We want OpenSearch to do for search what RSS has done for content.
OpenSearch, at the very least, is a step in that direction. Today, at the O'Reilly Emerging Technology Conference in San Diego, A9 introduced a new set of technologies to help syndicate search content.
OpenSearch is simple in concept -- there are thousands of different search engines (and by this I mean far more than just "Web" search), and each one of them presents the search results to the user in a different way. A bit surprisingly, there is not one common, popular format. OpenSearch is such a standard for search results.
OpenSearch has the potential to be transformative for those sites and markets that live along the long tail of search. To use the example of medicine, the specialized PubMed search engine has a strong appeal to the medical community, and it return very accurate domain-specific results for medical queries. Those results -- and the way the queries themselves are understood -- are entirely different then they would be in a lateral search (such as a surface web search). This is the perfect example of where a narrow band of the long tail has tremendous value to those that need it, but does not need to be promoted to the general audience. And with OpenSearch even those smaller verticals can have the same distribution channels as the large search engines.
Rather than invent yet another new standard, OpenSearch is built as an extension on top of the wildly popular RSS 2.0 XML schema. Since there are already hundreds of applications and tools that can read RSS, and there already so many developers who understand how to author RSS, OpenSearch launches with a generous head start.
In fact, just hours after the announcement, the first of many third-party OpenSearch columns appeared on the A9.com website. Transforming an existing search engine into an OpenSearch RSS compliant search engine is straightforward and intuitive -- the only real downside may be handling all the potential traffic that would come from people clicking on your results.
The launch of OpenSearch 1.0 is just the beginning of what is possible. Once search results are syndicated the same way content is syndicated, we will see an entire universe of tools and applications built up around the data. In the same way that RSS, or REST-based tagging APIs such as Flickr and del.icio.us, are offering a spring-board for truly useful services, the ability to query nearly any deep vertical, from anywhere, could change the way we thinking about search.
I've had an amazing time working on this project -- the team had the ability to turn ideas into reality almost before they were fully formed. I can't tell you the number of times I would check a development server only to find a feature implemented that was just discussed a few hours before. And, laptop reboots aside, having Jeff launch your project is just pretty cool.
Feel free to send me any questions, and I'd love to hear your thoughts about OpenSearch and where you want it to go in the future. And go build a column! If you really like doing it, who knows, maybe you'd want to do things like this full-time?
(PS: I promise a better image and a lot more content about OpenSearch as soon as I'm back on a Mac. But no matter what, it is great to finally be able to talk about it.)