If you're still with me, however, this book is absolutely plagued by the disease known as "technical writing". It's long been known that those who deal with computers and electronics on a day-to-day (hour-to-hour?) basis are not often the most linguistically inclined individuals. This "urban legend" is manifested predominantly in the work, taking its form through constant usage errors, many spelling errors, excruciatingly awkward prose (at times), and, although this is more of a non-issue considering the subject matter, just a little too much bias in one direction.
Just to show that I'm not making this up, Glyn Moody frequently refers to "X Window", rather than "X Windows" (even though that's technically incorrect), "XFree86", "X11", or the "X Windowing System". Other similar, subtle annoyances occur throughout the book, but make no mistake: they don't obscure Moody's points indecipherably, they just annoy. One of the sentences that forced me to question Moody's bias was from Chapter 11: "If the history of Microsoft shows anything, it is a dogged determination to improve its often inadequate first attempts at writing software, and Internet Explorer is no exception." This sentence, inserted just after describing Microsoft's assertions to the U.S. Department of Justice as "shameless", leaves the reader no choice but to second-guess Moody's intentions.
Is the book complete? Yes (at least, you won't find a more complete book around). Is the book perfect? No, and due to the frequency of grammatical and otherwise editorial errors, I have to reduce my would-be 5-star rating to 4. Do yourself a favor and read this (but if Perseus would release a corrected second edition, that'd be just fine with me :D).
As a side note, I'm a user of Linux and a supporter of the open source model. When you do read this book, you'll be forced to form an opinion of your own on software patents and the whole lot. I encourage you not to take the RMS standpoint of "free [open] is better, always", nor do I encourage you to take the Bill Gates standpoint of "protected code is better, always". Draw a useful parallel between the two opinions.