When instantiating a forkmanager, and then using it in another process,
the child/parent communication is broken as parent_pid is not updated
(still referencing the initial PID, not the real parent). I don't know
if this is crucial but it could introduce security issues: one could
have the time to forge a fake Storage file using the parent PID. This
could get mitigated really easily and I don't think it would have
side-effects to update parent_pid in start().
Good catch! I don't think (fortunately) that it's a security issue:
somebody could also fake the storage file with the current parent pid or
the original parent id (or even carpet-bomb the directory with files
with "$i-$j" for all $i and $j in a vast range). But the fix is indeed
easy, and will be pushed to CPAN real soon.
Good catch! I don't think (fortunately) that it's a security issue: somebody could also fake the storage file with the current parent pid or the original parent id (or even carpet-bomb the directory with files with "$i-$j" for all $i and $j in a vast range). But the fix is indeed easy, and will be pushed to CPAN real soon.