I'd wish there would be a way to measurably gauge how far, "for better or worse", books have strayed away from the carelessly lousy online documentation from a purely linguistic point of view.
~
This book has almost no actual examples of anything real in its 300 pages. It just the online docs with some Windows specific stuff thrown in, some of it relatively interesting/esoteric like using IIS front ending TC and some other stuff, which to me does not merit space in a TC book, e.g., Securing file system on a OS level, in which the author states (page 219) "Windows gives you much more flexibility when assigning permissions than Unix does" I am still wondering when, if ever, Windows will be able to achieve the level of security that hardened gentoo sports right from a live CD
~
Matthew Moodie's book, in addition to the very Windows-centric mindset, also contains quite a bit of bluff and insipid content, sections that appear to have no meat whatsoever like "Transactions and Distributed Transactions Support". I wonder why didn't they just leave this section altogether. It also has many little mistakes that at some point I simply started underlining with my comments next to them:
~
page 4: "Tomcat is always the first server to provide the new features of the spex when it is finished" Actually not true, Jetty has been better at that that TC itself
~
page 59: author refers to the "commons" branch in TC directory structure while talking about logging listing 4-7 (TC 6 did away with it)
~
page 161: "Using Tomcat's connectors"/"The Workers" it should be "CATALINA_BASE1", "CATALINA_BASE2", ...
~
Book also seems to be talking more about TC 5.5 than the new features of TC 6.0
~