I've read through the first half of the book, and am now at the catalog of algorithms that makes up the second half, and I've really enjoyed the book so far.
Most books on algorithms take a very formal, math heavy approach in which they present, analyze and prove things about a somewhat arbitrary collection of algorithms. There's usually no mention of real world applications.
This book, however, takes a different approach, and serves as a guide book for using algorithms in the real world. There's a heavy emphasis on formulating problems in terms of existing, solved problems. If you can "map" your problem to one with a known solution, then you can use the proven, existing solution to solve your problem. To emphasize that point, roughly the entire second half of the book is a catalog of known problems and solutions, with references to software libraries, books and other sources of information.
I also love that the example code is in C. Too many books give example code in languages with a lot of overhead, like Java, and end up obscuring the important parts with a ton of object-oriented crap. Yes - OOP is nice, but unless I'm reading a book on OOP, I don't want to dig through 30 lines of irrelevant boilerplate just to find the 10 lines relevant to the algorithm.
That said, it's not the best code in the world. Some of the snippets could be explained better. And there were a few stylistic issues, such as leaving off function return types and a bunch of global variables, that I didn't like, but I'm willing to forgive those because it's not a book on C, and the lack of syntactic clutter made the algorithm easier to see.
I also thought chapter nine was a bit too long. A good portion of the chapter is spent reducing various NP-complete problems to other NP-complete problems. Interesting, but it was a bit too theoretical, and didn't really fit with the with the rest of the book.