Skip to main content

Known Limitations and Regression Notes

This page records implementation lessons that should be considered during maintenance and testing. Confirm current behavior against the release being tested before treating an item as a permanent product guarantee.

Watcher lifecycle​

Watch mode does not automatically resume after a machine shutdown or reboot. The user must start it again with:

foldr watch <path>

Do not describe the watcher as a persistent operating-system service unless startup and recovery behavior has been implemented and tested on the supported platforms.

Failure isolation​

Processor or worker-thread failures have previously interrupted an organization operation. Changes to concurrency and file processing should test that one failed file does not silently invalidate the rest of a run, and that partial work is reported clearly.

Watcher observability​

Watch-mode diagnostics have previously been difficult to locate. Changes to logging should preserve actionable error details and make the active log location discoverable from the CLI or documentation.

Regression coverage​

When changing the watcher or file-processing pipeline, include tests for:

  • process startup and shutdown;
  • new files arriving while a directory is watched;
  • unavailable or restarted watcher state;
  • partial failures during organization;
  • readable diagnostic output.