I know this goes against everything ADSM is designed around BUT...
     Currently, here at AMOCO, we have two schedules defined for each
     machine.  One is INCREMENTAL and runs daily, the other is a SELECTIVE
     with all file systems listed and subdir=y (ie a full).

     Along the lines of a tape going bad.... if you are recycling your
     tapes you should catch a bad tape before a need to recover from it
     arises.  I have not had a bad tape yet but I would guess that if you
     had a bad tape and marked it destroyed that adsm would know what
     active datasets were on that tape and create a new backup during the
     next incremental cycle.  (I might test this out but am not very
     worried)  I use IBM 3590 cart tapes so as to eliminate tape worries...
     so far so good ;-)

     later
          Dwight