nextthing.org

This book is a bit dense as an introduction, but serves as an excellent reference. Supplement it with a good algorithms and data structures book, such as Bruno R. Preiss’ excellent and free Data Structures and Algorithms with Object-Oriented Design Patterns.

This is a book most people (including myself) have probably browsed through or skimmed, but never read. Nevertheless, it sits in the background in many CS programs, informing the way introductory computer science is taught. Alternatives for introductions to CS vary from school to school, but many use SICP. My personal favorite is the text used at Harvey Mudd College, “Computer Science: Abstraction to Implementation”, aka “Computer Science for Smart People.”

This is the canonical introduction to C, the lingua franca of systems programming, and a favorite source of interview questions.

AI is a fascinating field, and this book covers a lot of ground. Incidentally, Peter Norvig is the Director of Research at Google.

I work in an OS group, so OS books are important to me. This is the book my OS class at Mudd used, and I still refer to it at work.

An alternative to the Dino Book.

If you can’t joke about what the “U” in UDP stands for, you need to pick up a book on networking. Any good book will do.

There’s a new edition of this book coming out soon. For many years this was the standard in teaching compilers, though recently “Engineering a Compiler” has become more popular.

Often plugged as the modern alternative to the Dragon Book, this book has proven popular in recent years as the Aho book has grown increasingly out of date.

A whimsical introduction to Scheme and functional programming. In my opinion, CS graduates should at least be familiar with one functional programming language.

If you want to understand why throwing more engineers at a project may actually make the project take longer, this is the book to read.

Also known as the GoF (Gang of Four) book. There are also a variety of “Design Patterns in X” books for various languages, which may help make the various abstractions clearer.

Recently updated by Rago, the 2nd edition of this book is truly excellent. At the very least this is a required reference if you’re programming on a UNIX-based system or just want to understand the design of the API’s.

In much the way APUE is the classic for learning UNIX programming, this is the classic for learning network programming.

Whereas the first volume teaches you how to get your computer to talk to others on a network, this one teaches you about interprocess communication (or how to have programs on a local computer communicate).

Contains great descriptions of many of the parts of the old AT&T Unix System V. Given the increasing complexity of Unix systems, this a good introduction to the design of an actual operating system from when such things were much simpler.

SCO is not entirely evil, having given permission for the republication of this book in 1996. This book contains the source code for the 6th Edition UNIX kernel with excellent commentary on the code and a description of the PDP-11 architecture on which it ran.

The theory of computation is all about computability and complexity. In other words, it asks what can be computed, and at what cost? This comes up in programming surprisingly often, as tasks that are deceptively simple for people (like finding the shortest route between several stops on a map) can be quite complex for computers to perform. This book serves as a good introduction to the field.

Another introduction to the theory of computation, with lists of common hard problems.