From Coffee Shop Wi-Fi to the World Wide Web: The Surprising History of Ethernet

Imagine for a second that you’re in a busy tech hub circa 1973. You’ve got a few computers in a room, each one a massive, refrigerator-sized machine. They can process data like champs, but they’re basically lonely islands. If you want to move a file from one to another, you literally have to walk over with a stack of punch cards or a magnetic tape reel.

Enter Bob Metcalfe, a young engineer at Xerox PARC—the legendary research lab that gave us the mouse and the graphical user interface. Metcalfe wasn’t just trying to wire computers together; he was trying to solve a physical problem. How do you share a single, expensive laser printer among a handful of machines without a tangled mess of cables and without one machine hogging the line?

His answer? A coaxial cable, a set of rules, and a bit of luck.

The Birth of the “Aloha” Brainstorm

Metcalfe didn’t invent the idea from scratch. He was heavily inspired by the University of Hawaii’s ALOHAnet—a radio-based network that let computers on different islands talk to each other. Aloha had a clever trick: if two computers transmitted at the same time, they’d notice the collision, wait a random amount of time, and try again. It was chaotic but it worked.

Metcalfe realized he could apply the same logic to a wired network. He sketched out a system where computers were all plugged into a single, shared cable. Before sending data, a computer would “listen” to the cable. If it was quiet, it would transmit. If it heard a collision—two computers sending at once—it would back off and retry after a random pause.

He called it Ethernet. The “ether” was a nod to the old concept of a luminiferous ether—an invisible medium through which light waves traveled. It was poetic, but it also fit perfectly: a shared, passive medium that carried data invisibly.

The first Ethernet ran at 2.94 megabits per second. For context, that’s slower than a dial-up modem from the 1990s. But in 1973, it felt like magic.

The Xerox Gamble and the Great Standardization War

Xerox owned the patents, but they weren’t interested in selling Ethernet as a product. They wanted to license it cheaply. In 1980, they teamed up with Intel and Digital Equipment Corporation (DEC) to push a 10 Mbps version into the market. This was a bold move—most corporate networks at the time used bulky, expensive proprietary solutions from IBM or Wang Laboratories.

The problem? Everyone had their own standard. IBM had Token Ring, which worked like a digital relay race—a token passed from machine to machine, and only the holder could send data. It was orderly, but it was slow and finicky. General Motors had their own, called MAP (Manufacturing Automation Protocol). For a while, the future of networking looked like a fragmented warzone.

Ethernet won for two simple reasons: 1. It was cheap. The hardware was simple—just a coaxial cable and a transceiver. 2. It was good enough. Even if collisions happened, the random backoff meant the network self-healed.

In 1983, the Institute of Electrical and Electronics Engineers (IEEE) ratified Ethernet as standard 802.3. That stamp of approval made it the default language of local networks.

The Cable Evolution: From Thick Yellow Hose to Cat5

If you’ve ever seen a 1980s computer lab, you might remember the thick, yellow coaxial cable snaking along the ceiling—they called it “ThickNet.” It was heavy, rigid, and a pain to install. If a cable broke, the entire network segment went down.

Then came 10BASE-T in 1990. This was a game-changer. It replaced the single shared cable with twisted-pair wiring (think phone wires) and a central hub. Now each computer had its own dedicated connection. If one machine went haywire, the rest stayed online. Cat3 cable gave way to Cat5, then Cat5e, and finally Cat6—each step driving faster speeds.

Speed milestones: - 10 Mbps (1980) - 100 Mbps (1995) — Fast Ethernet - 1 Gbps (1999) — Gigabit Ethernet - 10 Gbps (2006) - 40 Gbps and 100 Gbps (2010s) - 400 Gbps (2020s) — Used in massive data centers.

Why Ethernet Won the Real World

Let’s be honest: Wi-Fi is convenient, but Ethernet is the workhorse of the internet. Every time you stream a 4K movie or send a Slack message, chances are your data goes through at least one Ethernet connection in a data center. Here’s why it’s still king:

• Reliability: No dropped signals, no interference from your microwave.
• Low latency: Copper or fiber delivers consistent, sub-millisecond ping times.
• Power over Ethernet (PoE): The same cable can carry data and power to devices like security cameras or VoIP phones.
• Cost: A gigabit switch now costs less than a pizza.

The Not-So-Secret Legacy

Bob Metcalfe eventually left Xerox and founded 3Com—a company that built some of the first commercial Ethernet adapters. In 2022, he was awarded the Turing Award (the Nobel Prize of computing) for his invention. When he accepted the award, he joked about a famous prediction he made in 1995: “The internet will suffer a catastrophic collapse in 1996.” He was wrong. The internet didn’t collapse, and Ethernet grew stronger.

Today, Ethernet is everywhere: in your office, in your smart TV, in the airplane you flew on last week, and under the streets powering cell towers. It’s a 50-year-old technology that runs on 1970s logic with 2020s speeds.

And the best part? You can still plug an original 1980s Ethernet card into a modern router—and it will work. That’s not just backward compatibility. That’s a legacy written in copper and light.