and “Postel’s law” which states a fundamental rule of networking: “Be liberal in what you accept and conservative in what you send.”