The trouble with book reviews

Book reviews - what do you look for?

Jon Skeet: Coding Blog — Sun, 14 Sep 2008 15:37:32 GMT

I've just started writing the book review for "Pro LINQ - Language Integrated Query in C# 2008"

re: The trouble with book reviews

skeet — Fri, 18 Jul 2008 20:08:38 GMT

(cc'd by email)

What makes me read a lot of stuff? Well, at the moment reviewing other books - I don't actually read many technical books usually (and sadly).

There's also the fact that I spend about 3 hours travelling into and out of London every day. That makes it easier to find time to read :)

For threading, I'd recommend Joe Duffy's upcoming book "Concurrent Programming on Windows".

You might want a more gentle introduction as well though. I've got an article at pobox.com/.../threads and Ben Albahari has one at www.albahari.com/threading

Not sure about advising how to read stuff and practise code examples, other than just to do it. When it comes to trying something new though, I do find it useful to only try one thing at a time. For instance, experimenting with WCF, WPF and LINQ to SQL all at the same time wouldn't be a good idea - if something didn't work, you may not know where to look, or be able to easily tweak other aspects to make debugging easier. I personally like console apps for experimentation with anything other than GUI toolkits.

Hope some of this is useful...

re: The trouble with book reviews

Kalpesh — Fri, 18 Jul 2008 17:43:46 GMT

Jon,

I buy books, but don't read them till the end :)

Could you advise me on how to read stuff & practice code examples (I know, I am acting like a baby).

I have found that it is better to buy books based on reviews by expert in the field. For r.g. I have seen several postings on csharp newsgroup & learnt a lot from it. So, it becomes my personal idea that book from you will have some solid content. Also, it is enforced by other good people in c# (Patrick, Ben as you said)

Blogs are more interactive, where author can correct mistakes, take feedback, improve stuff.

Books have been 1 way communication so far. I don't know how many people go & look at Errata (Errata will be really helpful for technical inaccuracy)

On a side note, I am a developer for quite sometime & have ignored threading for long. I am kind of a newbie on that end. Could you suggest me a book to start with on that topic (c#/.net flavour would be helpful)?

BTW, I have bought your book & have yet to start reading it. What makes people like you read a lot of stuff?

You can write to me at my personal email address, when you post back (shahkalpesh at gmail d0tc0m).

Thanks :)

re: The trouble with book reviews

Ian Griffiths — Thu, 10 Jul 2008 15:57:29 GMT

I'd add one point to your list of things to watch for when reading reviews. And it's particularly relevant to Amazon reviews:

Watch out for reviews from people who haven't actually read the book.

This phenomenon occurs because Amazon appear to offer some sort of reward system for people who post reviews. This encourages people to post as many reviews as possible, which is of course a lot easier if you skip the whole 'reading the book' part of the review process.

For example, if you go to the US Amazon site and look at the reviews for .NET Windows Forms in a Nutshell (my first book), you'll see a review from Phil Lee. If you're familiar with both Windows Forms and ASP.NET and you read what he has to say, it will be apparent that he doesn't realise these are two different technologies - he thinks that there's just "DotNet Forms".

So he dings us for spending only 6 pages on the data grid, while Dino Esposito spends almost half a book on the subject. Well yes, that's because the DataGrid that shipped in v1 of Windows Forms was a tiny, and almost useless thing (so much so that Microsoft felt the need to replace it entirely in v2), and there really isn't much more than 6 page's worth of stuff to say about it. Dino's book is about ASP.NET, which had a much more fully featured data grid right from v1.

It's clear from his review that the guy thinks my book was about ASP.NET. And since the title alone indicates that this is not the case, it obviously doesn't even have the first clue what Windows Forms is.

And I'm pretty sure my book wasn't *so* bad that you could possibly read it without realising that it has nothing whatsoever to do with the web...

So it's always worth checking how many reviews the reviewer has posted. This Phil Lee guy has written some 70 reviews, but I wouldn't trust any of them. And while it's plausible that he could have read that many books in the time he's been reviewing (it's only about 10 a year), he evidently did nothing more than read the ToC (and apparently not even the title...) of my book.

I also found the claim that the reference section "probably duplicates what's available on a MSDN subscription CD somewhere" pretty offensive too. He doesn't even pretend to have bothered reading that. I didn't write that half, but I read all of it, and my co-author's philosophy was specifically to provide hints and tips that are not in the MSDN library. The whole point of that section is it provides information you don't get in the standard reference material - otherwise what would be the point?

Grr!

re: The trouble with book reviews

skeet — Wed, 09 Jul 2008 18:53:08 GMT

@Ying: I'll look at reviewing my own book when I've done the others - i.e. not for a while. I'll need to actually reread it though...

@Kevin: The review of Joe's book will probably be after I've got hold of the real thing in hard copy. I think I'll find that much easier to read in full. At the moment I've only got a soft copy, which isn't letting me read quite as comfortably.

re: The trouble with book reviews

Kevin Kerr — Wed, 09 Jul 2008 17:16:27 GMT

I'm looking forward to the Concurrent Programming on Windows review. I've read the book (rough cuts), and would like to hear your take on it.

re: The trouble with book reviews

Ying Jin — Wed, 09 Jul 2008 16:08:08 GMT

Hi, Jon,

We would love to read your own opinions on your in-my-opinion very accurate and holding-those-greasy-c#nerds-hand-pretty-well book.

And as for reviews, the more, the merrier, especially with books about applying technologies. I would put most of software books I read into this category, the books that help people to digest a challenging technology or understanding a solution to a challenging problem. For these books, as you put it, accuracy and teaching styles rules. But both can be hard to define for these books, as they are not spec or scientific papers. The two actually intermingled. For some audience, it may be an engaging teaching and good enough accuracy, but for some other demographic, it can be too wordy yet too loose on accuracy. I guess at the end of the day, best way to judge is to have more reviews from all spectrum of the audience to get a good feel of it.

So everyone writes up and everyone benefits.

Ying

re: The trouble with book reviews

skeet — Tue, 08 Jul 2008 18:44:02 GMT

No, it certainly doesn't mean that my book doesn't suffer from the negative points! Obviously I'd like to *think* that my book is pretty accurate, and most of the errata I've collected so far (csharpindepth.com/Errata.aspx) have been non-technical, but I'd be very foolish to think that there aren't more issues lurking.

Likewise it's even harder for an author to know how effectively they're teaching than it is for an "expert reader". More opinions are always very welcome!

I'd love to see more authors review my book - Joe Albahari has already done so (msmvps.com/.../guest-post-joe-albahari-reviews-c-in-depth.aspx) and so has Patrick Smacchia (codebetter.com/.../book-review-c-in-depth.aspx) but I'd welcome more.

Some authors certainly don't want to publish reviews on related books for fear of either displaying bias or being perceived to do so. I respect that opinion, but my personal view is that so long as the review declares the relevant interest to start with, it's okay.

One advantage that an author of a related book has is that they're very likely to be experts on the subject matter - chances are they've been poring over the language specification (or whatever) more than they would have done otherwise. That's one area where I can add value to a review - I'm certainly a better judge of accuracy than I would have been a year or so ago. I hope that makes up for losing a certain amount of impartiality.

One interesting question is: would you like to see me review my own book? It would have to be brutally honest, of course, otherwise it would be pointless. I'm not yet sure whether I could actually bring enough objectivity to it to make it worthwhile, but if people are interested I might have a go.

re: The trouble with book reviews

Abdu — Tue, 08 Jul 2008 18:21:30 GMT

I have reservations about an author who reviews other books on the same subject. So does this mean your book doesn't suffer from any of the negative points you mentioned?

Now I want to see reviews from these authors on your book. (Might never happen)