A rainbow of Ethernet cables

40 years ago—in September 1981—DARPA published the finalized specifications of the TCP/IP protocol suite, which defines the basic rules for how the internet works. While TCP/IP didn’t become widely adopted until 1983, this milestone can help us understand why TCP/IP was so important.

What Is TCP/IP?

TCP/IP is a protocol suite consisting of two main protocols conceived by Vint Cerf and Bob Kahn, Transmission Control Protocol (TCP) and Internet Protocol (IP). Internet Protocol defines addressing and routing—how packets of data flow through the network. The Transmission Control Protocol handles making connections and ensuring data packets get to their proper destination. The two protocols work together to create the foundation of the modern internet.

RELATED: How Do IP Addresses Work?

Why Was TCP/IP Created?

Before the internet, the US Department of Defense (through ARPA), created a computer network called ARPANET that linked US government and university computers around the country. ARPANET came online in 1969. Prior to TCP, ARPANET used a protocol called NCP (Network Control Program) for making connections between machines on the network.

According to the NCP/TCP transition plan (RFC801) published in November 1981, the need for TCP/IP arose from multiple fronts. Increasingly, experimental computer networks had begun using radio and satellite links rather than physical wires. Also, organizations had increasingly researched local networks—groups of machines communicating together within the same facility rather than over long distances. The architects of ARPANET realized that the underlying protocols then in use were “inadequate” for spanning all these different and new types of networks.

A map of ARPANET from 1980
A geographical map of ARPANET in 1980. DARPA

At the same time, during the 1970s, companies such as IBM, DEC, AT&T, and Xerox had created their own proprietary, incompatible computer networks that fragmented the sharing of information. So the TCP/IP suite was immediately notable because it represented a non-proprietary, royalty-free, open-architecture solution that allowed computers of any kind to communicate through any medium, as long as the TCP/IP software was implemented on the system.

Development on TCP and IP began in 1973 by Vint Cerf and Bob Kahn. After development throughout the 1970s by Cerf, Kahn, and others, DARPA published specifications on TCP and IP in RFC documents 791 and 793, dated September 1981, which represented the first public introduction of the finalized TCP/IP framework.

How Does TCP/IP Work?

TCP and IP are two separate technologies that work together, hand-in-hand, to achieve reliable connections through a heterogeneous (many different types of computers and links) computer network.

As previously mentioned, IP handles addressing machines on the network and how blocks of data (called “packets“) reach the proper destination. TCP ensures that the packets reach their destination without error, calling ahead to make sure there is a host to receive the information and, if the information is lost on the way or corrupted, re-transmitting the data until it gets there safely.

TCP/IP’s architects purposely separated the implementation of TCP and IP to make the network more flexible and modular. In fact, TCP can be swapped out with a different protocol called UDP that is faster but allows data loss in situations where 100% transmission accuracy isn’t necessary, such as a telephone call or a video broadcast.

Network engineers call this modular design a “protocol stack,” and it allows some of the lower layers in the stack to be handled independently in a way that is most appropriate for the local machine architecture. Then the upper layers can work on top of those to communicate with each other. In the case of the Internet, this stack typically consists of four layers:

  • Link Layer – Low-level protocols that work with a physical medium (such as Ethernet)
  • Internet Layer – Routes packets (IP, for example)
  • Transport Layer – Makes and breaks connections (TCP, for example)
  • Application Layer – How people use the network (the web, FTP, and others)

The protocols that handle the web (such as the HyperText Transfer Protocol, or HTTP) are on the application layer, and they work on top of TCP and IP. Thanks to this model, HTTP does not need to know how to make or break connections at a low level—all of that is handled by the protocols lower in the stack. It makes for a very flexible system and is the reason why TCP/IP were so successful and why they still serve as the backbone of the internet today.

RELATED: The First Website: How the Web Looked 30 Years Ago

When Did TCP/IP Come Into Use?

While in development, TCP/IP came into experimental use as early as 1973. As its creators continued to refine the protocols, Internet Protocol (IP) went from version 1 to version 4 by 1981, which is still the version of IP still widely used today.

Although DARPA introduced the first finalized version of the TCP and IP protocols (version 4) in September 1981, some ARPANET computers continued using the earlier ARPANET protocols (such as NCP) for a time. As with any entrenched technology, change can take time, and the architects of the plan designed a transition period between NCP and TCP that would end on January 1, 1983.

Vint Cerf on the cover of the August 1996 issue of Boardwatch Magazine wearing an "I P on Everything" t-shirt.
Vint Cerf on the cover of the August 1996 issue of Boardwatch Magazine wearing an “IP on Everything” t-shirt in an era where IP wasn’t the clear protocol winner. Boardwatch Magazine

The January 1, 1983 “flag day” (a day when a dramatic change happens in computing), marked the beginning of widespread use of TCP/IP and the birth of the modern internet. Even then, other network protocols remained widely in use, and it wasn’t until the mid-1990s that TCP/IP became the clear “winner” in what some call the Protocol Wars.

RELATED: How Does the Internet Work?

The Future of TCP/IP

Currently, most of the internet runs on Internet Protocol version 4, commonly called “IPv4.” But there’s a newer version called “IPv6,” introduced in 1998, that is slowly rolling out over time (very slowly). Among IPv6’s most important features is support for 128-bit addresses, allowing for 340 trillion trillion trillion devices with unique IP addresses on the network.

In contrast, IPv4 supports 32-bit addressing, allowing over 4.2 billion IP addresses. While 4.2 billion sounds like a lot, we’ve already reached the limit of IPv4 addresses assigned some time within the 2010s, depending on how you choose to measure it.

Luckily, IPv4 and IPv6 are interoperable, so computer vendors, internet hosts, and assigning authorities have some breathing room while making the transition to IPv6 over time. Even with all its improvements, IPv6 traces it architecture back to the same research started by Cerf and Evans in 1973 and finalized in 1981. That’s quite a legacy. Happy Birthday, TCP/IP!

Profile Photo for Benj Edwards Benj Edwards
Benj Edwards is a former Associate Editor for How-To Geek. Now, he is an AI and Machine Learning Reporter for Ars Technica. For over 15 years, he has written about technology and tech history for sites such as The Atlantic, Fast Company, PCMag, PCWorld, Macworld, Ars Technica, and Wired. In 2005, he created Vintage Computing and Gaming, a blog devoted to tech history. He also created The Culture of Tech podcast and regularly contributes to the Retronauts retrogaming podcast.
Read Full Bio »