www.roanoke.com

A while ago an editor from USA Today who was teaching me about some software offered a truism about journalists with data: we always go straight for the biggest and baddest thing in a database.

So, you can guess what I did when Roanoke Times Metro Editor Brian Kelley told me about a the database at the Virginia Big Trees Program. I looked up Roanoke to find the biggest sucker in there.

The program, as the website itself puts it, "relies on volunteers to search for, nominate, and verify the measurements of big trees in Virginia. When a big tree is reported to the program, it is put into the Virginia Big Tree Database maintained jointly by the Virginia Forestry Association, the Virginia Department of Forestry, and the Virginia Tech College of Natural Resources."

The big winner turned out to be a 90-foot crack willow (Salix fragilis, for you Latin lovers out there) at 6311 Blacksburg Rd, "on right side of house in a field. Visible from road." That's in the picture.

You can search by locality, by the common or Latin name of a species, or other ways in the advanced search option.

This is terrific data partly because it's fascinating for what it documents, and partly for the way it documents it: by hand, by people out in the field with tape measures and cameras and pencils and paper. Many entries include hand-drawn maps to the locations.

In some ways, this is like the DataSphere's own black bear sightings map. It invites you to add to the database yourself by reporting what you've seen.

That's one of the best ways the Web works these days: the best content is created by the users themselves. Think YouTube for the most obvious example. YouTube for trees? For bears?

This is data doing what it can do best: becoming a conversation.

Ok, that's nerdy, but then this all started with my own glee over a database of big trees. I'm more of a data dork every day.

Lucky I'm married, because at this rate, I'd never get to kiss a girl again.