What Is a Computer on a Network - And Why an Address Isn't a Location
What Is a Computer on a Network - And Why an Address Isn’t a Location
After understanding why communication is built from layers, we can ask another basic question: for a computer to talk to another computer - how does it even know who it’s talking to?
The natural intuition is to think in terms of location: “the computer over there,” “this server,” “that machine.”
But on a network, physical location is almost irrelevant. What matters is identity.
A Computer on a Network Is a Logical Entity
A computer on a network isn’t “a box in a specific place.” It’s a logical endpoint that can:
- Send information
- Receive information
- Respond according to agreed-upon rules
It can move, change, be replaced - and communication will still keep working, as long as the identity is preserved.
That’s why a computer can:
- Move to a different server
- Move to a different network
- Or even be completely replaced
and still be perceived as “the same” party in the communication.
What an Address Really Is
An address on a network doesn’t describe a physical location. It’s a label.
Its job is simple: to let information reach the right destination inside a complex system.
The address says: “here is the party the message is intended for” - not “where the computer is located in the world.”
This is a critical distinction: in large networks, the ability to decouple identity from location is a prerequisite for scale.
Why You Can’t Rely on Location
Modern systems are constantly changing:
- Computers come and go
- Load shifts
- Infrastructure gets replaced
If communication depended on a fixed physical location, every such change would break the system.
Instead, the network treats computers as replaceable entities, and relies on logical addresses to route information.
An Analogy
Think of an internal mail system in a large organization.
An envelope isn’t sent based on the employee’s physical location, but based on their name or role.
The employee can move rooms, floors, or buildings - and the mail will still reach them, as long as the routing system knows how to update the link between the name and the location.
That’s exactly how a network works.
An Address as a Promise, Not a Commitment
An address doesn’t guarantee:
- That the computer is available
- That the response will be fast
- Or that the content will be received
It only guarantees one thing: that there’s a way to try to reach the destination.
Everything else - reliability, order, response times - is handled by other layers.
Looking Ahead
Once you understand that a computer on a network is a logical identity, the next stage emerges: how information finds its way among many identities, and why it’s almost never sent as a single unit.
In the next post we’ll cover the Packet - the fundamental unit of modern communication.
📚 More in this Series: How Computers Talk
- Part 1 What Is Communication, Really - And Why a Physical Connection Isn't Enough
- Part 2 What Is a Protocol - And Why Absolute Freedom Creates Chaos
- Part 3 Why Do We Need Layers in Communication - And Why a "One Solution for Everything" Fails
- Part 5 What Is a Packet - And Why Information Isn't Sent as a Single Unit
- Part 6 TCP vs UDP - Reliability or Speed
- Part 7 What Is a Connection - And Why It's a Logical State, Not a Cable
- Part 8 Latency, Bandwidth, and Throughput - And Why Everyone Confuses Them
- Part 9 Why Queues Are the Hidden Heart of Communication
- Part 10 HTTP - Why It Looks Simple, But Is Far From It
- Part 11 Why Protocols Evolve - And Are Never Fully Replaced
- Part 12 Communication as a Mirror for Systems Thinking
- Part 13 The Full Picture - How All the Pieces of Communication Connect Into One Living System