Revisiting the cosmic cooling crisis.

Recent measurements of the K-band luminosity function now provide us with strong, reliable constraints on the fraction of baryons which have cooled. Globally, this fraction is only about 5 per cent, and there is no strong evidence that it is significantly higher in clusters. Without an effective subgrid feedback prescription, the cooled gas fraction in any numerical simulation exceeds these observational constraints, and increases with increasing resolution. This compromises any discussion of galaxy and cluster properties based on results of simulations which include cooling but do not implement an effective feedback mechanism