Back Down the Rabbit Hole

Okay, so in case anyone is as forgetful as I am, here’s a bit of a recap. Basically, people were getting paranoid about their communication systems during the Cold War, so a man named J.C.R. Licklider developed the ARPAnet, which was supposed to serve as a small network of computers in order to keep the communication lines active in case of emergencies.

Apparently this ran into some issues with actually translating and transferring the messages from computer to computer–kind of like how I had troubles communicating with Bowser earlier. But, unlike me and my braindead dog, ARPAnet had a team full of scientists on their side to come up with solutions.

It was an MIT scientist named Leonard Kleinrock who finally figured it out in 1965 with the development of packet switching.

Packet Switching

Alright, if you’re thinking of the kind of packet switching you do on April Fool’s Day when you switch out someone’s shampoo for temporary hair dye, you’ve got the wrong idea–not that I entirely disapprove of that idea.

Before we get into packet switching though, let’s go back even further before 1965 to understand just how great an advancement packet switching was back then. No, I’m not talking carrier pigeons–that’s a bit too far back. I’m talking circuit switching.

Both circuit switching and packet switching were the processes through which data was transferred between devices. With circuit switching, the two communicating devices, or computers–we all know there were no smartphones or iPods back then–had to be connected through a direct, dedicated cable that would carry the data from one computer to the other. That data had to be sent all at once, and if there was another computer trying to communicate, all of that data would be lost. This usually led to lots of errors and lots of impatience and frustration.

It’s like when you forget to hit save at any point in typing up a large presentation on an old computer; your computer crashes just as you’re finishing up, and all your hard work is gone just because your computer couldn’t focus on so much at once.

Luckily, similar to the development of the autosave for most programs, packet switching fixed that issue. It couldn’t go so far as storing the data, but data was broken into chunks and sent in cute, little packets between computers. Those packets of data would travel along a web of wires that would connect at various nodes, so multiple data strings could be sent through the wires at once allowing several computers to communicate on the same lines.

More computer nerd jargon, huh? Think of it as sending out a book across that galactic web in the sky we talked about earlier. But instead of trying to shove the whole book through in one huge transmission, you take out the binding and send all the pages individually, as packets, because it’s easier to digest as well as deliver. Those pages go through a few switches before reaching their final destination, but if one goes missing or gets damaged, it’s easier to send just the single page and not the whole book all over again. Similar to only sending a page at a time, only small, basic packages were allowed when packet switching began in 1965.

Just think about that big web in the sky connecting all communicating devices all across the globe. With the development of packet switching, their little scientist spiders were just barely starting to weave that web, even if it did have to start out as a single, clunky thread.

Anyways, in 1968, an MIT researcher named Lawrence G. Roberts decided to make packet switching part of the ARPAnet and introduced specific package switches known as Interface Message Processors (IMPs), the very first generation of today’s routers, that would finally allow ARPAnet to be put into action.

You know how when you get a new car or a neat upgrade on your car, you instantly need to go test it out before the excitement overflows? That’s how the scientists of ARPAnet felt about this new improvement for their galactic network, and boy, did they go all the way on their test drive – but that will have to wait until our next dive down this rabbit hole. As we all know, curiosity is better taken in doses.

Next in the series comes the test driving of the ARPAnet and host-to-host protocols and applications like email, but we will discuss those intriguing developments in history in the next installment of the Evolution of the Internet.