Programming Language Usage Statistics

A statement like

: C++ has become not an easy language to work in. Nevertheless there are a great many C++ programmers and developments out there - more than for Smalltalk, Java, Lisp and Perl combined.

always makes me wonder on what statistic it is based. Not that I question the statement, but that is from my personal experience and the environment I work in. From time to time I run into situations where I get asked for numbers if I pose such a statement. So I am looking for a good source for ProgrammingLanguageUsageStatistics.

For example, an interesting number from the C++-FAQ (http://www.cerfnet.com/~mpcline/c++-faq-lite/):

: The number of C++ developers increases 20-30% every year. You can imagine that five people are becoming C++ developers while you read this FAQ.

Where does that number come from?

--WolfWolfswinkel

I also wonder what relevance there is in such statements (except when you're trying to recruit and need to know that there is a sufficiently large programmer base to choose from I guess). The important thing is what is the appropriate language for the job at hand.

That said, perhaps IDC will have statistics on this ? --ChanningWalton

I can think of two places where I've seen statistics of this general sort. One is at http://www.perl.com/pub/1999/08/onion/talk1.html : a talk by Larry Wall that discusses (among other things) the demand for different languages as measured by the number of job adverts on some online site or other. The other (now probably rather out of date) is in Richard Gabriel's PatternsOfSoftware, which contains an article called something like "The FutureOfProgrammingLanguages and the end of history", which gives some figures and concludes that no language since C actually gets used much :-). --GarethMcCaughan

The above URL contains a chart containing the demand for Python, Smalltalk, COBOL, JavaScript, VisualBasic, Perl, Java and C++. What surprises me at first glance, is the absence of DelphiLanguage. --WolfWolfswinkel

Alas, while the good talk is still there, all the pictures are currently offline -- 2004-12-21

42.3% of all statistics are made up on the spot. Including this one.

From Application Development Trends this month: (which month / year?)

"According to IDC, out of 13 million developer seats worldwide in 1999, Visual Basic had 7.2 million vs. 3.6 million for C/C++ seats and 1.3 million Java seats. By 2003, Visual basic should still be ahead with 7.4 million, but the race will be closer with 5.2 million C/C++ developer seats, and 4.4 million Java seats."

You can measure ProgrammingLanguagePopularity not only by how many job postings list a language, but also how many books are being written about it, how active websites, usenet newsgroups and mailing lists are, etc.

In some cases the newsgroup is decieving: the maintainer of the ObjectiveCee newsgroup's FAQ hates Apple, yet programming for Apple's WebObjects and CocoaFramework are the most common uses of Objective C. Postings in that newsgroup turn into flame wars very quickly, and people really interested in ObjectiveCee have to go elsewhere -- usually Cocoa mailing lists.

measure language popularity ... by ... how many books are being written about it, how active websites, usenet newsgroups and mailing lists are, etc.

Possibly. But some languages are inherently overly complex and difficult to work with. That's why they need a whole shelf full of 400 page novels to explain them.

I have this nagging fear that as I go about ChoosingTheRightLanguage (PickTheRightToolForTheJob), I'm going to overlook the IdealProgrammingLanguage. Since the IdealProgrammingLanguage will, of course, be easy-to-read, intuitive, etc., it only requires a thin book to describe it. And since it's so easy to describe, that first book (or perhaps the second) will be a jem-like work of art, and it will never even occur to other potential authors that they could possibly write anything better.

And programmers lucky enough to stumble across that language will quietly churn out quality code, while unlucky programmers fill up newsgroups and mailing lists with bug reports, requests for help, mis-guided attempts to help, work-arounds, etc.

So I fear that really excellent languages have low visibility, easy to overlook.

-- DavidCary

I compiled some statistics, and added a bit of commentary and conjecture. The results are here: http://www.langpop.com

- DavidWelton