The author recounts a frustrating experience with a C++ codebase that handled billions in online payments. A bug report revealed an impossible scenario where both "error" and "succeeded" fields in a response were set to true, despite expectations that only one could be true at a time. The code in question involved a struct named `Response` with boolean fields for error status and success, but the underlying issue stemmed from how the struct was initialized. Through investigation, the author discovered that the default initialization rules in C++ differ significantly from those in C. When a non-POD (Plain Old Data) struct is declared without an initializer, its fields may remain uninitialized, leading to undefined behavior. In this case, the `Response` struct had a default constructor generated by the compiler, but the boolean fields were not explicitly initialized, leaving them with indeterminate values. The author emphasizes the importance of proper initialization practices, suggesting that developers should always zero-initialize their variables to avoid such pitfalls. The article provides valuable insights into C++ initialization rules, distinguishing between different types of structs and their behaviors. It outlines how to properly initialize struct fields by defining default values or implementing a constructor. The author shares their eventual fix—changing the declaration to `Response response{}`—which ensures that the fields are zero-initialized. This experience serves as a cautionary tale about the nuances of C++ and the potential dangers of undefined behavior in software development.