This is pretty common in my area (theory / algorithms), but mostly to prove that you were the first to come up with something, in case someone else beats you to

This is pretty common in my area (theory / algorithms), but mostly to prove that you were the first to come up with something, in case someone else beats you to the punch while your paper is under review. It also provides open access and the opportunity to go beyond the space limits of conference proceedings (practically all work is published at conferences). This means you can show proofs etc in more detail, or present more detailed experiments, etc. I really like this way of going about things. Tons of stuff is available to anyone, conferences provide great opportunities to exchange ideas and see what others are working on, and journals don't have the enormous power they enjoy in other disciplines.