First of all, let me say, I think Sketchflow is amazing. Also, this book makes it clear that you can do very sophisticated things with the Blend UI before you have to slow down and code. Now for the very bad news.
This book held out ambitious promise. It promises to present how to use Sketchflow as a disciplinary backbone as you step through the sketching and prototyping phase of making things, such as software. And, I still believe Sketchflow can do that. But this book is a cut and paste catastrophe. You can literally see how they cut and pasted but did not fix sentences. The absence of the article "a" in very key places, was very distracting. Then in one caption there were additional "a"s. Critical words were left out of sentences, too. Explanations often complicated things that are just simple steps. You're designers! Don't you know when your work gets in the way of affordance? I hate it when tech doc writers think they're writing literature or even non-fiction and don't just number their steps. They're probably scarred they won't look smart if they do that. But smart tech writing is making functional users fast. (Lot's of Microsoft Blend blogs number steps and avoid Dynamic Prototyping's paragraphical convolutions). Luckily, I opened up Sketchflow and did a project before learning how to use it! So, I recognized, eventually, that they were attempting to explain something the UI just made very obvious on first use. STEPS!
I'd imagine I'm the only guy to ever make it to page 325 out of around 500, because there are no negative reviews. I also know a lot of people just read and don't do the hands on exercises.
I could forgive the writing sophomoricities, the worlds' darkest, tiniest screenshots, and stylish paragraphs that obscurred the steps, if the project steps, and the project files were actually correct. But, get this, a majority of the time, you are directed to project filenames that are different than those in the folder for that particular chapter. I reversed engineered. I set up my own files from scratch. I rewrote their sentences until I knew what they were trying to say, and I went into the actual code to fix things. During a chapter on data management, after hours of frustration, I found a blog promising to perform the same tasks. It had 3 times as many steps, because it ran for the code, but it all worked perfectly! Is that how you learn to do projects in expression blend? Random blogs?
This book and I are like a couple in an abuse cycle. I'd finally finish a chapter badly bruised and wondering if "this was it". And then the next chapter's introductory paragraphs would get my hopes up and the honeymoon was back on. An ever-so-brief honeymoon, abruptly ending when I'd try to open the next project file, or would follow the books steps which turned out to be wrong, and again I was fighting back, web-searching, blog reading, re-reading the paragraphs, restarting the files, failing back to older files, reinstalling the files, etc.
My friends keep telling me to give it up, and keep asking me why I can't seem to stop myself. But there are not plenty of fish in the sea. In fact, I've gone through a couple of much better written Blend books and the author's jump off into the code long before the UI requires it. Like, "How to make a button." Why would you want to know how to code a button to learn states, etc? Nobody does that. Blend comes with tons of buttons that can be easily customized. It's because the authors don't know how to use the UI to build cooler things than buttons, presuming that I will know how to take that button coding information and make really cool RIAs, really quickly without using the Blend UI--which is a contradiction in terms. This book at least excels at avoiding code to do what the UI can already do. It just fails to successfully present how they did it! Don't buy it. Find a better way to learn it, and then share it with me!
The last work files in the final chapters, where you're bringing it all together, WON'T EVEN LOAD in Blend 4. I don't know if it's a file compatibility issue, or what. By the time I was loading those, I'd quit reading the chapters becuase previous files were failing. Tragic. I'm puting it in the bookshelf tonight, the margins filled with my notes of frustration. "File didn't have this name. When I loaded the one that IS there, half the execises had already been performed on it", etc.
It's the only book of it's kind, and it doesn't work. So sad.
The well spoken wisened Bill Buxton wrote the preface. Bill, you're brilliant, your influence on how Blend and Sketchflow are built is fantastic. But why didn't you try the book's exercises before you associated yourself with it? You're why I bought it! I love your books, but in the future, please take the time to test technical books before you cost me[...] and 20 hours of frustration. Why is your time so much more important than my time, and the time of all the people who buy this book because of your endorsement?