Where Do You Live? IP Address: IPv4 and IPv6

Why do we need an address when we already have a name tag?

In the last lesson you learned that a MAC Address is only useful inside one LAN (Local Area Network). It is like a name tag: you know the person's name, but not which district or street they live on.

Imagine postal workers all over the world holding billions of names but no addresses. They could never deliver a single letter. The IP Address solves this problem. The address itself says which network a device is on, so a Router only has to read it to know which direction to send the data next.

	MAC Address	IP Address
Analogy	A name tag sewn on	A postal address
Question it answers	Who are you?	Where are you?
Who decides it	Set by the manufacturer at the factory (or generated randomly by the system)	Assigned by the network you are on
When you change network	Usually stays the same	Changes
Where it is useful	Within one LAN	Across networks, all the way to the other side of the Internet
Length	48 bits	IPv4 is 32 bits; IPv6 is 128 bits

The two work as a team. On each LAN segment, data is delivered to the next stop using the MAC Address. Over the whole journey, the IP Address decides the direction.

IPv4: an address made of four numbers

The most common IP Address today is IPv4. It is 32 bits long and is written as four decimal numbers separated by dots. For example, Mac's address at home:

192.168.1.10

• Each number stands for 8 bits, so it can only be 0 to 255.
• Four groups × 8 bits = 32 bits.
• 32 bits can make 4,294,967,296 addresses (about 4.3 billion).

This way of writing is called dotted decimal. 192.168.1.300 is wrong, because 300 is bigger than 255.

Two parts of an address: the network part and the host part

An IP Address can be split into two parts, just like a postal address has an "area and street" and a "door number":

Part	Meaning	Example in 192.168.1.10/24
Network part	Which network you are on	192.168.1
Host part	Which device on that network	10

The /24 after the address means the first 24 bits (the first three numbers) are the network part. All addresses that share the same network part belong to the same Subnet. A Subnet is a smaller piece of a bigger network, like a city divided into streets.

Some devices don't show /24. Instead they show a Subnet Mask: 255.255.255.0. A Subnet Mask is just another way of saying which bits belong to the network part. 255.255.255.0 and /24 mean exactly the same thing.

Diagram: The top half shows IPv4, four groups of 8 bits each, where /24 means the first three groups are the network part. The bottom half shows IPv6, eight groups of 16 bits each, and how it is shortened to 2001:db8::1.

Same Subnet or not: do we still need a Router?

Before sending data, a device compares its own network part with the other device's network part:

• Same → the other device is in the same Subnet, so the data is delivered directly inside the LAN (using the MAC Address from the last lesson).
• Different → the other device is on another network, so the data is handed to a Router first, and the Router forwards it.

The Router that a device hands its "outgoing mail" to is called the Default Gateway. The Default Gateway is your network's door to the outside world. At home, it is usually your home Router.

Notice that a Router stands between two networks, so it has two IP Addresses. Take Mac's home as an example:

• Inside (the home LAN): Private IP 192.168.1.1. This is the Default Gateway for the devices at home.
• Outside (the Internet): Public IP 203.0.113.1, assigned by the ISP (Internet Service Provider).

Example (Mac at home, /24):

My address	The other address	Same network part?	How it is sent
192.168.1.10	192.168.1.77 (the TV at home)	Same (192.168.1)	Delivered directly inside the LAN
192.168.1.10	198.51.100.1 (the coffee shop's Public IP)	Different	Handed to the Default Gateway 192.168.1.1

In a /24 Subnet, the host part runs from 0 to 255. But .0 stands for the whole network, and .255 is the Subnet's Broadcast Address, so only 254 addresses can actually be given to devices.

Public IP and Private IP

IPv4 addresses come in two kinds: one kind is used on the Internet, and the other is used only inside a home, school or company.

	Public IP	Private IP
Analogy	The street address of a building	The flat number inside the building
Where it is used	On the Internet	Inside the network of a home, school or company
Who assigns it	The ISP	Your Router
Is it unique?	Unique on the Internet	Different homes can reuse the same ones
Can it be reached directly from the Internet?	Yes	No

Private IPs can only come from the three ranges below. Routers on the Internet will not forward data addressed to these, so they never appear on the Internet:

Range	Commonly found in
10.0.0.0 to 10.255.255.255	Large companies, schools
172.16.0.0 to 172.31.255.255	Company networks
192.168.0.0 to 192.168.255.255	Home Routers, such as 192.168.1.x

Most phones and computers at home use Private IPs. When the whole household goes online, they share one Public IP on the Router. This sharing trick is called NAT (Network Address Translation), and Lesson 8 explains it in more detail.

Diagram: Mac's home and the neighbour's home both contain 192.168.1.10 and 192.168.1.11, and even the Router's inside address is the same, 192.168.1.1. There is no confusion, because a Private IP only counts inside its own home. When going online, the two homes use their Routers' outside Public IPs, 203.0.113.1 and 203.0.113.2.

How can you tell a Public IP from a Private IP at a glance?

Look at the start of the IPv4 address:

• Starts with 10. → Private IP
• Starts with 172. and the second number is 16 to 31 → Private IP
• Starts with 192.168. → Private IP
• Any other ordinary address → Public IP

Chart: Private IPs have only three ranges, and every other ordinary address is a Public IP. Be careful: an address starting with 192 or 172 is not always Private, so check the numbers that follow.

Here is every address used in this lesson, sorted out:

Address	Whose address	Type
192.168.1.10	Mac's computer at home	Private IP
192.168.1.1	The home Router (inside)	Private IP
203.0.113.1	The home Router (outside)	Public IP
192.168.0.23	Mac's computer at the coffee shop	Private IP
192.168.0.1	The coffee shop Router (inside)	Private IP
198.51.100.1	The coffee shop Router (outside)	Public IP

So in the comic, when Mac changes networks, his Private IP changes from 192.168.1.10 to 192.168.0.23. What everyone else on the Internet sees changes from the home's 203.0.113.1 to the coffee shop's 198.51.100.1.

Why do we have Private IPs? The story of a "temporary" fix

4.3 billion IPv4 addresses sounds like a lot, but the world's phones, computers and TVs added together passed that number long ago. Engineers saw the problem coming very early:

Year	What happened
1981	IPv4 became an official standard. There were very few computers in the world then, so 4.3 billion addresses looked endless.
Early 1990s	The Internet grew fast, and engineers realised IPv4 addresses would run out sooner or later.
1994	Engineers proposed NAT so that many devices could share one Public IP. Its authors said at the time that it was only a short-term fix.
1996	The three Private IP ranges were officially set, so any home or organisation could reuse them freely inside its own network.
1998	IPv6, the long-term solution, was officially published.
February 2011	The international organisation in charge of global address allocation handed out the last batch of IPv4 addresses to the regions.
April 2011	The organisation in charge of the Asia-Pacific region (including Hong Kong) was also down to its last batch and began to ration addresses strictly.

With Private IPs and NAT, a home, a school or even a whole company may need only one Public IP to get online. This slowed the running-out of IPv4 addresses a great deal, which is why IPv4 lasted until 2011 before the supply was fully handed out.

Interestingly, this "temporary" method has now been in use for thirty years, and it is still used in every home. It does have a downside, though: devices at home hide behind the Router, so other people on the Internet cannot reach them directly (Lesson 8 covers this). That is why, in the long run, everyone is slowly moving to IPv6, which has more addresses than we could ever use.

Example addresses in this lesson The Public IP examples in this lesson use addresses reserved for documentation (203.0.113.x, 198.51.100.x, 2001:db8::). The Private IP examples use the real private range 192.168.x.x, because private addresses never appear on the Internet and so cannot point to anyone's real device. This way, the type of every example matches the real world.

IPv6: a longer address

To solve the shortage of addresses once and for all, engineers designed IPv6. It is 128 bits long and is written as eight groups of four hexadecimal digits, with a ":" between groups:

2001:0db8:0000:0000:0000:0000:0000:0001

128 bits can make about 3.4 × 10³⁸ addresses, which is 34 followed by 37 zeros. That is enough to give every grain of sand on Earth many addresses.

An address this long is hard to write, so there are two shortening rules:

1. Leading zeros in each group can be left out: 0db8 → db8, 0000 → 0, 0001 → 1.
2. A run of consecutive all-zero groups can be replaced by "::", but only once in the whole address.

2001:0db8:0000:0000:0000:0000:0000:0001 → 2001:db8:0:0:0:0:0:1 → 2001:db8::1

Why can "::" be used only once? If it appeared twice, there would be no way to tell how many zero groups each "::" stands for. The letters in IPv6 addresses are usually written in lowercase.

Today, many mobile networks and home broadband services support both IPv4 and IPv6, so your phone probably has both kinds of address.

Where do addresses come from?

You normally don't have to set an IP Address on your device yourself. When you connect to Wi-Fi, the Router hands you an address automatically, like a hotel front desk handing out room keys. The protocol that does this is DHCP (Dynamic Host Configuration Protocol), which Lesson 8 explains in detail.