Like many who have commented before me, Ms Harris' attempt at humor tends to get in the way of reading. Time is the essence; crisp, concise and clear is the way to go.
There are also a couple of mistakes that I have taken note of:
(1) On page 12, the question "Which item is not part of a Kerberos authentication implementation?" has no answer because the given answer "A message authentication code" is wrong -- RFC 4120 clearly shows the application of HMAC. Even without the RFC, authentication using symmetric-key crypto clearly requires a MAC.
(2) IPSec is more commonly written as IPsec. The second paragraph on page 760 is outdated, as ISAKMP is integrated into IKE since RFC 5996.
(3) In Table 6-3, page 443, not all Halon substitutes work by interfering with, or more precisely, inhibiting the chemical reaction. Inergen and Argonite for example work by displacing oxygen.
(4) On page 936-937, the definitions of DCL, DDL and DML are not really consistent with the database literature. For example, instead of "defining the internal organization of the database" (which sounds very vague), DCL is really for "controlling access to the data and to the database". On page 942, the definition of savepoint is also not precise. Savepoint is really a point where you can roll back a transaction to. See "Oracle Database 10g: The Complete Reference".
(5) On page 948, the definition of durability is not precise. This statement from Haerder and Reuter's 1983 ACM Computing Surveys paper is better: "Once a transaction has been completed and has committed its results to the database, the system must guarantee that these results survive any subsequent malfunctions."
(6) On page 1087, the phrase "As in RAID 1 striping" is wrong, as RAID 1 is for mirroring or duplexing, not striping.