I like Joel. Well, I’ve never actually met him, of course. I mean I like his writing. I’ve read much of his website, I subscribe to his RSS feed so that I can see his new pieces as soon as they’re “published” and I’ve bought his other books even though they’re just rehashes of stuff that’s already on the net. That may seem a little crazy, as though I have more money than sense, but some things are much easier to read on paper than on LCD. And his writing is easy, humorous and engaging, making it worth dipping into occasionally.

So, cutting a long story short, I bought “Smart and Gets Things Done.” So what’s it all about? Well, one suspect is that it’s an advertisement for Spolsky’s software company, Fog Creek. But, in fact, it’s a book about attracting, recruiting and keeping super-star programmers. Indeed, chapter one is why you even want to hire top developers when you can hire a decent developer for rather less money and effort.

He covers the whole process, right from where to find your next recruit; how to sifting through the large number of CVs (resumes) that you’ll probably receive; how to interview people; and how to keep them once they do join. It covers a lot of ground in a couple of hundred pages and he almost makes it sound easy.

Joel is very much against the Google and Microsoft approach to interviewing. He (rightly I think) points out that asking brainteaser questions only finds out whether people can do… brainteaser questions and tells you little about whether they are decent software people or whether, as the title of the book suggests, are smart and gets things done. He spends some time discussing the title, and this is something that jells with my experience. For example, you don’t necessarily want PhD’s, as they are often smart but spend a lot of time thinking about theoretical problems rather than doing anything about it — clearly not the kind of person you want when deadlines are looming. In another chapter he notes that you really don’t want to be forming opinions of people before you meet them. Or, put another way, suggesting that PhD’s might not “get things done,” is not a brilliant idea.

In summary there’s plenty of good stuff to see. But, as I mentioned earlier, much of the content has come straight from his website. That was also the case for his previous book, “Joel on Software,” and isn’t necessarily a criticism in and of itself. However, that previous book was intended as a collection of short, separate essays tied together by a common, fairly broad theme — software engineering. In the sense that each chapter was distinct you could reasonably dip into it, reading one section but randomly skipping over another.

This book, on the other hand, is supposed to be on recruitment — a much narrower subject — and the chapters follow a kind of trajectory. Unfortunately the essays on the website generally work as stand-alone pieces, so when you bunch them all together in a single book and read them back-to-back you find that there is quite a lot of repetition. If it was all a thousand pages long and a recap was in order then that might make sense, but “Smart and Gets things Done” is only a couple of hundred pages long.

So overall it’s a nice book. It has a lot of good advice, even if some of the suggestions are not achievable by the typical employee. As is generally the case with Spolsky it is entertainingly written and is engaging, witty even. However, given the length of it and the fact that there is a considerable amount of overlap you may be better served by reading it all on his website.