Wednesday, January 26, 2005

Building a better search

I talk a lot about Google and the interesting things that are coming out of the Google Labs. But interesting things are also happening happening elsewhere, and Scientific American has an intriguing article about search engines and data mining.

People have been talking about augmented reality for over ten years now, and we are perhaps finally reaching the stage where connected technology is becoming ubiquitous enough to support it. Data mining, and especially integrating GIS and other related systems into a location dependant search, are becoming increasingly common. Services such as "Find my nearest..." are now commonly offered by the mobile operators, but they're still at a fairly primitive stage. It doesn't really help me to know that my nearest postbox is on Church Street, if I don't know where I am, or where Church Street is either. The integration of GPS, either directly into the mobile handset or via a Bluetooth connection to a seperate unit, will improve things a lot. However, the real improvements needed are in the software, we already have the hardware to do a lot of very powerful stuff, but our software is falling behind. Like a lot of other people I'm currently being put off writing the software to make use of the hardware by the steep learning curve of the mobile platforms and by the knowledge that this is a going to become a great deal easier very, very, soon. So why struggle now?

The introduction of Python for Series 60 is, I think, going to prove to be a ground breaking move by Nokia. For the first time writing code for a mobile platform is easy. We've all gotten lazy, and we're too used to high level languages these days to want to spend time talking to the hardware directly. Of course, I'd rather they'd released Perl for the Series 60, but I can't have everything...

So maybe I should stop talking about this and write some interesting software, of course, I'd have to buy some more hardware to do anything interesting. Pity that!