Author thinks "octet" has something to do with "octal" and claims that "it's wrong that they use octet to refer to bytes because bytes have nothing to do with octal numbers, but oh well" in one of the early chapters. Then in later chapters, repeatedly makes snide comments about the use of "octet" instead of "bytes".
Todd, the reason for the use of "octet" has nothing to do with octal numbers. It's because historically bytes could be other than 8 bits. Some systems had 7 bit bytes. Some systems have 16 bit bytes. "Octet" is used to ensure that we're talking specifically about a grouping of 8 bits.
Several chapters are a chaotic mess of information presented out-of-order. For instance, in the Routing Protocols chapter, author repeatedly makes comparisons to OSPF before the OSPF section, making such comparisons empty and meaningless. Alphabet-soup acronyms are constantly being referenced several sections before their actual definitions. Whoever edited and proof-read this book is a moron.
Some quiz questions were inaccurate. For instance, one question shows an image with a crossover cable already present but the answer thinks its a straight-through cable. Another question asks what phrase describes network traffic that remains within a data center and claims the answer is "north-south" when it's in fact "east-west". Avoid this book.
The last seven chapters were 2, 3 pages each and did not cover all of the things they claimed to. Several sections were literally just copy/pastes of earlier sections. Then out of nowhere comes a chapter that's supposed to be about cloud-computing but instead rehashes the entire book, repeatedly copy/pasting several earlier sections. Filler, much?
Author also brags about deleting all of the partitions off of a student's harddrive to make a point, instead of just replacing the desktop image or something less harmful. "The student was angry at me, but I told him I had to make a point." What a dick move. There are better CompTIA Network+ books.
Oh, and by the way, in the subnet section, there is no need to subtract from 256 to get the grouping size. Just use the value of the bit position instead. 128, 64, 32, 16, etc. /26 is 64 because it's the 2nd most significant bit. Much simpler.