Code

Web of Data

Adam Bosworth of Google gave a great keynote speech this morning at the MySQL Users Conference, focusing on doing for data what HTML over HTTP did for content. This is an interesting and slightly uncomfortable idea for database folks in the sense that it relies not on a centralized database but on harnessing data on the Web through an open standard. And it was interesting to see which standards he cited as models: RSS and Atom, and Amazon’s OpenSearch RSS:

Folders are not a very efficient way to find things - searches are. Can we take a database and do the same thing for the web that we’ve done for data? We haven’t done it in a way where you get back information. We get content. The lesson of the last ten years is that you open up your formats directly. When you do that, you get a huge increase … an equivalent explosion in data.

Querying XML data, Bosworth said, was too complicated and took too long to build standards. But a framework for querying web data (not content, but data, which means tagging and categorization) could be easier to build and would scale “effortlessly.” He cited the success of RSS 2.0, which he said “won in a walk” because of its simplicity. Finally, he dropped this tantalizing prediction for anyone who accomplished this vision: “You guys can be Google too. You can scale up and out in ways Oracle can only dream of.”

Update: An earlier PowerPoint version of Bosworth’s talk.

Other reaction: Scott Mace.