An Open Source Silo

Okay, so Doc Searls has been blogging for a long time about how you don’t want to get involved in a silo (context: a silo is a proprietary user interface built on top of proprietary APIs built on top of a proprietary operating system running on proprietary hardware. If you can’t think of any examples, pull your cellphone out of your pocket and think about it for two seconds). I’m working for Cloudmade, which hopes to make money by offering a silo of services over OpenStreetMap. The thing that makes our silo different (bigger! better! stronger! more open source!) is that it’s built on a foundation of openness that allows other people to fork (or perhaps “clone” is a better word) our services. But like all services, they’re never open source, because they aren’t just software. They’re an ongoing commitment of resources for which any sane person demands payment (unless you really WANT to own the village commons).

I say this not to justify our silo, but instead to point to a different kind of openness: open data rather than open source. If the world of the future is to be open, and yet built from services, it’s going to be the data that’s open.