I just finished writing a engineering textbook, in the area of applied electromagnetics, specifically the Method of Moments. One of the key motivations was the need for a book with recent, up-to-date information in the field. However, my primary goal with this book was to alleviate problems I saw with other books in the field (as well as many journal papers): Many authors just don't understand or care about What Engineering Textbooks Should Be Like.

Having read through many good (and bad) engineering books, I'd like to first expand upon several points made by Born2rule, as well as add my own.

Key Points for the Design of Good Engineering Textbooks

  • 1. Standardized Variables: Most authors do a fair job at this throughout their own work, however if drawing upon another reference, the author should do their best to rewrite and integrate the equations using their own notation if possible. It confuses the student when different variables are used for the same quantity. This is especially problematic when the book comprises a collection of chapters from many different authors, or if it reproduces exactly the equations for a journal paper without adapting it for context and application.

  • 2. Thorough Explication: This is where in my opinion most authors simply fall flat on the job: They just leave out too much information. The worst thing a book can do is mumble "it can be shown that" and just write an answer down. This means "I am too lazy to show you or derive it myself." The author needs to show the student how it's done. Do the hard work and derive the equations, then summarize them in the text. Show the steps of the derivation. This is not spoon-feeding. It shows the learning student how problems can be solved, and provides them with insights into solving other problems of similar design. They will be able to pull out your book once they are practicing engineers and refer to the derivations for refreshers as well as new ideas.

  • 3. Showing The Money: Some authors omit key parts of an algorithm or the equations actually used to compute important quantities. Or they refer to a different book, or to "private communications." This is often done so that the knowledge remains with only a select few (the author) who can then charge their astronomical consulting fees to help you solve problems. This doesn't help the student, it pisses them off. It makes them mad that they paid $150 for your book. This is bullshit. This is the reason I will get piles of hate-mail from spiteful academics: I put it all out there. My follow-up book will be a collection of this hate-mail. It will be awesome.

  • 4. Words: Engineering books are by nature technical and obtuse. When learning a new subject, the student will be harmed by brevity. The author should do their best to provide appropriate context for the equations found in their book, which will help the student to flesh out the concept in their minds as they learn. Quantitative can proceed qualitative, but only if the reader already knows what you're talking about.

  • 5. Examples, Examples, Examples: Nothing helps a student more than well thought-out examples. These examples should explain all the key details and reasoning, as well as a thoughtful and thorough summary of the results. Figures and plots should be clear and to the point, and be free of extraneous data. Using the information contained in the book, it should not be unreasonable to expect that the student would be capable of reproducing the example for themselves (provided they have the appropriate software/computer/equipment, etc).

  • 5. Good Equation Typesetting: LaTeX is obviously preferable, but the author should make a decent attempt at well-formatted equations. Using the Microsoft Equation Editor is grounds for an immediate stab in the face.