In the spirt of Google Image Labeler, Ryan Grimm has added a code quiz to the O’Reilly Labs site. The quiz selects a code example from any one of the books in our complete Xquery database of all our content, and offers you a choice of four books that the code might be from, only one of which uses the language in question. For bonus points, you explicitly identify the language. Like the Google Image Labeler, this game does actually gather useful metadata for us, but Ryan’s real interest is in finding a game that helps “teach us what we don’t know.” He writes:
After a game is finished I generate a few tag clouds similar to the ones in the content statistics based off of how a user answered the questions. The hope is that these clouds can represent what a person knows, what they thought they knew, and what they knew they didn’t know. If we can turn this hope into a reality it might help show users areas they can improve on.
I found the game curiously addictive. It seems obvious, but after a few tries, you realize what an engaging way it is to explore code snippets doing interesting things. (Also see the Content Stats and Code Search tools on Labs.)
We’re interested in opening up our content database to other developers who might have cool ideas about what applications they could build with access to all our content. If you have ideas, let me know, and we’ll see what we can do about giving you the keys to the family car. We don’t yet have a formal application process, but will develop one out of our back-and-forth with the early birds.