This is probably only of interest to the frequent flyers among you, but I thought I'd post it because I've been intrigued by the site Farecast.com since its inception a year or two ago.

Farecast, if you haven't used it, applies predictive analysis to airline airfare and booking trends and makes predictions about whether they will rise or fall for a given route over a certain period of time. From this, you can make a determination as to whether or not it's a good idea to buy now or hold off, a perpetual question for travellers who plan well-ahead, since it's not always true that booking early will get you good fares... the opposite is equally possible. The site also provides some good tools to visually adjust your schedule for the best possible rates in a given window of potential travel times.

To be honest I haven't tracked their success rate in predicting fare changes but I see they are now doing something similar with hotel room rates so someone must think it's working. And Microsoft isn't particularly interested in airfares... according to Todd Bishop of the Seattle PI, the Redmond company's interest is in the underlying technology rather than the specific applications. That makes all sorts of sense to me; it's the data analysis that is interesting and most applicable in today's world of massive and rapidly accreting datastores.

Whether or not Microsoft will be able to leverage this easily and in a useful timeframe is another question. According to Farecast's website, they are "committed" to using Open Source technologies and have built their system up on a grid composed of Microsoft nemeses Linux/Unix/OS X, Java, Perl, and MySQL. As we've seen with previous Microsoft efforts at integrating Open Source based companies (a la Hotmail), sometimes the cultures and capabilities don't translate well or easily. It took four years or more to get Hotmail switched over to Windows; maybe they've learned from that process.