Understanding IP addresses: the four numbers that decide whether your internet works
Last month a client called me in a mild panic. Their office printer had vanished from every computer on the network. Nobody had touched the printer. Nobody had touched the router. And yet, one morning, the thing that had printed invoices flawlessly for two years was simply… gone.
The culprit? An IP address conflict. Two devices had been handed the same numeric label, and the network — being a rigid, literal-minded beast — could no longer tell them apart. Ten minutes in the router's admin panel fixed it. But it reminded me why understanding IP addresses matters, even if you never plan to touch a command line.
An IP address is the reason your laptop can find a server in another country, the reason your smart doorbell knows where to send footage, and the reason things quietly break when two devices claim the same identity. It's not glamorous. But it's the plumbing of everything online.
Key Takeaways
- An IP address is a numerical label that identifies a device on a network — nothing more, nothing less.
- IPv4 uses 32 bits (four numbers from 0 to 255); IPv6 uses 128 bits and looks completely different.
- The subnet mask tells a device which part of an address is the network and which part is the device.
- Most home connections use dynamic IPs assigned by DHCP, which is why your address can change overnight.
- IPv1, IPv2, IPv3 and IPv5 were either never built or never released publicly — IPv4 jumped straight to IPv6 for a reason.
- Reading an IP address is a learnable skill, and you can practise on your own printer in about two minutes.
How do I understand IP addresses without a networking degree?
Strip away the jargon and an IP address is just a mailing label. Every device on a network needs one so that data knows where to go. When you type a website into your browser, your request doesn't travel as a vague wish — it travels as packets stamped with a source address and a destination address. Routers read those stamps and pass the packet along, hop by hop, until it lands.
That's the whole idea. The complexity comes from the bookkeeping.
How to read an IPv4 address
An IPv4 address looks like this:
192.168.1.42
Four numbers, separated by dots. Each number is called an octet, and each octet can range from 0 to 255. Why 255? Because each octet is 8 bits, and 2⁸ = 256 possible values (0 through 255). Four octets × 8 bits = 32 bits total. That gives roughly 4.3 billion possible addresses.
Which sounded like plenty in 1981. It wasn't.
Here's the practical part: when you see an address, you can usually guess what kind of device you're looking at, because certain ranges are reserved.
- 192.168.x.x and 10.x.x.x — private addresses. These only exist inside your home or office network.
- 127.0.0.1 — "localhost." Your own machine talking to itself.
- 169.254.x.x — the "I couldn't reach the DHCP server" address. If you see this on a device, something is broken.
I once spent an embarrassing hour troubleshooting a "network outage" that turned out to be a laptop stuck on a 169.254 address because someone had unplugged the wrong cable. The address told the story before any diagnostic tool did.
What are the four parts of an IP address?
Technically, the four parts are just the four octets. But there's a more useful way to think about it: an IPv4 address has a network portion and a host portion. Which is which depends entirely on the subnet mask — and that's where most beginners get lost.
Which brings us to the question everyone eventually asks.
What does a subnet mask of 255.255.255.0 actually mean?
Think of the subnet mask as a stencil laid over the IP address. Where the stencil has a 255, that part of the address identifies the network. Where it has a 0, that part identifies the device.
Take 192.168.1.42 with a mask of 255.255.255.0:
| Component | Value | Meaning |
|---|---|---|
| Network portion | 192.168.1 | Which network this device lives on |
| Host portion | 42 | Which device it is, within that network |
| Subnet mask | 255.255.255.0 | First three octets are network, last is host |
| Usable hosts | 254 | 256 minus the network address (.0) and broadcast (.255) |
So a mask of 255.255.255.0 means: everything on 192.168.1.x is on the same local network. Your router at 192.168.1.1, your laptop at 192.168.1.42, your printer at 192.168.1.87 — all neighbours. They can talk to each other directly, without any packet ever leaving the building.
Change that mask to 255.255.0.0 and suddenly 192.168.1.42 and 192.168.9.15 are on the same network too. Bigger neighbourhood. More devices that can find each other without help.
Can you explain subnetting to a beginner?
Subnetting is the act of chopping one big network into smaller ones. That's it. The reason you'd do it: to reduce traffic, isolate departments, or just keep things manageable.
Say you run a small office with 200 devices on a single flat network. Every time a computer sends a broadcast — "who has this printer?" — all 200 devices hear it. Wasteful. Split that into four subnets of 50 devices each, and broadcasts stay inside their own slice.
You express subnets with CIDR notation: a slash followed by the number of bits used for the network portion.
- /24 = 255.255.255.0 = 254 usable addresses
- /25 = 255.255.255.128 = 126 usable addresses
- /16 = 255.255.0.0 = 65,534 usable addresses
When I first learned this, I tried to memorise the whole chart. Bad idea. The trick is to remember that each step down the slash adds one bit to the network side, which halves the host side. /24 gives you 256 total, /25 gives you 128, /26 gives you 64. Once that clicked, I stopped needing the chart.
How do IP addresses actually get assigned?
Nobody hands out IP addresses by hand. There's a hierarchy.
At the top, regional internet registries allocate large blocks to internet service providers. Your ISP then carves those blocks up and hands addresses to customers. When your router connects, it usually asks for an address using DHCP — Dynamic Host Configuration Protocol — and gets one on loan.
That's why your home IP can change overnight. Restart your router at 3 a.m. and you may wake up with a different public address than you had yesterday. Most people never notice. Some people care very much — anyone running a home server, for instance, which is why static IPs exist as an option (usually for a small monthly fee from your ISP).
Inside your home, the router does something clever called NAT — Network Address Translation. All your devices get private addresses, and the router presents a single public address to the internet on everyone's behalf. That's why your laptop, phone, tablet and TV can all browse simultaneously without needing four separate public IPs.
Franchement, NAT is one of those inventions that quietly saved the internet. The IPv4 address pool would have run dry far earlier without it.
What happened to IPv1, IPv2, IPv3, and IPv5?
Short answer: almost nothing. They were never released for public use.
IPv1 through IPv3 existed as early design work from the 1970s, back when the project was still experimental and nobody knew whether it would survive. The version that actually got standardised and deployed was IPv4 — and by the time it went live, the earlier drafts were long obsolete.
IPv5 is the interesting one. It was assigned to an experimental protocol called Internet Stream Protocol, designed for streaming audio and video with quality guarantees. It was never rolled out at scale. So when engineers needed a successor to IPv4, they skipped straight to IPv6, partly to avoid confusion with the already-taken IPv5 slot.
So no, you didn't miss a chapter. The numbering just has gaps.
IPv4 vs IPv6: what actually changes for you
IPv6 addresses look alarming at first glance:
2001:0db8:85a3:0000:0000:8a2e:0370:7334
That's eight groups of four hexadecimal digits, separated by colons, giving 128 bits. The address space is so large that if you assigned an address to every grain of sand on Earth, you'd still have room left over.
| Feature | IPv4 | IPv6 |
|---|---|---|
| Bit length | 32 bits | 128 bits |
| Format | 192.168.1.42 | 2001:0db8:...:7334 |
| Total addresses | ~4.3 billion | Effectively unlimited |
| NAT required? | Commonly, yes | No |
| Header complexity | Simpler | More fields, better routing |
For most people, IPv6 is invisible. Your ISP may already be running it alongside IPv4 in what's called dual-stack mode. If you want to check, look at your router's status page — you'll often see both addresses listed.
The one practical headache: if you're configuring a static IP for a server and you mistype a single character in an IPv6 address, it won't work. IPv4 gives you a little more room for sloppiness. I learned that the hard way after forty minutes of staring at a firewall rule that turned out to have an extra zero in it.
How to check your own IP address in two minutes
Everyone should do this at least once. It turns an abstract concept into something concrete.
- On Windows, open Command Prompt and type ipconfig. Look for "IPv4 Address."
- On macOS or Linux, open Terminal and type ifconfig or ip addr. Same idea.
- On any device, open your router's admin page (usually 192.168.1.1 or 192.168.0.1) and check the connected-devices list.
What you'll see is your private address — the one your router knows about. Your public address is different, and you can find it by searching "what is my IP" or visiting your ISP's help page.
The printer example from the start of this article? That's where you'd go to fix it. Find the printer's address on the router page, check whether another device is using the same one, and either reassign or set a static reservation. Five minutes of poking, and the printer comes back.
Why knowing this pays off
You don't need to become a network engineer. But understanding IP addresses changes how you troubleshoot. When the internet dies, you'll know whether to look at your device, your router, or your ISP. When a device disappears from the network, you'll know where to look. When someone tells you to "just check the IP," you'll know what they mean.
The next time something stops working for no apparent reason, the answer is probably four numbers and a dot. And now you know how to read them.
