Months ago when Jimmy Wales announced he was going to develop an open-source search engine, he talked about starting off with "a couple of open source and Java Search engines that are based upon the Apache Jakarta Project called Lucene and Nutch. Programmers would be able to adjust, copy and exchange the code among one another while others give feedback."
He's taking the next step now as he told participants at the O'Reilly Open Source Convention in Portland that his for-profit Internet publishing company, Wikia, has now acquired a search technology called Grub, which will be the foundation for the search engine itself.
What the technology does is index the Web "by borrowing the processing power donated by volunteer computers." It is said to work like the SETI@home project that looks for extraterrestrial life. Wales will allow the code behind it to be tinkered with and improved upon through an open-source license.
While the product will compete with existing search companies, Wales says the ultimate goal is to make Internet search more accurate.
"It's not a good thing that we are getting search results from a handful of very large players and we have no idea how they are generated," Wales said.
When Wales gets the index built, from there useres will be able to do their searches and receive answers from community feedback and computer algoithms.
The strategy of positioning himself as the David vs. Goliath, or the small guy taking on the hoards of Google "rocket scientists," isn't a bad one, but the problem is far more than marketing and public relations, it's also about cost.
Google isn't only about their huge numbers of programmers it retains, it's also about the hundreds of thousands of computers it has. Google spent about $575 million on data centers and computer equipment the last quarter alone. This is why Wales started this project up differently than Wikipedia, by making it for-profit company from the beginning.
The major question I have about this is if there is a real market demand for this. You don't really hear a lot of people asking for this type of model, although I'm sure in these types of seminars you get the usual cheerleading and go-get-em stuff.
This sounds to me like a real long-term project that will take a lot of time to take hold ... if it ever does.
If it continues on and expenses don't become an issue, there will probably be a small percentage of people that like to tinker with open source get involved and use it. But the average user will have a hard time knowing why they should use this against something else.
As Google (GOOG) gets bigger and bigger, they are looked upon more as a typical corporation. That could open up a crack in their armor that could make smaller rivals look good. It will be interesting to watch the evolution of the project as it goes on.
Comment Preview