Å·±¦ÓéÀÖ

Not what I expected, but still very relevant. I expected something very academic and mathematical. The author claimed many times that his principles were "axioms", and that they were pristinely mathematical in nature and all self evident. This was a rather annoying claim, since the book was not mathematical at all, nor were the axioms necessarily self-evident (though good supporting examples were provided). Despite this, it all still rings perfectly true. A system can be a blessing or a curse, but it is guaranteed to have unexpected behavior. When it does something bad, you'd better hope that your system is flexible, changeable, monitorable objectively somehow, that it doesn't completely dominate everything and only allow positive feedback.
The style is kind of a mix of Taoist philosophy, design, a tiny bit of math (really, barely any), common sense, self-improvement, endearing Latin textbook, and more. Lately I'd been thinking about all of the generalizable things I'd learned from programming (and especially my strengthened dislike of large bureaucracies): that things need to be flexible and interchangeable, testable and constantly tested, designed with user experience in mind, tested before deployed, etc. All of that is abstracted out of software design and into the real world with this book, which is really quite phenomenal. Software happens to be just one type of system.