Hi,

I am trying to refine a couple of waters in a structure anisotropically in the latest version of CCI-Apps (Feb-08) . The selection is:

refine.adp.anisotropic = "chain W and (resid 25 or resid 54)"

The refinement works fine if "main.ordered_solvent = False". When I enable "main.ordered_solvent"  the refinement fails with:

"Sorry: Bad (empty or mixed) selection in adp_individual.aniso" during the first update of the solvent in macrocycle 1.

A workaround is easy to implement, but I think this kind of scheme should work. Although it has all kinds of tricky implication for the downstream handling of the selection if the residues/waters in questions are deleted....

Cheers,

        Carsten