Why Compile-Fail Tests Can Pass for the Wrong Reasons in Rust

A Rust developer discovered that some compile-fail tests in their codebase were rejecting code for unintended reasons, such as type-inference errors, rather than the privacy boundaries they were designed to verify. The core issue was that a test failing to compile does not confirm the specific property being tested — the reason for rejection matters as much as the rejection itself. In one case, omitting const parameters caused an inference error before the compiler could even reach the private-field access check. Another test used tuple-struct syntax on a named-field type, making it evidence of a wrong API assumption rather than a privacy violation. The developer now applies a four-point checklist — covering the claim, reachability, diagnostic accuracy, and API currentness — before accepting any compile-fail test result.