The book starts off with a mild surprise: its definition for what “legacy code” is. It is defined as “Code without unit tests”. A more appropriate title for this book would’ve been “Making Legacy Code Unit Testable”. Luckily, my colleague warned me beforehand, so I knew what I was getting into. So I started reading with adjusted expextations, and the book lives up to them. It inspires to take a new look at the legacy code I’m currently dealing with, explains pretty well what refactoring- and dependency breaking techniques are available. Some chapters I skimmed (C macros for unit testing, anyone?), but on the whole: recommended!