errnobased

Errnobased is a term used in software engineering to describe error-handling approaches that rely primarily on the errno mechanism to signal and convey error conditions. In this approach, library and system calls typically return a value indicating success or failure, and, on failure, set a global or thread-local error variable (errno) to a identifying error code from the errno.h set. Programs may then inspect the return value and, if needed, examine errno or translate it to a human-readable message via functions such as strerror.

The concept is most closely associated with C and POSIX interfaces, where many functions adhere to a convention of returning -1 or NULL on error and populating errno with a numeric code. This pattern enables compatibility with a wide range of existing libraries and system calls, but can require additional boilerplate to translate codes into meaningful messages or to clear errno before calls.

Advantages of an errnobased approach include low overhead and strong compatibility with legacy C APIs, as well as straightforward interoperation with system interfaces that already use errno. Disadvantages include reliance on a global or thread-local state that can lead to subtle bugs if errno is not managed carefully, limited context for errors beyond the numeric code, and difficulties proving correctness in multi-threaded or asynchronous environments. Critics also argue that errno-based error signaling can hinder composability and modern error strategies that emphasize rich exceptions or explicit result types.

Origin and usage: the term derives from errno, the conventional error indicator in C, with the suffix -based indicating reliance on that mechanism. Errnobased design remains relevant in low-level or performance-sensitive code, particularly in legacy codebases and systems programming.