This is a fairly large behavioral change, and requires more careful bookkeeping in the configuration. Meanwhile, most of the actual reporting functionality is already handled by FragmentWatcher. I'm going to close this PR/Issue, and we can revisit if an experiment specifically asks for it.
This is a fairly large behavioral change, and requires more careful bookkeeping in the configuration. Meanwhile, most of the actual reporting functionality is already handled by FragmentWatcher. I'm going to close this PR/Issue, and we can revisit if an experiment specifically asks for it.