The time variation of physical constants has been much discussed in the literature, motivated by claims of fine structure constant variations together with several theoretical ideas. Although it is well understood (by most, but not all!) cosmologists that one must consider only dimensionless constants, most discussions of the strength of gravity involve "G", which is of course dimensional. I discuss some applications of variations of "G" on cosmological observables, stressing the need to stay dimensionless. "Constants" might also vary in space.