The dry technical term "content-addressable storage" doesn't sound exciting, but superficial explanations can make it look like magic, to the point raising suspicion. In this article, I will show that content-addressable storage is a technology that works, is already in widespread use, and holds many promises for the future of both scientific programming and the management of scientific data. I will start by outlining the theory, and then illustrate how it works in practice, using IPFS, the Inter-Planetary FileSystem (http://ipfs.io/), as a vehicle.