Is JavaScript the Most Common Language in the World?
Daniel Rowinski, Editor-in-Chief at ARC (Application Resource Center) from Applause, wrote an outrageous editorial proclaiming that, “It’s Official: JavaScript Is The Most Commonly Used Programming Language On Earth.”
This is such an overwrought conclusion. StackOverflow’s developer survey is just one metric. It may well be anomalous. It is certainly open to interpretation and critique (for example, “survey analysis hurts women”). The survey purports to reach a whopping 0.4% of all developers around the globe. This is absolutely stunning. Not.
Extrapolating from 56,000 developers to glean some kind of insight for over 14 million people is quite a reach. For one thing, we don’t know what types of developers choose to participate in the survey and what types decline (for whatever reason). There are undoubtedly a few kinds of selection bias built into the survey.
We also know polls and surveys are unreliable. They don’t tell the whole story. Political polls, for example, have poor predictive power and are frequently wrong. Polls, and the StackOverflow survey, provide interesting indicators that can help guide you, but you shouldn’t make important decisions based on just one survey.
The truth is that most language rankings do not place JavaScript on top. IEEE Spectrum, which is based on 12 metrics from 10 data sources, places JavaScript at #8. PYPL puts JavaScript in #5 position. TIOBE says JavaScript is at #7. CodeEval ranks it as #6.
And consider this: web development is the hottest technology today. JavaScript is the native language of the web browser, so if you want to develop for the web, what choice do you have? It doesn’t make JavaScript “popular.” It makes it an unfortunate necessity; you simply have to hold your nose. In truth, though, you do have choices.
Redmonk’s language rankings are based on StackOverflow and Github numbers, too. JavaScript is on top because the number of lines of code committed to Github is a staggering 4 billion. (The number is greatly inflated by web projects in other languages; when you commit them to Github, you also commit a measure of JS code.) We have always known this, and it ought not to be surprising: web applications rely on JavaScript, and most people do not deliberately use JavaScript by choice. The choice is made for them. Sane people do not choose a broken language.
To be sure, many, many, many JS programmers ask questions on StackOverflow. This also should not be surprising. JavaScript is so perplexing that flummoxed newcomers naturally gravitate to the most common programmer resource available. In fact, this understanding was one of the foundations for Seerene’s controversial study released just two weeks ago: Which is the most complex programming language? (JavaScript!)
Also, The Hacker Life presents some interesting analysis about tech jobs in various kinds of companies versus programming languages. For example, in the San Francisco Bay Area, JavaScript is less popular with startups than with older, well-established companies. Startups generally favour Ruby, C, and Python. Even the older companies show a greater preference for C++, C#, PHP, and Perl! At least in the job market, JavaScript is nowhere near the most popular.
There is absolutely nothing “official” about Mr. Rowinski’s thesis. It is simply hyperbolic ignorance masquerading as intelligent insight. The best we can say is that StackOverflow’s survey is an interesting data point. We must, however, look for confirmation and good judgment to go with it.
