11-15-2018, 05:26 PM
The traditional answer:
This is a known issue. Because rules are applied on GPU, this can't be fixed for fast hashes without significantly slowing down cracking - and doesn't need to be fixed for slow hashes, because you can filter out long candidates before they're passed to the final instance hashcat (so you'd apply rules in one instance of hashcat with --stdout, and then pipe them to the cracking instance of hashcat).
That being said, this may change someday with the new -S / --slow-candidates option.
This is a known issue. Because rules are applied on GPU, this can't be fixed for fast hashes without significantly slowing down cracking - and doesn't need to be fixed for slow hashes, because you can filter out long candidates before they're passed to the final instance hashcat (so you'd apply rules in one instance of hashcat with --stdout, and then pipe them to the cracking instance of hashcat).
That being said, this may change someday with the new -S / --slow-candidates option.
~