What is the Internet?

The Internet is a world-spanning network of computers. A variety of wired and wireless technologies all working on concert with each other to build a system where information can be shared between any number of machines separated by globe-spanning distances. There are even internet-connected devices in outer space. Billions of people use the Internet every day to find news, recipes, cat pictures, and election results. We watch videos on YouTube to learn how to fix our cars. We order products from our phones and they are delivered to our homes in a few days.

So what is it really? It’s hard to talk about what the Internet is, without talking about how it works.

How does the Internet Work?

Vocabulary:

Protocol: A set of rules and conventions that allows all machines to talk to each other. It can, in a way, be thought of as a common language for two or more computers.
TCP/IP: The most important protocols on the Internet. They are not the only ones, but since the beginning and still today, these two protocols mediate most communication on the Internet.
Packet Switching: Breaking a message into small pieces and sending each piece independently across a network over whichever route is most efficient at the time.
Packet: A single message sent across the internet. It may contain the whole message, or part of a much larger message.
Domain: An easy-to-remember name for a web service that can be translated into an IP address.

The Internet is based on the principle of packet switching. This is the idea that we can reliably transmit any digital file over long distances by breaking up larger files into many smaller chunks at the sending end, and re-assemble them on the other end. It has been through a variety of incarnations over the years, but eventually, the suite of protocols that became the defacto language of the internet was the TCP/IP stack.

A brief overview of how this all works in practice (from a very high level of abstraction) is discussed on this page.

The Stack

Application Layer
TCP
IP
Hardware Layer

Essentially, the stack works like this:

An application (a Web browser, for example), decides it needs some resource from the Internet. It asks the computer to send a message to a computer somewhere else on the web.
Next, the Transmission Control Protocol (TCP), takes the information the application is going to send and breaks it up into pieces (if necessary) and takes each piece and bundles it into a string of binary called a packet.
Next, it sends it on to the next layer, the Internet Protocol library. The IP layer stamps the packet with an address and sends the packet to the next leg of its journey.
The system hardware interface talks to your network hardware and encodes the packet into an electrical which can be sent on a wire.
From there, it winds its way across the Internet looking for its destination. Each packet is passed along from stop to stop like a hot potato and at each stop a special piece of hardware called a switch decides where to send it on to next.
When it reaches the destination computer, it climbs back up the stack. The network hardware deserializes the signal (puts it back into its packet form), the IP layer verifies that this is where it belongs, the TCP layer collects all of the packets and reassembles them if necessary.
Lastly, and finally, whatever application (possibly some web server sofware) looks at the request and decides how to respond to it. For example, it might serve up a Webpage and send it back down through the stack and across the internet back to your computer to be rendered in your web browser.

One interesting thing to note is that on modern computers, this process happens very, very quickly (on the order of seconds), but in computer time, it’s an incredibly slow and laborious process.

Infrastructure

Briefly, it’s useful to understand the Internet in terms of network hardware. There’s a sort of hierarchy of hardware on the internet and each level of that hierarchy, a different level of service exists and a different customer, business, peer relationship exists. For example, you may purchase access to the internet from a local Internet Service Provider. They might purchase their access to the Internet through a regional ISP, who in tern will be purchasing network access from what’s called a Network Service Provider. These are massive telecom companies and the huge bundles of fiber-optic cabling that stretch all the way around the world are collectively often referred to as the Internet Backbone. Each NSP is legally required to connect to at least three Network Access Points, where packets can jump from one network to another, thus ensuring that the network is really a network and that all computers are connected to all other computers, however circuitiously.

Domain Routing

Routing on the internet relies heavily on the Domain Name System. Since the Internet actually relies on IP addresses, which are just long strings of numbers like 192.168.1.1, which are hard to remember, we have a system of Domain Names, which are much easier to remember, like google.com. Each time you look for a particular server, its domain name is resolved into an IP address by looking it up in a very big table, something like a phone book (remember those?).

This page of this book is located at https://itech.erickuha.com/book/ch6-internet/what_is_internet.html. But what does that mean? Let’s take the address apart and examine its parts.

https:// - is the protocol. In particular, http is the protocol of the World Wide Web, called hypertext transfer protocol. It’s a collection of commands that can be used to retrieve, update, or delete web pages. The :// part is just a some junk, called a delimiter, that the earlier designers decided to put there to separate the protocol from the rest of the address. You don’t actually have to type the protocol or the delimiter anymore as most browsers will insert it automatically and for free!
com - The top-level domain of a website is where, organizationally, it exists. There used to be only a few of these, but now there are are many. How many can you think of? Their main purpose is to make table-lookups faster by only searching part of the table (the part with all of the .com addresses, for example). List of available top-level domains. Pricing on Google Domains.
erickuha - is my domain, the domain that I registered through a domain registrar and have exclusive rights to. It is, somewhere along the line, associated with an IP address, but I do not know what it is.
itech - The sub-domain is the specific part of my domain which can be routed differently than the rest. For example, the root erickuha.com is hosted in one location and itech.erickuha.com is routed to a completely different service, server, and part of the internet.
/book/ch6-internet/what-is-internet.html is the called a path. It leads to the specific file on the server that is requested when someone navigates to this page. It works almost identically to the file hierarchy on your own computer’s file system. A system of forward slashes separates folders/directories and file names with file extensions are used to denote final endpoints.

There are other schemes for locating resources on the internet, but this is the sanest and simplest one and can be used as a model for understanding the rest of it on down the line.

Other thoughts

There is a lot more to talk about and a lot more to learn. You can look into all of the different protocols that are used on the Internet and the World Wide Web, or you can look deeper into how packet-switching is actually implemented in the TCP/IP stack. This is a jumping-off point to get you started down the road to understanding more about how the largest piece of technology ever created actually works.

Next, let’s look at some practical applications of all of this.