My personal expectation regarding database and file tree is if I click
on a file in the file tree or database, I expect it to resume from
whatever file I stopped playing no matter what I do to the disk (other
than of course removing that file).
With the database, if I insert and artist, then later stop, add tracks
from that artist, then resume, to me I'm resuming that artist and expect
anything new to be added.
With clicking on an individual track in a folder in the file tree with
"select" it's less clear to me what should happen to the rest of the
playlist if the folder changes contents. By less clear, I mean neither
"having it change to reflect the new contents" nor "stay the same as it
was before the change" would surprise me much. But with the database, I
feel like a playlist created by inserting queries should be more like a
query's results than a strict list of explicit files.
