Critical reviews are my least favorite aspect of my Amazon experience, but I believe readers expect me to be honest with them. Gray Hat Hacking, 3rd Ed (GHH3E) has a lot of potential, but it needs a reboot and a ruthless editor. I read and reviewed the original edition 6 1/2 years ago but skipped the 2nd Ed. This 3rd Ed (published in Jan 2011) features several exceptionally talented authors (such as Allen Harper and Chris Eagle), so my expectations remained high. Unfortunately, after finishing the book I had collected a pile of notes that I will try to transform into constructive commentary for a 4th Ed, which I would enjoy seeing!
The GHH team needs to revisit first principles and decide just what it is trying to accomplish. I recommend the authors ditch the first three chapters, or radically concentrate on the ethical disclosure debate. The rest of the so-called legal material reads like a brain dump, almost like a blog post that never finishes. In some cases the authors of the sections stray from their topic, such as the "Vendors Paying More Attention" section on p 71. Cut it out! Be ruthless! Similarly, the section on social engineering (ch 4) needs a major overhaul if it is to survive into the next edition.
Other chapters have issues. Ch 7, on BackTrack, is basically just installation instructions. Ch 17 only devotes 17 pages to Web app security; either remove it or add substantially to the material. Ch 18 is supposed to be about VoIP, but it's mainly a discussion of the VoIPER tool. Ch 19 is supposed to be about SCADA attacks, but it's really just talk of the Autodafe and TFTPFuzz tools. In ch 28, the author doesn't explain how Nepenthes acquires a malware sample, besides letting it run on a cable network for a few weeks. Having deployed Nepenthes I know how it works, but I expect a reader who wants to learn about Nepenthes would want to understand it based on the text he or she bought.
The organization of the book needs an overhaul too. It seems to promote a progress of less complicated to more complicated, but at this point it needs to be reconstructed in a fourth edition. Why does Part IV, Vulnerability Analysis, follow Part III, Exploiting? Doesn't exploiting require doing vulnerability analysis? In other cases, material seems redundant. Ch 28 and ch 29 cover similar material but are likely by different authors; I recommend combining them and dropping duplicate material.
There's a lot of good technical information in GHH3E, but I don't see myself recommending it to analysts in a CIRT or similar group. I think if the book rebooted with a focus on specialized material not found elsewhere, leveraging the talents of people like Harper and Allen, GHH4E would be THE book to buy on those topics.