Programming Languages Are Not Zero Sum

From Wikipedia, the free encyclopedia:

In game theory and economic theory, zero-sum describes a situation in which a participant's gain or loss is exactly balanced by the losses or gains of the other participant(s). If the total gains of the participants are added up, and the total losses are subtracted, they will sum to zero.

Being advocate, implementor, tester and co-designer of a new programming language, I often hear objections along the lines of you are killing $other_programming_language, combined with a mixture of fear and resentment. People are afraid that having a new player on the market will decrease market share of their own, favorite programming language.

While I can understand these thinking patterns, there is no reason for concern. The market for programming languages is not a zero-sum situation. While I don't have hard data, I have the impression that the programming job sector is growing, and the US government expects it to grow further, too.

Certainly the growth of world population sets a rapidly increasing baseline, and even if we assume a constant percentage of all people related to programming in some way, the total number of programmers rises, and will continue for quite some time.

(I'm pointing to some resources about programming jobs, and I fully realize that it's not the same as number of overall programmers; but it's easier to get data for jobs, and I do think that the general trend statements are true for both).

So as long as the total number of programmers increases, a decrease in relative market share doesn't automatically mean a loss. In fact the job trends show an increase for "scripting" languages, and while Ruby is certainly the winner in terms of growth, Python, Perl and PHP win too!

Non-job data shows for example a noisy but steady growth of uploads to the Comprehensive Perl Archive Network (CPAN) -- data from a programming language that is often perceived as a loser of ruby's and python's success.

A recent Linux distribution trend analysis fell into the same trap: it shows relative numbers of search terms, and talks about a decline for all distributions except Ubuntu. Again I don't have hard numbers (the mirror infrastructure of most Linux distributions makes it nearly impossible to get accurate download counts), but I haven't seen any evidence that total usage numbers of any of the Linux distributions actually decreased.