[Translation] Average income developer depending on the programming language

Good afternoon, dear readers of Habr. How often do you look for the relevance of a particular platform? If you are a good specialist, then probably always. If it is not often, perhaps it is time to start doing it. We have prepared a translation of the study of the American blogger Ben Podgursky , which allows us to understand the level of payment for programmers, depending on the programming languages. Ben used as a source user data on github. This post scored a large rating on Hacker News.



There are a lot of stereotypes about the developers of various languages, but I was curious how they correspond to reality.

• For each repository, I used the evaluation of GitHub projects by the composition of programming languages. For example, this GitHub project values ​​as 75% Java.
• For each language, I collected revenue data for all developers who contributed to the project, which is at least 50% in that language (according to the above assessment).
• I filtered by languages ​​with> 100 available data points for income (it helps to understand the popularity of the platform - the more, the more popular).

Here is a table of income, sorted from lowest average income to highest one:
')


Here is the same data in the form of a diagram:



Most of the languages ​​in the sample roughly met my expectations:

• Haskell is a very academic language that is not famous for its generous salaries.
• PHP is a very accessible language, and young, less qualified programmers can easily master it.
• Java and ActionScript are used to a large extent in enterprise software, and, as is well known, enterprise software development is paid fairly well.

On the other hand, I am unfamiliar with some languages, such as XSLT, Puppet and CoffeeScript. Do you have any idea why these languages ​​took such places in the sample?

But before drawing conclusions from the above data, I want to note something:

• All these projects are open source projects (open source) that cannot provide an accurate estimate of the cost of work among developers of closed projects.
• Not all repositories are analyzed on GitHub, so some of the users are not taken into account.

Although these numbers may be inaccurate, I think that they are a good base and starting point when comparing the cost of developer labor by programming languages.

I will listen to all your comments and opinions about the methodology and results.

Yaroslav Bosenko, a Ukrainian start-up, helped to translate.
Thank you for your attention and for the positive feedback. Subscribe to our blog on Habré , Twitter (an aggregator of interesting news about Silicon Valley), Facebook , YouTube-channel to follow the updates. Previous posts can be found here .