Free lesson

What Really Happens When You Open a Website

What Really Happens When You Open a Website — How the Internet Really Works: Its History from “LO” to ChatGPT. Start with the mechanism; A network of networks; From a site name to packets; Routers, hops, and delay; The internet is not the web; How to read traceroute. Practical assignment: Inspect the route of one request.

Demo edition. The materials are undergoing editorial review.

Topics covered

  • Start with the mechanism
  • A network of networks
  • From a site name to packets
  • Routers, hops, and delay
  • The internet is not the web
  • How to read traceroute

Start with the mechanism

The goal of this lesson is to explain, in plain language, the path from tapping a link to seeing a page. The internet is not an invisible “cloud.” It consists of many separate networks connected by cables, radio links, routers, and shared protocols.

Later lessons will examine who developed these parts and when. First, we need to see how they work together today. At the end, you will run traceroute or tracert and inspect part of the route taken by test packets from your device. The result will not be a complete map of the internet, but it can reveal real intermediate nodes and delays.

A network of networks

Your home Wi‑Fi, a mobile operator’s network, a university network, and a data-centre network are separate systems. The internet exists when such networks exchange traffic under shared rules. It has no single central computer or master switch.

The connection has a physical basis. Inside buildings, data may travel through copper or fibre-optic cables, or over a Wi‑Fi radio link. Over long distances, terrestrial and submarine fibre-optic lines do most of the work. Satellite access also exists, but it is a mistake to imagine that all international traffic “goes through space.”

Access is not universal. The ITU estimated that 6 billion people, or 74% of the world’s population, used the internet in 2025, while 2.2 billion remained offline. This is a dated estimate that the ITU may revise.

From a site name to packets

When you open an address, the browser first needs a network address for the relevant service. DNS helps obtain an answer for a domain name such as example.org. The answer may already be cached, so a full lookup does not happen every time. The browser then establishes a suitable connection, negotiates protected communication for HTTPS, and sends an HTTP request. A server or intermediate cache returns documents, style sheets, images, and other resources that the browser uses to construct the page.

Network data travels in pieces called packets. A common Ethernet MTU limit is 1,500 bytes for an IP packet, not 1,500 characters of page text; headers occupy some of that space, and other networks may use different limits. Protocols manage ordering, loss, and retransmission where their designs provide those functions.

Routers, hops, and delay

A router reads a packet’s destination address and uses a routing table to choose the next node. Movement between two such nodes is called a hop. Packets belonging to one exchange do not necessarily follow the same route every time, and the return path can differ from the outward path.

Fictional example. A laptop requests a news site. The first hop reaches the home router; later hops cross the provider’s network and other networks until they reach a server or a nearby CDN node. The server returns many packets, and the browser assembles the resources and draws the page.

A common mistake is to treat a large delay increase as proof of an ocean cable. It could also reflect a long terrestrial route, a queue, a route change, or the way one router handles diagnostic packets.

The internet is not the web

The internet is the infrastructure and protocol system through which networks carry data. The web is one service that runs over it: linked resources that browsers retrieve mainly with HTTP or HTTPS. Email, video calls, games, and messaging also use the internet, but that does not make them web pages.

Historical dates make the distinction concrete. Tim Berners-Lee’s proposal for the web appeared at CERN in March 1989; he produced the first browser-editor and server in 1990, and web software became available over the internet in August 1991. Inventing the web therefore did not mean inventing all the network infrastructure on which it operated.

How to read traceroute

traceroute sends a series of probes with progressively larger TTL values, while Windows tracert performs a related test. Each router decreases the TTL; when it expires, the router will usually return an ICMP message. This lets the tool attempt to identify successive hops and response times.

On Windows, open Command Prompt and run tracert example.org. On macOS or Linux, try traceroute example.org. If the command is unavailable, do not install something solely for this task; use an existing network utility or a trustworthy web service. A web service traces from its server, not from your phone.

* * * means that no reply arrived for those particular probes, not necessarily that something is broken. A router may filter ICMP, rate-limit replies, or give diagnostic traffic low priority. Hostnames and timing jumps are clues, but they do not prove a physical location or a particular cable.

Practical assignment

Inspect the route of one request

Continue to learning

Course outline

  1. What Really Happens When You Open a Website
  2. 1969: The Two Letters That Started It All
  3. Computers: From a Room to Your Pocket
  4. A Common Language: TCP/IP, Email and Names
  5. A Computer on Every Desk
  6. The World Wide Web: A Gift to the World
  7. The Dot-Com Bubble: When Everything Was ".com"
  8. The Open Internet: Linux, Wikipedia and Shared Knowledge
  9. The Internet in Your Pocket
  10. Social Networks and the Feed: How Attention Became the Product
  11. Who Runs the Internet — and Why It Ran Out of Addresses
  12. AI and What Comes Next: How to Read Predictions