Runtime Checks Disabled in Nim Release Builds Cause Silent Failures After Local Debug Testing

When developing a Nim application, developers typically compile with the default -d:debug flag, which inserts runtime checks for array bounds, nil pointer dereferences, and integer overflow. These checks help catch mistakes during local testing. However, the production build is usually created with -d:release (or -d:danger), which omits those checks to improve performance. Because the when defined(release) guard excludes debug‑only code, the same source can exhibit different behavior: errors that abort in a debug build may proceed silently or cause undefined behavior in a release build, only becoming apparent after deployment.

The goal is to understand whether relying on debug‑only checks is acceptable for correctness, or if additional safeguards are needed to guarantee identical behavior across build modes without sacrificing the performance benefits of release compilation.

• What strategies exist to detect the class of errors that are only caught by debug‑time runtime checks in a release build?
• Should projects enable specific runtime checks in release mode via compiler flags, and what performance impact can be expected?
• How can the when defined(release) conditional be used safely to avoid divergent code paths while still allowing debug‑only diagnostics?