Hi, short version of the question: when elements are 'drop'ped from a list, when is the memory for those elements released? And is there some way to control that? Longer version: I have this situation where I've got a State monad carrying a Data.Matrix, and I... 1) generate in infinite list of monadic operations with (repeat mf) where mf is a monadic function. 2) 'sequence' (from Control.Monad) to get a monad containing the infinite list of all states (i.e., a list of Matrices). 3) evalState(T) to extract the list 4) extract the nth state with (head (drop (n-1))) I like this approach. However, there seems to be a memory management issue: a Matrix itself should only be, I'm guessing, somewhere around 100KB. But instead I'm maxing out the 3GB of RAM on my old T60 laptop. Maybe I'm generating list elements (Matrices) a lot faster than memory management is releasing them...? -- 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).