Some ideas for a databank of macroseismic data