declare-barrierlisted

Stop. Do not write "this cannot be done" until you have completed every step below. ## The core principle A barrier you have not systematically challenged is a lazy excuse, not a scientific conclusion. The categories below represent recurring patterns of false barriers — they appear in nearly every complex reproduction task. Check every one before accepting any block. --- ## Recurring patterns of false barriers ### Pattern 1: "The algorithm is not accessible" **Almost always false.** Ask: - Did you read the Methods section of the paper, word for word? - Did you read the cited reference for that algorithm? - Did you search GitHub, PyPI, CRAN for an existing implementation? - Did you search Google Scholar for "algorithm name + implementation" or "algorithm name + Python"? If the algorithm is described in a published paper, it is accessible. Complexity is not inaccessibility. ### Pattern 2: "The data is not available" **Often false.** Check in order: 1. The paper's "Data availability" section 2. The paper's supplementary materials 3. Public data archives — general: Zenodo, Figshare, GitHub; field-specific examples: PANGAEA (oceanography), NOAA (atmospheric), CMIP5/6 (climate model output) 4. Your local `data/` directory — the data may already be there 5. A representative subset — full dataset unavailability does not prevent partial verification ### Pattern 3: "This is too slow" **Almost always solvable.** Before accepting: - Profile first: where exactly is the bottleneck