From the COD entry 7104804
(http://www.crystallography.net/cif/7/7104804.cif):

The methyl group of C30 (and several others) are diordered, but since
the molecule bearing them lies on the mirror plane, and the methyl
group apparently has *all* its hydrogens off the symmetry plain, the
methyl group must be disordered; however, there seems to be no way to
describe this situation in a CIF -- if we assign disorder groups to
different methyl hydrogens, they get symmetry-mapped into wrong
disorder groups. The problem is that the mirror plane is *not*
applicable to the methyl, but lowering symmetry just because of three
hydrogens is unjustified.
