This was a disappointing purchase for me, as I read some of the book on-line (in CMU-CL's "Encycmuclopedia") and was rather stunned at the beauty and possibilities of the MOP (which, in short, is defining the core object system itself in terms of the object system, allowing you to use the full power of the object system class hierarchy/relationships to control your object semantics). The book is a pretty straight forward implentation discussion, which is certainly nice as a case study in implementing such an interrelated system and boot-strapping the MOP into use, but it is only that. The MOP is one of those perfect ideas with such vast potential that I would much rather have seen actual expressions of that potential instead of mere inner working details. If I'd know that, I would not have bought it--but then neither will I be selling my copy.