The child years of the internet
Part 2: The internet series (part 1)
The moon landing was an event the whole world watched. In many ways, it regarded the end of the era of the space race between the Soviets and the Americans. The American flag was planted. The astronauts returned safely home, but the impact on everyday life for people on earth was very limited.
On the other hand, the birth of the internet a few months after the moon landing, quietly marked the beginning of a revolution that impacts almost every life on earth. But, its revolution didn’t happen overnight.
The internet is now everywhere, but it’s already over 50 years old. What happened during the child years of the internet?
ARPNET first years
ARPANET created a network of networks. From there on, the precursor of the internet starts to grow. In 1971, it begins with around 14 nodes in operation:
In this year, The Network Working Group completes the Telnet protocol and makes progress on the File Transfer Protocol (FTP) standard. These protocols are still in use today.
Two years later, ARPANET already contains thirty nodes:
In 1973 another important protocol was established, the Transmission Control Protocol (TCP). This would standardize ARPANET into the internet and sparks enormous growth.
Today, we often think about the internet in terms of shopping on Amazon or watching series on Netflix. But internet surfing is just the surface of what the internet is. The internet itself is more about data. When we surf the web, check emails, play a video game, download music or use online streaming sites data is sent in packages between two or more computers.
Technically, every computer connected to the internet has an address. This IP address (Internet Protocol address) is a number that is used as identification. Sending data over the internet goes like this:
But in the real world, data doesn’t travel that straightforward. The internet was explicitly designed for failure from the start. But how does this mechanism work? A small analogy:
Say, I am living in Oakland and I want to move to New York. For this, I need to move my stuff first. I order two moving vans that will do the job.
My stuff is split into two parts. Every van enters the new address in its navigation and starts driving. The first one drives through Austin and because of a traffic incident the second one drives through Tampa. For me, it doesn’t matter, at the end my stuff is brought together in my new apartment.
The internet works similar. When sending a file, this file is chunked into various data packets. These packets all travel to the destination address. Say the data packages are going to a destination in New York, it can travel to an internet hub in Austin:
However, other packages can travel a whole other route:
Network of networks
Another important concept on the internet inner workings is a “network of networks”. Take for example Oakland which has its own network of computers:
But Austin has another network. The data must move from one network to the other network:
Conceptually, multiple networks where data is travelling in a redundant way is a simple concept. Travelling by car from one place to another is also a simple concept, but we all know it only works because we have a standardized infrastructure and traffic regulations. For the internet, these set of rules are set in the TCP/IP Protocol.
TCP/IP (Transmission Control Protocol / Internet Address) are the basics which make location and transmission of data possible. Other protocols like HTTP (Hypertext Transfer Protocol) are built on top of this.
Basically, TCP/IP is a set of layers:
When the data travels, it goes through each layer:
When you send a WhatsApp message, this message is broken up into multiple data packages. These packages go through the various layers of the internet and the receiver will receive the data packages which are turned into a message again.
From ARPANET to the Internet
in 1974 the first specification for the Transmission Control Protocol (TCP) was published. In 1975 the ARPANET grew towards 61 nodes:
Meanwhile, Berkeley adds the TCP/IP protocol support to the Unix operating system. From then, these Unix servers became important nodes on the internet.
Like traffic regulations, just a protocol wasn’t enough. Cars need roads, and the internet needs cables. Big steps were taken by Robert Metcalfe working at Xerox. He developed a system using cables that allows for transferring more data packages. This is known as the Ethernet.
Metcalfe started his own company, 3com that developed Ethernet further to create local area networks (LANs). This allows workstations to form a network.
In January, the ARPANET standardizes on the TCP/IP protocols adopted by the Department of Defense (DOD).
Another big step was the introduction of the Domain Name system. In 1985, the first domain was registered: symbolics.com, a domain belonging to a computer manufacturer.
Between the beginning of 1986 and the end of 1987 the number of networks grow from 2,000 to nearly 30,000. This was the time that internet not only had academic institutions, but other organizations like NASA, the Energy department and some commercial businesses.
The term “internet” was already used in the first RFC published on the TCP protocol in 1974. However, the transition from ARPANET towards the internet took a lot longer. ARPANET evolved into NSFNet. And in Europe a separate TCP/IP based network was created by CERN, CERNET.
It took until the year 1988 that the first international connections to NSFNET was established by France’s INRIA and Piet Beertema at the Centrum Wiskunde & Informatica (CWI) in the Netherlands. The internet has come of age.