San Francisco is hopping this week over Web 2.0.

What is it? It's a database.

When you use a database as your basic site design template, then everything becomes a database call or a database interface. Thus, you can do anything. You can do blogging, you can do identity, you can do customization, you can do community.

The problem is getting stuff into that database. Can you share data among databases? Users don't like that. But how do you get permission for all the relevant, needed data to get into the database?

The obvious answer to that is that users have to live inside the database. A lot. This restricts choices, because time is limited. It means there are only a few "winners" -- a few sites will scale to get everyone's data and everyone else will lose out.

Thus there's a self-liimiting aspect to Web 2.0 trials. Unless....as with Sxip, you can take your personal data (the stuff that would fill a database) with you, and control it. Then you point that data to whatever database you choose to be a member of.

Another important point, which I looked at on another blog. Because a really scaled Web 2.0 effort requires so much data, and so much programming from so many different directions, it's a great opportunity for open source.

There's one great site design out there, something that unifies a whole lot of bits and pieces. Open source lets you get at those bits and pieces. Trying to build an AOL from scratch won't work, because there are too many bits and pieces that go into making something the scales and works.

So this is open source's chance. And I hope they don't blow it.