My hunch is that the book sales is relevant metrics – a decrease means less learners, less “new entrants”. Probably Java is preparing for a smooth landing in the coming years. Which doesn’t necessarily mean it will be “replaced” with PHP/Python/Ruby. Internet is the fastest growing area in software today, and Java isn’t even considered in most of the cases.

But I think there’s still a chance for Java expansion in the mobile world.

And yet, those statistics are worthless. The question is “is Java getting more unpopular, now that the books sold by O’Reilly are going down?”. From Probabilistics 101 we have conditional probabilities …

P(unpopular|books sales dw) =
P(unpopular * book sales down) / P(book sales down)

But the trouble is that the two events MAY be independent of each other, and that could mean that …

P(unpopular|books sales dw) = P(unpopular)

This can be solved elegantly with the Bayes theorem, which would need the answer to this question … “how many times was the O’Reilly Radar right before?”

But I have a hutch that Java is mature enough that it doesn’t generate book sales anymore. And it won’t be replaced by Python or Ruby because it has different use cases (which happen to overlap).

That may be right in terms of numbers, but when it comes to quality Python has some really excellent online resources: Mark Pilgrim’s “Dive into Python” (http://www.diveintopython.org/), the Python Cookbook (http://aspn.activestate.com/ASPN/Python/Cookbook/) or the online documentation on the official website (http://python.org/doc/), which I, for one, have found very good. And there’s the also the mailing list (http://groups.google.com/group/comp.lang.python/topics) which I’ve heard is more newby-friendly compared to other programming communities.

http://www.indeed.com/jobtrends?q=java%2C+c%23%2C+javascript%2C+python%2C+actionscript&l=

No growth, but no decline either.

The books just aren’t good enough quickly enough to keep up with the technology anymore, and online info/blogs/searching has improved to the point where it’s generally more effective for me. So the few books I do buy tend to be for late-night just-for-interest reading about things I’m not actually doing. If anything my own book-buying tends to indicate the exact opposite of which technologies I’m learning.

Thus the stats may be true about books sales, but don’t necessarily mean anything with respect to technology trends.

C# has many forms – pure C#, ADO.NET, ASP.NET etc. In case of Java there are god amount of online tutorials/documentation for the many Java frameworks.

The stats is in line with my expectations.

http://stephan.reposita.org/archives/2007/05/24/books-as-a-success-indicator/

“So what O’Reilly really says is that they have interesting books for PHP and C#, but less interesting books for Java, Javascript and Python.”

Nothing has changed since then.

Peace
-stephan