Tidelift's annual managed open source survey explores how technologists use open source to build applications at work. Over 600 people shared how they use open source software today, what holds them back, and what tools and strategies would help them use it even more effectively.

In this post, we share the sixth of nine key findings. If you don’t wait to wait for the rest of the results, you can download the full survey report right now.

Seeing your favorite language gaining popularity is professionally affirming. The RedMonk Programming Language Rankings, TIOBE Index, IEEE Spectrum Interactive Rankings, The State of the Octoverse, Stack Overflow Developer Survey—all use different methodologies to measure things like attention, job opportunities, and the prevalence of new code being written. 

None will tell you which language is functionally better, but they provide actionable insight into 1) which languages you should be learning and 2) which languages you should be using for certain types of projects. 

We wanted to add some meaning of our own to this crowded pool of data with a few questions about the programing languages technologists rely on most. We started by asking respondents to select the top open source languages their organization relies on, allowing them to choose up to five languages.

Not surprisingly, the top three languages by far were JavaScript (78%), Python (52%), and Java (41%). In fact, the top five languages reported by our respondents exactly match the most recent RedMonk language ranking from June 2020, with PHP and C# (.NET) ranking fourth and fifth.

Go ranked higher in the Tidelift survey than in RedMonk’s analysis (7 vs. 15), while Ruby ranked lower in the Tidelift survey (9 vs. 7). At the bottom of the Tidelift list were up-and-coming languages Rust (6%) and Swift (7%), which are often used for mobile app development.

RedMonk’s analysis looked at over 50 languages, of which eight ranked higher than Rust, which by some measures has never been hotter, with TypeScript showing up at #9. The next time Tidelift asks this question, we expect to include TypeScript and Kotlin in the list of languages, as both received many write-in responses in the 2020 version of the study.

Fans of Java often complain that it is underrated in language studies because its use is concentrated in larger enterprises. The survey supports this conventional wisdom, as 66% of organizations with more than 1,000 employees rely on Java, which moves it ahead of Python among this cohort.

Although Python is popular among students and hobbyists, our data shows that larger organizations are more likely than smaller organizations to rely on it (61% of organizations with more than 1,000 employees vs. 49% of organizations with less than 1,000 employees).

It is important to note that larger organizations selected more languages. Organizations with more than 1,000 employees on average identified 3.5 languages, while those in organizations with 1,000 employees or fewer only chose 2.9 languages. Since larger organizations have more applications, it is not surprising that they rely on more languages.

We also wanted to understand how critical each of these languages is across an organization’s applications. Respondents were shown the languages they had chosen in the previous question, and asked what percentage of their organization’s applications relied on each language.

JavaScript stayed in first place, with the average organization using it in 57% of their applications. Java jumped to second place at 54% since it is often an integral part of larger enterprises’ internal applications.

Most notably, C# (.NET) and Ruby improved significantly by this metric at 50% and 41% respectively. Even though the sample sizes are smaller (100 respondents for C# (.NET) and 74 for Ruby), this tells us that the organizations that rely on these languages do so quite heavily.

The other outlier here is Python, which was the second most relied upon language overall, but dropped to sixth in terms of the percentage of applications that rely on it with an average percentage of 43%. In keeping with its dual role as a primary language and a glue-layer for smaller projects, Python is used by a lower percentage of applications than JavaScript and Java.

Older, entrenched languages battle for developers’ attention against new languages with new approaches. The two metrics we are tracking—top languages being used and percentage of applications using each of those top choices—provide valuable data technologists can use to make decisions on which languages are most established, which are gaining momentum, and which are losing momentum.

Want the full survey results in one report? Get them here now.

Read more about how we conducted the survey, see the survey demographics, and learn why we call it the managed open source survey.