I *think* this is wrong currently, but I'm not sure.
Somebody (Magnus?) close this if this isn't broken.
Currently, the FairDJ tracks database does its own
track discovery by scanning a bunch of directories for
songs.
That should be done by the MServ compatibility backend.
If somebody creates a new backend, that one might not
be file-based at all, which would make FairDJ's
built-in scanning useless.
If some other backend needs to do directory scanning as
well, the scanning code should be placed in its own
class(es) for reusability when that situation occurs
(but not earlier).
Logged In: YES
user_id=131866
Currently done by the very MServ specific TracksDbImpl.
Simply moving it from net.sf.fairdj.tracks to
net.sf.fairdj.mservdb would be a good start. Merging it
with MservDb and simply calling the resulting class
"MservDb" would be even better.
I'm about to lose my Internet Access for a while now so I'm
whining here rather than coding.