Computers

The Internet

There’s a lot of misunderstanding about the Internet, and, to be fair, not many people care where it came from anyway. It’s pervasive. It affects just about everything we do every day, to the extent that we don’t really think about it much.

But it has a history, and bits of it are interesting.

It started in the early 1960s, during the Cold War. The US military, along with most other countries, fully expected to have to fight World War III at some point. Along with the weapons, was a growing realisation that the various pieces of the military needed to exchange information reliably, and the computer industry was growing to support the information flow.

Computer networks at the time were very different. There were a dozen main manufacturers, each of which has a unique networking protocol. So, if an IBM computer had to talk to one from Honeywell, there had to be specific software to translate between the two. If your mainframe needed to talk to one from DEC or Data General, two more specific pieces of software had to be created. You can imagine that with a dozen manufacturers, the whole situation got complex quickly. And complexity in software is definitely smething to avoid if you can !

So in 1963 an organisation called ARPA (stands for ‘Advanced Research Products Agency’) issued a requirement for a new network. It had to be more general – ‘neutral’ was a word that was used, and it had to be more efficient. Rather than have single purpose links between two computers, it had to be more like the phone system. The data needs to know where is come from, and where it need to go, and the network works out how to get there.

The problem with networks back then, was that the links were point to point. For example, a computer in London would have a single connection to a computer in Manchester that it needed to share information with, and a second dedicated link to Bath, a third to Paris, and so on. Not only is that inefficient, from a military perspective, it’s vulnerable. Break one of those links and two sites can no longer communicate.

US Universities responded to the ARPA document and by 1969 had created the first packed switched network. With hindsight, it’s fairly simple, but at the time it was radical and completely new.

It works like this. If I want to send a long email from Australia to the UK, my computer will break the messages up into chunks, attach my address and the destination address to each chunk, add a small amount of management information and send it out into the Internet. Each chunk is called a ‘packet’, and they all have to arrive at the destination and be assembled in the right order before my email is complete.

The Internet is actually a huge number of smaller individual networks which are joined together and can share data. Bigger companies lay long distance fibre optic networks under the sea and over land to connect countries and cities. Smaller companies either create their own local networks or lease parts of bigger networks. It’s these interconnections that make diagrams of the Internet so like diagrams of the human brain – bewlideringly complex.

But it can still be understood. Think of it as two things working together – hardware and software. The hardware is the network (phone lines, optical fibres and wireless links), the computer systems that move data from one place to another (routers and servers) and the software provides the services to make it all work.

To take one small example, NASA have a web site called ‘www.nasa.gov’. If you type that into your browser, the front page of their web site will appear. The name ‘www.nasa.gov’ might seem a bit arcane, but at least it’s something humans can understand. NASA is the organisation name, the ‘.gov’ on the end means that it’s part of the US government and the ‘www.’ on the front means that it’s a web site.

But computers don’t find that name particularly useful, so they convert it into a number, and it’s the number that controls how the request from your computer reaches NASA and allows them to send you the web page. For NASA, the number is ‘203.77.190.71’. Unless you spend your life working with computer networks, that number doesn’t mean anything which is why the human-readable names exist. There’s software to convert between the two of course, and it’s software that you don’t need to care about. Your ISP will take care of the details on your behalf.

This particular software is called DNS – Domain Name System, and it’s probably the biggest distributed database in the world. Distributed because it’s spread out over tens of thousands of computers throughout the world, each of which has a small amount of the total infomation needed to convert between names and addresses. But each also knows where to go to convert names or addresses that it knows nothing about.

DNS is one of the fundamental services that makes up the Internet. Another is email, and there are dozens more, many of which operate in the background making your experience as smooth as it can be. It’s not always as smooth as we’d like of course, but it’s a start, and getting better all the time.