I wanted to share my contributions to open-source, and especially the work I recently did for the TypeScript-ESLint project. For this week, I challenged myself to contribute to a much bigger and more complex project than I had worked with before. I wanted to push past basic fixes and actually make a change that improves developer experience for a large community. After exploring the repository and reading through open issues, I found Issue #11758, which described a bug in the no-base-to-string rule. The rule was mistakenly warning when calling .toString() on a class that extends Error and uses generics, even though that should be allowed.
Before writing any changes, I needed to understand how the rule checked TypeScript types. I spent a lot of time reading through the code and learning how the TypeScript type checker works with inheritance. Eventually, I discovered the cause of the bug: the rule only checked the type’s own name to decide whether it should be ignored, but it never looked at its parent type. Because of that, a type like a generic subclass of Error was treated as unsafe, even though it should be allowed. To fix the issue, I updated the rule so it now looks not only at the current type but also at all of its base types, even when generics are involved. This means the rule can now correctly recognize inherited behavior.
In the end, the fix turned out to be much smaller than I first expected. After updating the rule logic, I then needed to prove that it worked. I added a new valid test case that represents the situation from the issue, where a generic class extends another class which ultimately extends Error. This test confirms that the rule no longer reports .toString() in that scenario. Running the full test suite showed that everything passed successfully, which gave me confidence that my fix solved the problem without causing any new issues elsewhere in the project.
This contribution was a big learning step for me. Compared to earlier work, I gained much deeper experience reading unfamiliar code and learning how to efficiently navigate through large codebases. It also showed me that I am capable of contributing to major developer tools used by many people every day. I hope to continue doing more contributions like this going forward.
Top comments (0)