The Relative Strength Index for Programming Languages (RSIPL) is a metric I devised and first published starting in March 2017. RSIPL was created after reading several similar articles and lists, most notably, the RedMonk Programming Languages Rankings. The ranking methodology RedMonk uses is based upon popularity of the languages on Github (measuring “code”, as they put it) and StackOverflow (measuring “discussion”). As good as this list is, I feel that it is perhaps less timely than it could be, and that it is missing a third critical piece of the puzzle: job postings.
I am sure there are many who look at these lists and wonder what the potential is for landing a job if they invest their time and effort into learning one or more of these languages. Because of this, I felt it important to reflect this in the score. By using job postings as a criteria, the list reflects what I feel is a pragmatic view into the popularity of languages.
Another important note regarding RSIPL: the rating attempts to reflect popularity based on current activities, rather than cumulative popularity over the years. To achieve this, I looked at recent Stack Overflow tags, recent GitHub activity, and recent job postings from both Indeed and Dice. The job postings were blended together to form one aspect of the score.
So, to summarize, this ranking is intended to reflect very recent activity from the perspective of Open Source code activity, tagged discussions, and job postings. Each criteria was normalized on a scale of 0-100. Results were then averaged for each language.
10 Most Popular Programming Languages
Following are the top 10 languages in March 2017, using the RSIPL formula:
While I feel RSIPL does a better job at capturing current popularity of programming languages than other lists, it is far from perfect. Although job postings do add a meaningful aspect to the analysis, it is somewhat flawed, because job postings tend to lag behind the latest and greatest trends, as industry is typically slow to respond to change. In addition, there will always be jobs related to corporate legacy systems that will also skew such ratings. But to counter that argument, even though there are jobs available for legacy technology, they are still jobs nonetheless, and they still do reflect on the popularity of a given language (so COBOL programmers take heart!).
I intend on posting RISPL ratings over time, perhaps monthly. I also intend to revise the index as better data becomes available. Hopefully it is helpful to some in its current form.