Learning preserved

Archive

Ideas that were intentionally stopped. Not failures — each one keeps the original problem, what was tested, and why it ended.