[res0nat0r]

Is there any way to speed Unison up on large datasets? It runs unusably slow even to scan changes of something like my mp3 directory.