The Inquirer-Home

Museum unearths world wide web origins

How British boffins invented the Internet
Wed Dec 09 2009, 14:08


THE WORLD'S FIRST museum gallery showcasing the technology of the Internet opened recently at The National Museum of Computing (TNMOC) and the interactive exhibit tells the story of the pioneering British boffins without whom the World Wide Web could have been a very different place.

The gallery was opened on December 4th in the presence of the world's press and as lively a gang of OAPs as your ever likely to encounter, many of whom were former employees of the National Physical Laboratory (NPL), one of the UK's leading science and research centres, and the birthplace of packet switching, the technology that underpins just about every element of the Internet.

The gallery traces the time line from the first telegraph links right up to todays ultra high speed broadband using audio-visual as well as traditional displays and real time mapping of internet connections to the Museum's own Website.

time-line But it's the fascinating and mostly undiscovered history of the invention of packet switching by a team of British boffins working at NPL which will be of most interest to visiting geeks.

In the 1960s, most computers consisted of a huge remote mainframe connected to a number of screen and keyboard terminals by means of a modem connected to a standard telephone line. The technology was cumbersome - you actually had to ram a telephone handset into a device which was to all intents and purposes two rubber cups containing a speaker and a microphone - unreliable and incredibly slow. In fact, even at that time, computers were capable of transferring data at much faster rates than the telephone system.

Donald Davies, who was superintendent of the computer division of the NPL at the time, realised that traditional telephone techniques weren't going to cut the mustard as far as computers were concerned and set about finding a more appropriate and efficient alternative.

The eureka moment came when Davies realised that, unlike human speech, computer data does not require a time line. He suggested that data could be sent in small packets containing destination addresses, and that these packets could be interleaved on the same line with chunks of data from other sources, which were going to different destinations.

The packets of data would know where they had come from (the source address) and where they were supposed to be going (the destination address) and contain a packet sequence number (the order in which the separate chunks of information would need to be reassembled to make sense).

The network would decide the most efficient route for each packet of data and since each packet was travelling independently, several computers and terminals could use the same physical network simultaneously.

In 1965, Davies wrote a paper describing how a national computer network could be built using computer-based switching centres and the term "Packet Switching" was born. At the same time in the USA, Paul Barand of the Rand Corporation was working on a military project concerned with voice communications. He had also come up with a computer-based message switching technique.

Meanwhile, deep in a secret bunker somewhere in America, the Arpanet project was busy trying to connect various bits of the US military and government together. But it wasn't until the NPL's Roger Scantlebury presented a paper on the work Davies had been doing at an ACM (Association for Computer Machinery) symposium that the men in black from Arpanet started to take notice and tracked Scantlebury down.

Soon after that meeting, the work of Davies, Baran and Arpa was brought together and Arpanet subsequently adopted packet-switching as the underlying technology for what we now call the Internet.

The NPL Network

By 1971, the NPL has set up its own local area network (LAN) connecting the various buildings of its Teddington laboratory together using discrete 11-way coaxial cables. This was the first large scale use of packet switching, predating the first draft standard for Ethernet by nine years. According to NPL, the success of this network persuaded Arpa to adopt packet switching and introduced the idea of high speed lines. At the time NPL was achieving speeds of 1.5 megabits per second which everyone else involved in network technology thought was crazy. Then again, the folks at IBM once very publicly stated that no-one would ever want a computer in their home.

Because the network had been built, it now fell to the engineers to demonstrate practical uses. Very soon the various departments of the organisation were creating and storing text documents, which could be retrieved, edited and saved back to the central mainframe. Although they didn't know it at the time, the file server was born. Soon after a system for sending text documents to individual terminals was developed - a precursor to the email systems we all now take for granted. µ


Share this:

blog comments powered by Disqus
Subscribe to INQ newsletters

Sign up for INQbot – a weekly roundup of the best from the INQ

INQ Poll

Happy new year!

What tech are you most looking forward to in 2015