Friday, November 06, 2009

Book Review: “The myths of Security” by John Viega

My review for “The myths of Security” by John Viega has been posted to Amazon; I gave it 4 out 5 stars.

Think about this book as a printed collection of blog posts – some a dozen pages, some half a page. John’s essays – all 48 of them - read like a typical blog: fun views on hot subjects, controversial opinions, new ideas for the future, dispelled myths, cool technology ideas, etc. I definitely enjoyed reading the book, even if most of the material was at least somewhat familiar to me.

For starters, this was the first time that I have seen a book written by somebody employed by a major antivirus company, who would agree that antivirus solutions don't work too well and slow down systems. It was very impressive to read that the author himself does not use an antivirus solution and didn’t even use one when he' was in charge of building one! (Understandably, he does recommend that consumers use one on their systems)

The following are some of my fave chapter highlights. “Security:”Nobody Cares” is one of my favorites; it covers why people, on average, don’t care about information security. His analysis matches that of some other industry thinkers, but it is presented well in the book.

I also enjoyed his thinking about why Microsoft antivirus solution would never pick up and never present a threat to the big AV vendors. In his opinion, most people do not trust Microsoft as a security brand. He thinks that customers would always go to security specialist and not to MS for antivirus tools, even if such specialist is located in Russia or Czech Republic. Also, it looks like the 30% success ratio for antivirus solutions is pretty much a commonly accepted number nowadays; it is mentioned in the book more than a few times.

One chapter that made me angry was chapter 7 on Google. He basically makes the insinuation that the Google in particular and pay-per-click advertising in general motivates people to hack into systems; a view as illogical as it is silly.

In chapter 26, John has an interesting idea for a Social Security number replacement scheme. Many other chapters contain ideas for improving major parts of security technology, even if in some cases the author has to disclaim them with his disbelief about their implementation potential.

It is quite interesting that in chapter 28 John dispelled the myth that including security early in the application design is cheaper. Compared to ignoring the problem until notice from customers, it is certainly more expensive. He touches most other known security industry “pain points” such as vulnerability disclosure. He proposes to replace “responsible disclosure” with a new scheme from my view looked kinda similar, less dangerous for the world at large but less motivating to software vendors. He also discusses whether disclosing vulnerabilities reduces or increases the risk for consumers (in his view seems to increase it).

Closer to the end of the book chapters get shorter and shorter. For example, chapter 42 ends up being half of a page in length. It pretty much states that he would sacrifice some privacy for more functionality and so would most of the others, which seem to be a very popular view nowadays.

I was very happy to find that he devoted an entire chapter - 2 pages in length - to criticizing academic security research (one of my pet peeves!). He says “lots of academics are reinventing what security industry has been doing for years. “ [They are also reinventing a lot of “epic FAIL”, proven to not work.] The book also mentions that there is nowhere near enough data sharing between security industry, where the problems are, and academia, where - supposedly - the brains are.

Other reviewers have pointed out that it is not clear what is the audience for the book. Many of the chapters seemed written for the “curious consumer” while others are clearly intended for security practitioners or even security managers and imply a degree of IT industry savvy.

Finally, I have to say that multiple mentions of McAfee did not annoy me at all. I fully realize that if somebody employed by the vendor criticizes the very livelihood of that vendor (classic signature AV, in this case), you must throw your employer a major bone. You absolutely have to mention your employer positively to counterbalance the criticism and he does – in many chapters.

To conclude, I read books on information security for fun. This book was a lot of fun to read even if I did not agree with some of his opinions. It is well-written, has light writing style and touches most if not all controversial issues in security; the book also has a lot of fun novel ideas for the future to think about.

Dr Anton Chuvakin