I've been thinking about DabbleDB quite a bit lately -- Avi Bryant's approach to solving the super-spreadsheet problem. That is, the problem that small organizations often run into when utilizing spreadsheets. They start out good, but they don't scale to more complex relational data. Of course, the solution that is typically pointed at is a database.

The solution is correct, but unfortunately databases are harder to do, and even more difficult to do them Right, than spreadsheets. So in a nutshell dabble seeks to make databases easy. Looking at their product it does some obvious things, making it easy to build a database. But I keep trying to figure out what does (or could, I haven't used it myself yet) make their product better than others.

I believe that the answer is near-transparent refactoring. They make it easy to incrementally transform more free-form data into typed and structured data. So if you have a field with just a few different things you put in it, they make it easy to suck this out into a related table.

Good idea :)