The C Companion

First of all I should introduce myself as one of Allen's former students at the University of California Extension in Berkeley, CA. I started with him about 1990, and have been in contact with him off and on ever since. I last spoke with him last year. In order to do justice to a review of this book, I must interweave it with a discussion of what it was like to be in one of Allen's classrooms. This book was one of the required books in Allen's course, "C For Professional Programmers". And believe me, he meant it when he said, "Professional." As a teacher, he can only be described in one word, "brutal." For the first week of that class, the assignment was to write the "Hello, world" program (which he considered non-trivial just starting out). In the second week we were to read our compiler's documentation -- "all of it" (yes, in one week!) That's when Microsoft C version 6.0 was out and the documentation consisted of three huge binders with hundreds of pages each. His homework assignments were hideous for those just learning C (fully justify a string in place, horizontally concatenate an arbitrary number of files... well, you get the idea). If he thought your C was not C-like enough (more like Pascal), he had no problem giving you a zero on a homework assignment, or subtracting a huge number of points for syntactical infelicities (even if they worked). This class, like all the others of his I attended, was stocked with Computer Science students and some graduates. Even these talked about having to spend 40 hours a week on homework and reading assignments (Allen nodded approvingly -- he insisted on at least 20 hours a week). As you might imagine, the attrition rate was high. The class ended up with less than half of those who started. But, for those who survived, the rewards were enormous. For one, Allen was a very personable and engaging person and a great lecturer. The book under review was formed from the crucible of his classes. As Freeman J. Dyson wrote in his preface to the book, "Readable Relativity" by Clement V. Durrell (one of his grade school teachers), "Schoolboys are the best possible corrective to pious inanities in any field." That certainly applies to this book. It is simply excellent. The raison d'etre of the various topics in this book is indicated by various little secrets that Allen told us which I will now share. Allen once mentioned that C is a fast (to write) assembly language. (Hence the tutorial on, and connection with assembler in the book.) Another time he mentioned that it was written for those who want to write compilers. He then described what he meant. The Pascal Writeln statment cannot be implemented in (classical) Pascal -- it would violate the calling conventions of the language. However, the printf statement can be written entirely in C. This book gives a sample implementation of printf, and you get the point when you see pointers backing up into outer (or inner) space, something impo

I had been introduced to C before I knew assembly and was left wondering the magic behind the curtains.. what happens when u call a function.. when u declare an array.. I mean I knew that it must mean something in assembly but the link was not there in any of the C books I looked up. For an engineer who knows hardware and is curious, there is no other place than this book. Holub breaks up C into assembly and show you what happens as you compile and run it, side by side. He takes the mystery out of C programming. Wonderful for folks who missed the link between assembly and C. C has been called as High level assembly language and the proof is here in this book.

Personally, I think this book is better than the K & R book. If I read this book when I was just learning C, then I would give up and say to myself "It's too complicated. Why do I have to understand all about these..." But hold on. Don't you want to know what the stack frame is and how it works during recursion and how a variable number of arguments are handled, how the automatic type-conversion is related to the portability issue? Then read this book. If not, just skip this book. The only short coming of this book is that it's not ANSI C compliant. But it matters little if you already know about ANSI C. You will be able to discern the difference yourself, and probably can smell the history of the evolution of C. It's sort of a bonus. Allen I. Holub, what about writing the 2nd edition of this book? Then, that would be a great a help even to many embedded C programmers, I think. Sencon-to-none C book for me.

Of all the C books I had, I would rate this book my number 1. Just the section on pointers is worth the price of the book. The way it explains pointers, pointers to pointers, etc is so done so elegantly and clearly. No complex double talk or technical quibbs.I wonder why Mr. Holub stop authoring anymore books... Mr Holub, if you are reading, please continue your excellent work ...

When I started programming in C I was doing embedded operating system design and this was the only book I found that clearly explained what a stack frame was and how to unwind it in assembler to handle errors. It contains information for the experienced C programmer that I have never found elsewhere on how C code works 'behind the scenes' (assembly language, binary arithmetic, complex pointer math, stack frames, function call jump tables, etc.). It is not a book for beginning programmers, and it is not in any way an introduction to the C language, it is however, an unusual resource for the experienced programmer.