It is the 20th anniversary of the first Web client-server communication over the Internet. E&T looks at how the Web was won and where it's got us.
If you want to know the ultimate origin of the World Wide Web, you have to go back to World War Two. It was only a month before two atom bombs were unleashed on Japan after the culmination of the top-secret Manhattan Project. Vannevar Bush, a computer scientist who provided the organising force behind the project, distilled his thoughts on a system that would, instead of keeping information secret, make it far more accessible into an article for the general-interest magazine The Atlantic.
But Bush lacked the appropriate technology to make his vision work. As an inventor of a microfilm reader, Bush saw his memex being based on mechanical switches and photographic film. But the core of his invention was the idea of links between information. It took 20 years to find a name, when Theodor Nelson coined the term 'hypertext' in a paper published in 1965.
Programmers around the world tried to put Nelson's ideas into practice, some found their way into the mass market, such as Apple's Hypercard - popular software that was ultimately killed off by the computer maker.
In 1980, several years before Hypercard appeared on the market, a software engineer at CERN thought hypertext might deal with a problem the physicists and engineers faced on the control system for the Swiss centre's Proton Synchrotron (PS).
'The PS control system faced a crunch. It was going to be turned on but [the people on the project] didn't have time to finish it. They brought in contract programmers and I was one of them,' said Sir Tim Berners-Lee at an anniversary event organised by CERN last year.
One of the big problems on the project was coordinating different pieces of information, Berners-Lee says: 'I came along and found this melee of people doing different bits.'
The answer seemed to lie in software that could collate the information generated by people working on the project. Berners-Lee explains: 'I wrote something that ran on a text terminal: you could follow links by typing in numbers. When you found a relevant place in a document, you could add a new note card and say what the connection was to the existing card.'
Berners-Lee gave the system the name Enquire, after the Victorian book Enquire Within Upon Everything. 'The topology of the system was very hypertexty,' he says.
Enquire featured in the initial proposal for a distributed information system that Berners-Lee wrote almost ten years later in the spring of 1989, again as a plan for a tool that would let physicists and engineers share what they knew about their systems. But the Web did not come to life until more than a year later when Mike Sendall, Berners-Lee's boss, OKed the purchase of a NeXT workstation and a plan to use the prototype World Wide Web project as a way of testing the machine for use more widely at CERN. Berners-Lee says: 'He didn't give me the nod [to do the World Wide Web]. He gave me a wink: use it to test the new machines.'
Berners-Lee wrote the browser, so that it was easy to edit pages, and he used stylesheets to change formatting. The system was devised to let researchers pool information on the systems they were building, often working in very loose groups. 'Students would join the project for a year and then leave,' Berners-Lee explains.
'I wanted to get people up to speed using this hypertext environment like a big sandpit. Together you would construct hypertext about how you were going to do one big thing. You would have all these pointers to explain why each decision was made. 'It was just meant to be an editable hypertext mess. People now have Wikipedia, but they have to write in this weird language.'
The slogan for the early Web was simple and echoed Bush's concept of the memex: 'Let's share what we know.'
Although he had the initial software up and running by the end of 1990, there was only one machine on the World Wide Web for several months. It was someway into 1991 when the software made it onto a second machine. The first one to run a Web server in the US appeared on the network almost a year after the first implementation at CERN. Fortunately, CERN had lifted a ban on using TCP/IP - the protocol over which Web requests are made - to make external connections in 1989, according to Ben Segal.
A couple of days after the first US Web server appeared, Berners-Lee gave a demonstration at a conference that had rejected a paper on the system, because the organising committee considered his proposal too simple an implementation of hypertext.
Outside CERN, the demonstrations were text-only and did not show what was possible on the NeXT workstation.
Jean-Francois Groff, a software architect who worked with Berners-Lee and is now chief technical officer at online-music company Fairtilizer, wrote a hack that made it possible to access the machine from any Unix terminal, but only in text mode. 'If you typed 'two' followed by 'enter', it would follow the second link. This was how people experienced the World Wide Web in the early days,' said Groff at CERN last year. 'People didn't know we had this read-write editor: it's all my fault.'
Robert Cailliau, who worked with Berners-Lee at CERN and wrote the first MacOS browser, added: 'People saw text and thought: 'we should improve on this'. They didn't come to us. They didn't realise we had gone so much further. We lost about five years conservatively.'
Some things are still missing from the original plans for the Web. Berners-Lee says he wanted the ability to create groups on the Web. 'The ability to create groups is something we are still trying to push for,' he says, adding that social networks such as Facebook have gone some way towards that plan.
A crucial decision lay out of the hands of Berners-Lee, Cailliau and Groff. Although the idea of charging royalties for the use of HTTP and HTML had been floated, CERN ultimately decided to make the Web free. 'A very important step,' says Berners-Lee. Within a few years of its invention, CERN relaxed its hold on the protocols, putting them in the hands of the Massachusetts Institute of Technology (MIT) in the US and INRIA in France.
Berners-Lee himself moved to MIT as the centre of gravity of the Web had shifted across the Atlantic, largely because so many institutions were connected to the Internet already. The mid-1990s would see the creation of the Mosaic browser, responsible for putting the Web in the hands of users outside educational institutions. As that morphed into Netscape, the Internet boom began, sealing the success of the Web. 'There is a lot of innovation on the Web because it's an open, royalty-free platform. You can invent and put things out there and do things that are important for society,' Berners-Lee claims.
F Berners-Lee is actively pushing two themes that he sees as good for society. One is for governments to open up their data to make it more widely accessible by Internet users by means of semantic-Web technologies (see p52 for more). The other could have a much wider impact on people around the world. 'The Web has got a lot more fancy,' said Berners-Lee earlier this year. It favours 'geeks in the West, people who are very lucky'. But the lucky ones on the Web account for only 20 per cent of the global population. 'What about the other 80 per cent?' he asks. This thinking led to the creation of the Web Foundation, an attempt to bring Web-powered IT to practically everyone on the planet that might let them change their environment for the better.
Aid agencies have argued that the priority is to feed people rather than provide ready access to the Web. But Berners-Lee argues: 'The nifty thing about information is that when you move it, it exists in both places at once. The sort of effect that IT has can be very different to other things. People can begin to solve problems like getting clean water and healthcare for themselves.'
Steve Bratt, CEO of the Web Foundation, says: 'We want to promote the Web as a tool of empowerment.'
At a seminar organised by the TED non-profit group in Kampala last year, locals and aid workers pointed out that there are big problems with language in countries such as Uganda: there are just so many local, unwritten dialects. Barbara Birungi of Appfrica pointed out: 'Most people can speak the local language but not read it,' adding that English is the only common written language in the country. Stephane Boyera of the Web Foundation argued that the emphasis should be on protocols such as VoiceXML rather than HTML because it focuses on the spoken word.
However, Paul Bagyenda of Digital Solutions claimed that resorting to Google text-based services to access healthcare information 'show high levels of usage'. He added: 'Once you know what you want to achieve, the technology exists.'
Bratt says the organisation's staff have been 'talking to people in Africa, finding out whether these hypotheses are correct'.
At Nokia World, where Berners-Lee delivered a keynote speech last September, he claimed a big obstacle to usage of the Web in the developing world is bandwidth. Developers and operators should concentrate on delivering systems that can supply information over slow connections. Mobile applications can be good at this, he says, but they have the disadvantage that 'you can't link to them'.
'I would like to see the code for the Web apps stay resident on the phone so they can work across a low-bandwidth network,' Berners-Lee says. And, as with the Web's beginnings, he wants the new community of Web users to be adding their own information.
'People haven't realised they can create content. People in Ghana browse blogs but hadn't thought of writing their own. They haven't added their own towns to Wikipedia. We can enable them by giving them the technology. The entrepreneurial spirit is there. It is a matter of connecting it up.'
The Web could help people in developing countries improve their lives by sharing what they know. Twenty years on, Berners-Lee's plan for the Web has come full circle.