Backwards C Error-Check Becomes an Unwritten Team Convention

In a mostly C project, a veteran developer inverted the usual Boolean convention: functions that return TRUE on success are stored in a variable named error, and the code branches on error == FALSE to handle failures. The pattern spread through the codebase by sheer inertia, pulling new hires into the same confusing style.

A CodeSOD article on The Daily WTF, dated 2026-07-15 and written by Remy Porter, discusses a pattern in a largely C project that a developer named Stevie submitted.

In that codebase, functions return a BOOL to indicate success or failure: TRUE means the call succeeded, and FALSE means it failed. The normal way to detect failure is `BOOL success = someFunc(); if (!success) { ... }`.

One long-time developer, however, used the opposite form: `BOOL error = someFunc();` followed by `if (error == FALSE) { ... }`. Inside that block the code handles the failure, often calling `errorCode = GetLastError();`. The variable name suggests that a true value signals an error, but it actually stores the success status.

Because this developer has enough seniority and the existing code has mass, the pattern has become the de facto convention in much of the project, even though it was never written down. Newcomers start with the clearer style, but the volume of existing code exerts stronger pull, gradually forcing them into the same confusing idiom. The author calls it less a genuine WTF than a persistent annoyance.