Asymptotic and exaustive
One of the problems in the literature is that "Information Theory" is not well defined. It started off meaning that which Shannon accomplished in Communication Theory, then had evolved into what this book defines as IT. The problem that I have with this is that the theory still relies on asymptotic arguments. We do not do asymptotic experiments. Also, while it mentions Maximum Entropy methods, they are out of date. This method has evolved to the point where it is now fused with Bayesian theory (Giffin 07, arxiv.org). All of that said, this is an invaluable book. It is the most comprehensive work that I know of regarding the asymptotic view of IT.
