The Internet and the World Wide Web

"The Internet" and "the World Wide Web" are used interchangeably in everyday speech, but in computer science they are two distinct things — and conflating them is one of the most reliably penalised errors at A-Level. The Internet is the global infrastructure: the physical cables, routers and the TCP/IP protocols that let billions of devices exchange data. The World Wide Web is just one service that runs on top of that infrastructure — a system of interlinked pages fetched with HTTP. Email, video calls, online gaming and file transfer are also Internet services, but they are not the Web. Getting this relationship — service on top of infrastructure — clearly in your head is the foundation of the whole topic.

At A-Level you must distinguish the Internet from the Web; explain the Domain Name System (DNS) — its hierarchy and how a name is resolved to an IP address; dissect the structure of a URL; explain packet switching and contrast it with circuit switching; and describe the roles of routers, gateways, ISPs and the Internet backbone. We will trace a real DNS lookup and follow a packet across the network to make these concrete.

Spec Mapping

This lesson addresses the Communication and networking area of the AQA A-Level Computer Science (7517) specification. It covers the distinction between the Internet (infrastructure) and the World Wide Web (a service); the structure and operation of the Domain Name System, including its hierarchy and the resolution process; the components of a URL; packet switching versus circuit switching; and the roles of routers, gateways, ISPs and the backbone in carrying data. It builds on the TCP/IP protocol stack and IP addressing content (routing, IP addresses) and links to protocols in detail (DNS, HTTP) and web technologies.

The Internet — The Infrastructure

The Internet is a global network of interconnected networks: a physical fabric of fibre-optic and copper cables (including vast undersea cables), wireless and satellite links, and the routers and switches that join them — all communicating using standardised protocols, principally TCP/IP. Critically, the Internet is not one network; it is the agreed interconnection of thousands of independently run networks (ISPs, universities, governments, companies) that exchange traffic with one another.

Aspect	Detail
What it is	Global network infrastructure (hardware + the TCP/IP protocols)
Origin	Evolved from ARPANET (late 1960s)
Core protocols	TCP/IP
Services it carries	The Web, email, file transfer, VoIP, streaming, gaming, IoT, and more
Physical media	Fibre-optic (incl. undersea), copper, satellite, wireless

How a Device Reaches the Internet

A home device's data climbs a hierarchy of progressively larger networks to reach a distant server:

flowchart LR
    D["Your device"] --> HR["Home router<br/>(gateway to the ISP)"]
    HR --> ISP1["Your ISP"]
    ISP1 --> IXP["Backbone /<br/>Internet Exchange Points"]
    IXP --> ISP2["Destination's ISP"]
    ISP2 --> SRV["Destination server"]

Your device joins a local network (LAN) by Wi-Fi or Ethernet.
The LAN reaches the wider world through your ISP (Internet Service Provider) via a modem/router that acts as the gateway out of your network.
ISPs interconnect with one another at Internet Exchange Points (IXPs) and across high-capacity backbone links — the long-haul, high-bandwidth fibre that forms the Internet's core.
Data is then routed hop by hop across these networks to its destination.

The backbone and IXPs matter because they are why the Internet scales: rather than every ISP connecting directly to every other (impossible for thousands of networks), they meet at shared exchange points and over backbone links, keeping the global mesh manageable. Notice too that the data does not travel over one continuous wire from your device to the server; it is handed from network to network, each independently operated, with routers at the boundaries forwarding it onward — which is exactly the decentralised, no-single-owner character that defines the Internet. A useful mental model is a parcel passing through a chain of sorting depots: each depot only needs to know which depot to forward the parcel to next, not the whole route, and the parcel is re-handled at every boundary until it reaches its destination — precisely how routers forward packets hop by hop.

The World Wide Web — A Service on Top

The World Wide Web is a service (an application of the Internet) consisting of interlinked hypertext documents and resources, accessed with a web browser using the HTTP/HTTPS protocol. It was invented by Tim Berners-Lee at CERN in 1989 — decades after the Internet itself — which is the cleanest proof that the two are different: the Internet existed and carried email and file transfer long before the Web was conceived.

Aspect	Detail
What it is	A system of interlinked hypertext pages and resources
Invented	Tim Berners-Lee, 1989 (CERN)
Protocol	HTTP / HTTPS
Accessed via	A web browser
Content	HTML pages, images, video, scripts, APIs
Resources identified by	URLs (Uniform Resource Locators)

Component	Role
HTML	Markup language structuring web pages
HTTP/HTTPS	Protocol for requesting and delivering web resources
URL	The address identifying a specific resource
Web browser	Client software that renders pages (Chrome, Firefox, Safari, Edge)
Web server	Software hosting and serving pages (Apache, Nginx, IIS)
Hyperlinks	The links between pages — the literal "web"

The Crucial Distinction

	The Internet	The World Wide Web
What it is	Global network of networks (infrastructure)	A system of interlinked pages (a service)
Nature	Physical hardware + protocols	An application running on the Internet
Origin	ARPANET, late 1960s	Tim Berners-Lee, 1989 (CERN)
Protocol	TCP/IP	HTTP/HTTPS
Can exist without the other?	Yes — predates and underpins the Web	No — the Web needs the Internet to function
Scope	Carries many services	Only web pages and web resources

The key takeaway: the Web is one of many services that use the Internet. Email (SMTP/IMAP), file transfer (FTP), voice/video calls (VoIP) and online gaming all run over the same Internet infrastructure but are not part of the Web. So when you send an email or make a video call, you are using the Internet without using the Web at all. Stating this relationship — the Web is a service that runs on the Internet infrastructure — is exactly what an exam answer needs.

The Domain Name System (DNS)

Computers route data using numeric IP addresses (e.g. 93.184.216.34), but humans remember names (e.g. www.example.com). The Domain Name System (DNS) is the Internet's distributed directory that translates domain names into IP addresses. Without it, you would have to memorise IP addresses for every site — and worse, a site could not change its IP without breaking every link to it.

DNS is a distributed, hierarchical database — no single machine holds every name, which is precisely what lets it scale to the entire Internet and distribute administrative control. The hierarchy runs from the root, down to top-level domains (TLDs) like .com and .uk, down to the authoritative name servers that hold the actual records for an individual domain:

flowchart TB
    ROOT["Root name servers<br/>(.)"]
    TLD["TLD name servers<br/>(.com, .uk, .org ...)"]
    AUTH["Authoritative name servers<br/>(example.com)"]
    HOST["Host record<br/>www.example.com -> 93.184.216.34"]
    ROOT --> TLD --> AUTH --> HOST

Tracing a DNS Resolution

When you type www.example.com, here is the sequence by which it becomes an IP address. This recursive resolution is a staple exam walk-through:

sequenceDiagram
    participant B as Browser
    participant R as Recursive resolver (ISP)
    participant Root as Root server
    participant TLD as .com TLD server
    participant Auth as example.com authoritative server
    B->>B: Check local DNS cache first
    B->>R: Resolve www.example.com (if not cached)
    R->>Root: Where is .com?
    Root-->>R: Ask the .com TLD servers (referral)
    R->>TLD: Where is example.com?
    TLD-->>R: Ask example.com's authoritative server (referral)
    R->>Auth: What is the IP of www.example.com?
    Auth-->>R: 93.184.216.34 (the answer)
    R-->>B: 93.184.216.34 (and cache it for its TTL)
    B->>B: Open a connection to 93.184.216.34

Step by step:

The browser first checks its local cache (and the operating system's). If the name is cached, resolution stops here.
Otherwise it asks the recursive resolver (usually the ISP's), which does the legwork on the browser's behalf.
The resolver asks a root server, which refers it to the appropriate TLD server (here, .com).
The resolver asks the .com TLD server, which refers it to the authoritative server for example.com.
The resolver asks the authoritative server, which returns the actual IP address for www.example.com.
The resolver caches the answer (for a period set by the record's TTL, time-to-live) and returns it to the browser, which can now open a connection.

Why Caching Matters

DNS results are cached at many points — the browser, the operating system, the recursive resolver — for the duration of each record's TTL. Caching is essential: without it the root and TLD servers would be overwhelmed by the sheer volume of lookups, and every page load would incur the full multi-step resolution. The trade-off is propagation delay: when a domain's IP changes, old cached entries persist until their TTL expires, which is why DNS changes are said to "take time to propagate". DNS normally runs over UDP on port 53 (a small query/response where the occasional loss is cheaply retried), falling back to TCP for large responses.

DNS Record Types

DNS does more than map names to IPv4 addresses — different record types serve different purposes, and a domain's authoritative server holds a set of them:

Record	Purpose
A	Maps a domain name to an IPv4 address
AAAA	Maps a domain name to an IPv6 address
CNAME	An alias — points one domain name at another (e.g. `www` → the main domain)
MX	The mail exchange server for the domain — used by SMTP to deliver email
NS	The authoritative name servers for the domain

This is why a single domain can serve web pages and receive email at different machines: the A/AAAA records direct web traffic to the web server, while the MX record directs mail to the mail server. The MX record is a neat synoptic link to the email-protocols content — when an SMTP server relays a message, it performs a DNS lookup of the recipient domain's MX record to find where to deliver it. The distributed, hierarchical design also distributes administrative control: whoever operates example.com manages its own records and can add a new subdomain without asking any central authority — a key reason the naming system scales to hundreds of millions of domains.

URLs — Addressing Resources on the Web

A URL (Uniform Resource Locator) is the address of a specific resource on the Web. Dissecting one is a common exam task:

Example: https://www.example.com:443/products/shoes?colour=red#reviews

Part	Value	Purpose
Protocol (scheme)	`https`	How to fetch it — HTTPS is TLS-encrypted HTTP
Domain name	`www.example.com`	The host, resolved to an IP via DNS
Port	`443`	The TCP port on the server (default 443 for HTTPS, 80 for HTTP — usually omitted)
Path	`/products/shoes`	Which resource on the server
Query string	`?colour=red`	Parameters passed to the server (after `?`)
Fragment	`#reviews`	A section/anchor within the page (after `#`, handled client-side only)

Two subtle but examinable points: the domain name must be resolved by DNS to an IP before any connection is made — the URL names the host symbolically, not numerically; and the fragment (after #) is processed entirely in the browser and is never sent to the server, since it merely scrolls to a part of an already-fetched page.

A further distinction worth grasping is domain name vs IP address vs URL, three terms students often blur. The IP address is the numeric, routable address of the host (e.g. 93.184.216.34); the domain name is the human-friendly label (www.example.com) that DNS resolves to that IP; and the URL is the full address of a specific resource, wrapping the domain (or even a bare IP) together with the scheme, path, query and fragment. So the domain name is one part of a URL, and DNS is the bridge that turns that part into the address packets are actually routed to. This is also why a website can move to a completely different server — and hence a new IP — without any of its URLs changing: only the A record in DNS is updated, and the human-facing domain name and URLs stay the same. That indirection between name and address is one of the quiet but powerful design ideas of the Web.

Packet Switching

Almost all data on the Internet travels by packet switching. Rather than reserving a dedicated path, data is split into small packets that are routed independently and reassembled at the destination:

flowchart LR
    SRC["Source<br/>(splits data<br/>into packets)"] --> R1["Router"]
    SRC --> R2["Router"]
    R1 --> R3["Router"]
    R2 --> R3
    R1 --> R4["Router"]
    R3 --> DST["Destination<br/>(reassembles<br/>by sequence no.)"]
    R4 --> DST

The Internet and the World Wide Web (how they differ)