It's really easy to make mistakes when you're working with floating-point values—especially when you're comparing them. Thankfully, Bruce Dawson has written up a very thorough treatment on the topic.