There are many manipulations available in Ingrid, some more useful than others (see function list). Of particular interest is a function like svd, which does singular value decomposition, i.e. returning a dataset (several variables) from a variable. The averaging function average takes two dependent variables and a list of independent variables, regridding the second to match the independent variables of the first. In both cases the metadata is an important part of the calculation.
But no number of available functions could stand in the way of a user who was determined to type in his url the way the DODS-Grads server had taught him. So we installed a data server just for him. It would be nice if we could standardize.