What Is Communication, Really - And Why a Physical Connection Isn't Enough

📚 How Computers Talk - Part 1 Communication #Communication Fundamentals
Table of Contents

What Is Communication, Really - And Why a Physical Connection Isn’t Enough

Intuitively, it’s easy to think of communication as something technical and simple: there’s a sender, a receiver, and a connection between them. If you connect two computers with a cable - it seems like the problem is solved.

But in practice, that’s only the most superficial layer. A physical connection allows signals to pass, not meaning.

Two computers can be connected to each other, and still not exchange any useful information at all.

A Signal Is Not Information

At the lowest level, computers know how to send and receive signals - electrical changes, pulses, sequences of zeros and ones.

But a signal, by itself, is not information. For a signal to become information, both sides need to agree in advance on some very basic things:

  • What counts as a unit of information?
  • Where does it start and where does it end?
  • What does each part mean?
  • What do you do when something goes wrong?

Without such agreements, the sequence that arrives on the other side is just noise.

Communication Is Coordinating Expectations

In the deeper sense, communication isn’t transmission - it’s coordination. Not “I sent something,” but “we sent and understood the same thing.”

A computer sending data always makes assumptions:

  • At what rate the data will arrive
  • In what order
  • Whether it’s okay to lose parts
  • Whether it needs to confirm the other side received it

If the receiving side doesn’t share those same assumptions - communication fails, even if the connection is perfect.

The First Problem of Distributed Systems

Here a foundational principle emerges that will accompany the entire series: communication fails not because of a lack of connection - but because of a lack of agreement.

The failure isn’t necessarily dramatic:

  • Data arrives late
  • Requests are answered partially
  • Results look “strange”
  • The system doesn’t break, but also doesn’t behave as expected

And that’s the most dangerous kind of failure - a silent failure.

An Analogy

Imagine two offices connected by a pipe for passing documents. The pipe works perfectly. But one office sends documents without a title, without a page count, and without an agreed-upon order.

The other office receives a pile of papers - but doesn’t know what to do with them.

The problem isn’t the pipe. The problem is the absence of rules.

The Key Point

Once you understand that communication is, first and foremost, an agreement, it’s easy to understand why you can’t “just send data.”

You need to define:

  • How to send
  • In what structure
  • What’s expected from the other side
  • How to handle uncertainty

None of these are minor technical details - they’re the foundation on which protocols are built.

In the next post we’ll dive into that exact concept: what a protocol is - and why giving up freedom is a condition for reliable communication.

Comments