It is sincere. If you are working with structured documents, then accessing only an arbitrary line of the file in a performant way (especially if the documents are large) is a useful task. I see it also has facilities for handling CSVs, where its unlikely the user wants individual lines but may in fact want "columns."