simply type in some keywords, such as "flapper," "hang-dogging" and "dirt me." OpenCola will construct a software robot (sort of a single-purpose search engine) that goes onto the Net to dig up stuff that matches your criteria. It does two kinds of digging. First, it fetches files from the Web just like any other search spider. But the robot goes one step further by looking for other users' OpenCola robots looking for the same kinds of things you are. Then, your robot grabs files the other users have fetched and found worthwhile. In other words, you benefit from the decision-making of the other like-minded OpenCola users. It's like a Web-size version of Amazon.com (AMZN)'s "people who liked that book will like this book" service.
When you get the results of an OpenCola search, you can click a button to "reward" the robot for finding good files, which encourages it to bring you similar documents, or click another button to "punish" it when it finds bad files, which will teach it to stop dragging that kind of trash into your computer. This information is then shared with like-minded users on the OpenCola network and can lead to interesting surprises, says Doctorow. For instance, your robot will eventually start to bring you documents that aren't about rock climbing, but which have been pegged by other climbers as good.
The good stuff gets pushed to the top even more quickly because OpenCola gives extra weight to the decisions of users who consistently pick things that a lot of other like-minded people find useful, creating a sort of reputation economy in which the behavior of the most trustworthy users gets propagated through the network.
Buzzwords aside, the aim of OpenCola is to replicate that infinitely hip friend of yours who points you to songs, video clips and articles he just knows you'll love.
This passion for pushing the obscure comes from Doctorow's experience as the quintessential niche guy. As a science-fiction writer, a Mac user and a Canadian citizen, Doctorow knows all about niche markets. The biggest science-fiction magazine in the world, Analog, has fewer than 70,000 subscribers. There are 10 million Mac users, but it is considered by most Web-application developers to be too much of a niche to bother with. And Doctorow is from Canada, a country of 25 million English speakers who are often not considered cost-effective for publishers.
Doctorow's dream is that OpenCola's technology will help the millions of artists who can't afford to publicize themselves in a worldwide market. He believes OpenCola will make it possible for any arbitrary piece of media to find its audience.
That's a much different approach than Napster or Gnutella, which don't allow musicians to contact or transact business with potential listeners. OpenCola has a mechanism in place for publishers to sell their material, or give some or all of it away. Naturally, OpenCola takes a cut of the transaction; that's the business model. A musician might offer a song or part of a song for free, allowing access to a full download for a fee.
OpenCola also supports the "tip protocol" a la Stephen King, where users download material and pay for it on the honor system, and the unfortunately named "street-performer protocol," where a creator releases a work into the public domain after a certain amount of money has been "thrown into the hat" by any number of customers.
Of course, Doctorow hopes that big companies as well as niche publishers will see the power OpenCola has in connecting consumers with producers.
One gets the feeling that Doctorow is doing OpenCola partly as some kind of postmodern science-fiction performance art experiment, where humans and their personal robots team up to score more whuffie points (used in OpenCola's reputation-ranking system) than anyone else on the network. Or at the very least, it resembles an information-age Rube Goldberg contraption that is dependent on the perfect operation of each component, lest the whole
