Lies, damn lies, and book sales

I learn via Matt that the Tim O’Reilly has been posting his updated state on the computer books market. I have to confess I don’t quite see why those data are in such a high consideration as technology trend indicators. Other than seeing a value for book sellers (good for planning their business, indeed), judging technology adoption from book sales looks to me like doing weather forecasts by looking at the sky behind a curtain: good enough if you’re figuring out what to do with your umbrella, but that’s almost it.

Looking at Tim’s data, it seems Java is on a declining curve, while C# and .Net are on the rise. Apart from being a scenario I’m not experiencing (Java is clearly still the strongest player in the enterprise market), what we should read in that analysis is just that book sales for C# are stronger than the Java ones. It’s easy to explain that, though, if you consider the following:

  • Java is well established and mainstream. One can clearly imagine how the need for general Java books is declining, if you just consider the saturation effect. C# and .Net are still ramping up in the adoption curve, so it’s no surprise their book sales are better: people looking for generic books don’t hunt for Java anymore, as the Java section of their bookshelf is probably filled up already;
  • Java is nowadays about the “long tail”. The neat stuff isn’t big anymore: there are a lot of small gems  (libraries, frameworks, tools, plugins, applications, you name it) gathering a lot of attention, yet nearly none of them is big enough to justify more than a few tutorials, some articles, and maybe an instant book or two. The interest in C#/.Net is still about the language and the surrounding framework, and that’s book galore;
  • the book industry as a whole is clearly declining as more and more information gets on the Web. The small room left for books is about reference stuff and manuals that last for more than the time it takes to write, find, order, ship and read them. Why bother buying a book on framework X, when in six months time it will be more than outdated? Java nowadays is more about frameworks and user-space software/libraries than about the language itself, and that’s not stuff for books.

Given the reasons above, I’m actually quite surprised to see how Java is just going through a small decline which smells stabilization to me. While as a publisher I wouldn’t bet on many new Java books, as someone who needs to place a bet on technologies, those results are a clear sign of very good health for Java.



1 thought on “Lies, damn lies, and book sales”

Comments are closed.