Weaving the Web by Tim Berners-Lee (Texere)
Developers looking for a model to follow could do worse than adopt the approach of Tim Berners-Lee world wide web.
Berners-Lee set the whole glorious thing in motion back in 1990, writing the first web client and first web server while at CERN, the particle physics lab in Geneva that dabbled in such projects for a while. Now he’s the director of the World Wide Web Consortium (W3C), which works on web technology standards.
Berners-Lee came to the idea of the web by way of wanting to keep track of people, projects and computers and seeing the value in nascent hypertext systems to do this on the internet. The web, which could have been called Mesh, MOI or even TIM if events had unfolded in a less self-interested way, was always driven by a few key principles: universality, of presentation, access and the ability to link, was one.
And stick to them the principles does. Berners-Lee is gung-ho on open source, having made all the web code he has written available for inspection and alteration since his first release in 1991. This approach, he suggests, is ideal for collaborative work, as long as version control is closely maintained. All W3C software is still open source. It’s unlikely the web would have been as freely navigable if this wasn’t the case. So it’s good news that W3C continues to be heavily involved in developing the standards of web services, the XML-based interfaces that will enable inter-program communication.
But the way Berners-Lee tells it (with the help of Mark Fischetti), people took some convincing of the web’s potential and the desirability of working toward the same ideal.
Mostly to the good, the progress of the web hasn’t been strictly linear, even if the ideas behind it haven’t deviated much. For instance, a group of programmers created Apache out of a stalled project that planned to compete with the web server Berners-Lee had already written — now W3C is among the millions that use Apache. And it’s still going on: in the book Berners-Lee talks about PNG as a scalable web image spec to replace GIF, for which Unisys held the compression technology patent. WC3 proposed the second edition of PNG just a few weeks ago.
In the same rigorous approach to openness, Berners-Lee is particularly scathing of patents as a method of protecting intellectual property in technology that hardly innovates. Some patents, he says, just take a well known business process and do it in software.
“Others combine well-known techniques in apparently arbitrary ways to no added effect — like patenting going shopping in a striped automobile on a Thursday.” The US Patent and Trademark Office is ill-equipped, he says, to check for “prior art” in the relatively new field of e-commerce.
But internet time moves quickly, and since this 200+-page paperback was published (2000) a few things have changed. Berners-Lee speculates on what the semantic web — a web of data that can be processed by machines — and smart search engines will be able to do. While some fret about the relentless rise of Google (even the w3.org site now uses Google for search), the quality-of-links approach to finding the right site it spawned has revolutionised how we use the internet. The leap it’s helped us make has made the horizon seem that much more interesting.
Broatch edits the books pages for Computerworld.