A language poll can tell you what its respondents enjoy. It cannot tell you which language will make your particular project easier to build and maintain. The old embedded poll has been replaced with a decision you can use without loading a third-party widget.
Start with the program you need to ship#
List the target platform, required libraries, deployment constraints and the people who will maintain the result. A small script, a browser interface and a device driver place very different demands on a language and its tools.
Then build a small representative piece: read the input you expect, handle one failure and package the result for its target environment. That exercise exposes integration problems that a syntax comparison misses.
Give the trade-offs names#
Consider the debugger, dependency maintenance, runtime requirements and how easily another person can reproduce the build. A language familiar to the team can be a better choice than a fashionable one that leaves the project dependent on a single person.
You do not need to settle on one language for a career. You need a defensible choice for this job. For a more playful contrast, the esoteric-language article looks at languages designed to explore unusual constraints rather than ordinary product requirements.
