FileBot

Fpcalc results should be cached preventing regeneration if the file was not changed.

That's a essential requirement for using it in a media center where you use symlinks to generate the the browsable directory structure.

Without this feature the processing would take hours and hours.

filebot shouldn't be run on files you have already processed in the first place

You're wasting the free AcoustID webservices unnecessarily!!!

EDIT: FileBot imposes a 3 requests / second flood limit on AcoustID usage. So this is the bottle neck, not the fpcalc call. And the requests are already cached 2 days / 20k items max. I guess I can increase that.

EDIT2: With FileBot r2266+ caches for AcoustID webservices are now set to 2 months / 200k items