[Haskell-cafe] efficient operations on immutable structures

Hi again all. From some online research, I understand that operations on complex immutable structures (for example, a "setter" function a -> (Int, Int) -> Matrix -> Matrix which alters one element in the Matrix) is not necessarily inefficient in Haskell, because the (compiler? runtime?) has some means of sharing unchanged values between the input and output structure. What I am not clear on, however, is how this works, and how you ensure that this happens. Could someone perhaps elaborate, or point me to a really good read on the subject? -- http://qlfiles.net To protect my privacy, please use PGP encryption. It's free and easy to use! My public key ID is 0x340EA95A (pgp.mit.edu).