Customer Reviews

The Art of Designing Embedded Systems (Edn Series for Design Engineers)

5 star:		(4)
4 star:		(3)
3 star:		(2)
2 star:		(0)
1 star:		(3)

 

 

In Mr. Ganssle's book he states, "Engineers are notoriously poor communicators..." and some of that truth shows in his book. The chapters, sections, and appendices are well organized but the content in them is not. The writing style has a tendency to ramble like a friendly conversation, sometimes deviating into tangential material. However, the experience and wisdom held in the book's content (even the tangential parts) are its greatest prize. As an embedded developer with seven years of experience, I found myself frequently relating to the author's own experience. When I read the sentence, "We've got to avoid quoting a long, arbitrary time impact as a knee-jerk reaction to any change request." I actually laughed out loud. Mr. Ganssle does a good job of contrasting a disciplined, well-documented development process versus a panicked, haphazard one and he explains the trade-offs between the two by quoting respectable studies on the subject. The material on writing and debugging good embedded code is all stuff that most folks will, hopefully, learn after few years on the job. But, if you're just starting out, or if you're looking for a better way to do your job, I recommend that you read this book.

This text is obviously a distillation of many years of experience in the field. I have programmed enough ISR's and programmable devices to know that it is very good advise indeed. The book, however, is not for the beginner. His assumed audience are people who are already in the field and would like to take advantage someone with greater experience. The book is loaded with hard learned nuggets of wisdom that are less likely to be found in classroom textbooks. He has something to offer both software and hardware types(I'm software). For example, the advise about software partitioning, ISR development and debugging, and device programming is excellent, and for hardware types the advise on building boards that are both software and hardware debuggable and the myriad tips on finding problems from the common to the obscure will be appreciated. He also provides suggested project management guidelines that help in getting functionality operating as quickly as possible, and, from my point of view, guidance on how to deal with irrational schedules and managements that are reluctant to provide adequate project resources. I appreciated expecially his citing research that quiet private offices provide the single greatest boost to engineering productivity. Probably the only thing I found lacking in the book was a Bibliography of other book references and useful web links.

This is a really poorly written and poorly organized collection of ramblings from an embedded systems "expert." Each topic is covered at a grossly superficial level and most sections consist of short random thoughts. Here's a typical paragraph (from the 2 pages on real time operating systems).

"Improvements to these tools come almost daily. Keep on top of the field to avoid the fate of the dinosaurs."

Speaking of dinosaurs, the author's favorite tool for debugging and performance analysis seems to be the oscilloscope.

I get the feeling that the author "wrote" this book by talking into a tape recorder while in an LSD haze on the way to Las Vegas.

BTW, the page count is 243 and not 352 claimed in the advertising for this book. Thankfully, the editors saw fit to trash at least a portion of this book.

My only regret in writing this review is that the Amazon rating service doesn't offer the option of giving 0 stars.

Like Ganssle's other books, this one is best suited to those working on big teams and/or for big companies. That's clearly his background and much of his writing and suggested methods reflect big company bureaucracy. His writing is also generally geared towards larger 16 and 32 bit embedded projects with external memory versus smaller self contained MCUs. Most examples are x86 or 68k based.

Ganssle's style is humorous at times, and generally easy to read. Sometimes he exaggerates, however. For example, he makes interrupt routines sound like the hardest thing on earth. This book is far from being comprehensive but contains some solid advice--especially for those managing large high-budget projects.

This may NOT be the best book f you're a student, a one-man-show, work on a small team, or are using a small single chip MCU (ditto for Gannsle's other books).

If Mr. Ganssles's embedded disign skills are even slightly inversely proportional to his writing skills, then he undoubtedly is one of the true experts in the field. This book has a few small pieces of interest but the cost of getting them is a bit high. Take this book to read at the dentist's whil you are getting a root canal. You won't even notice the drill. Almost zero technical content for anybody who wants to learn how to make embedded systems. Mostly an eclectic ramble about Mr. Ganssle's view of management, engineers, software development, and life in general.

There are many embedded books that will take you real deep and never let you come up for air - this one helps you stay afloat (and awake). Yes, the book organization may not seem to be in perfect order - I have only found a few that were. But, the main difference that sets this book apart is the insight to problems in designing a system - especially the bug's that might creep up later in software. Here, the author gives his experience on how to design embedded systems not build them (there is a difference). Practical advice is presented that may save time, your system, and headaches.

Though this book is not very big, it is filled with gems that apply not only to embedded systems programming, but to general software engineering practices. Ganssle's writing is readable, humorous, and very short on fluff, as readers of his email newsletters and ESJ writings have come to expect.

This is a very good book... but it is meant for people with experience in the industry.... the book is about industry and the dynamics within it...

not for "student" who are just in college/univ.

excelent book once you've had at least 6 months industry experience

In a perfect world managers grant unlimited budgets to designers, upgrade development tools as soon as a new tool is released, managers accept the answer "It will be done when it's done", and every prototype works the first time. In Jack's world (the real world) the embedded systems designer must deal with facts, experience, and limited resources to get a product to market. That's what this book is about. The book does not arm you with sure-fire advice to persuade your boss to purchase those new development tools, nor does it contain "how-to" advice, code snippets, or schematics. The book does contain a collection of solid advice that Jack has accumulated during his career as tool-vendor, developer, consultant, and other positions associated with embedded systems design and development. There are "musings" from a collection of areas including rational for developing time estimates for firmware schedules, troubleshooting, code organization, and advice on how to improve your approach to development. I was particularly impressed with how Jack's approach to scheduling based on many studies to justify his advice. In fact, much of first part of the book applies equally to system software development. How many times have you been asked for a schedule estimate knowing fully that your boss has already determined the number waiting for you to commit to that number or have to pull a number out of a hat to satisfy the boss? Jack discusses solid advice on how to arrive data-driven estimates.

I felt the book deserved 4 stars primarily because the book is a little out of date, but is otherwise fun to reads, interesting, and introspective.

The chapter titles are:

1. Introduction
2. Diciplined Development
3. Real Time Means Right Now
4. Firmware Musings
5. Hardware Musings
6. Troubleshooting Tools
7. Trougleshooting
8. People Musings
Appendix A: A firmware Standards Manual
Appendix B: A Drawing System

The author of this book has lot of practical experience in the field of embedded systems, and for the embedded system programmer with 1/2-2 years of experience I think this is a must for reading. Yes the book is unstructured, but it is filled with usefull practical advices. With respect to debugging I just think that the author underestimate the power of thinking and experience and therefore praise the value of ICE's, which I mostly consider as a tool for beginners, where one of the few exceptions are debugging ISRs with trace.