Track scanning can take a long time. When starting
track scanning, we should log a message saying that it
can take up to half an hour on a big archive. We
should log a progress message about once a minute
saying how many tracks we've found so far, and that
we're still scanning.
When done, we should list how many tracks we found, in
how many albums, and for how many tracks we found any
metadata (or whatever is doable). The user should get
some kind of numbers at least that could be used for
judging whether or not things seem to have gone well.
After having read the user database, we should say from
where we read it (if applicable) and how many users we
found in it (always).
Logged In: YES
user_id=20061
Even better, we should not (like Mserv) automatically rescan
all tracks at startup. That's just silly, especially on a
huge archive. Instead, the reset command should be used to
forcefully initiate a rescan.
However, if *no* track data is found (the first startup
ever), I think a rescan should take place.