Weblogs: Web Standards

Royal Society: Tim Berners-Lee lecture

Tuesday, September 23, 2003

The Royal Society invited Tim Berners-Lee to give a keynote speech on the topic of "The future of the World Wide Web". The event was last night, Monday 22 September up in London. TBL is the creator of the World Wide Web, and one of the top 100 Britons, and I get the opportunity to see the man in action.

Getting there and initial expectations.

And I wasn't disappointed. (Although taking thirty minutes to get from London Bridge to Charing Cross by train is a little frustrating). The venue was packed - probably over two hundred people crammed into a long lecture room, plus a large number of people watching on the live webcast. Looks like tonight won't be the night to get his autograph in my copy of "Weaving the Web".

Where the Web has been

Tim's speech started off about the building blocks of the web: Hardware independance, software independance, Accessibility and most importantly open standards. There's a strong conviction that the web has succeeded because of the openness of those standards. One of the key points that came across was that good design involves the separation of content from its form.

Incidental anecdotes of his time with CERN and the people he has worked with through the years. There is particular emphasis, not on the web itself, but the people who made it happen. Even those people out of the public light, but still making valuable contributions to the Web.

Documents are for people, data is for machines

The second main point that came through is that documents are for people and data is for machines. This particular point is the cornerstone of the next step of web evolution: the Semantic Web. Both forms of information will live together in this mesh of knowledge. The Semantic Web is about knowledge representation, and using simple programs to do useful stuff with that application. The Semantic web is just a web based database, where hierachial databases are limited by hierachy, and relational databases are limited by its tabular nature, the Resource Description Framework can handle both at the same time.

RDF Enlightement

I understand the concept of RDF a little better now. A lightbulb snapped on in my head when Tim explained the combining of disparate RDF documents almost as "overlays" showing more relationships between data. This layered flexibility allows the merging of local, global, public and private knowledge into new patterns that can then be manipulated. These overlays also help link relational databases to hierachial databases, forming more complex and realistic patterns.

This got me thinking... If an RDF application combines RDF documents - how does it find the relevant documents out there on the web. Is there an RDF search engine, and how would it work? The solution seems relatively straightforward. Since every resource has a URL, just index all URIs along with the document it is described in (another URI). Neat and simple.

Semantic Web still sparkling

Through the talk I was struck by the enthusiasm and excitement Tim has with the potential around the Semantic Web. I would have thought after four years, the idea would have lost some lustre, but its all there sparkling like a new toy.

The ludicrity of software patents

The funniest moment for me is Tim talking about software patents and the need to uncover prior art. The example he gave was someone holding a patent for holding triples in a database, and the W3 team found prior art - dating back to 364 B.C.!

Concluding remarks

I came away from the talk with a new sense of excitement - its infectious. The Q&A session included reminiscences of early adopters, notably Pei Wei, Marc Andreesen and the Midas browser. The biggest regret was the two forward slashes in the http schema. AI was referred to, along with Arthur C. Clarke's "Dial F for Frankenstein" in reference to a question about when the web would become self-aware.