On systems having over a thousand tickets the current spool directory can become slow on some filesystems. We should split the code for creation and accessing a ticket ID so that:
ticket creation splits the spool directory into subdirectories starting with the ticket prefix
ticket access checks for those subdirectories
essentially mirroring the current git object hierarchy.
On systems having over a thousand tickets the current spool directory can become slow on some filesystems. We should split the code for creation and accessing a ticket ID so that:
essentially mirroring the current git object hierarchy.