It allows you to check or modify a data set referred to in your job without leaving the job
itself, going in split mode etc. It can also be used to make sure a specific data set is there
instead of waiting for an annoying jcl error. It supports PDS, sequential and VSAM organizations.