Funders and journals increasingly have open data mandates, but adherence is partial. Data "available on request" is often not available and editors asking for data leads to manuscript retractions.