Hi,
I hope this is a good place to make suggestions. This is not technically a bug, but a suggestion that could make things slightly "better".
Suppose I've edited tags in MP3 files stored under C:\z1\z2\z3\z4.
Then I delete folder z3 with Explorer, with mp3tag still running. Obviously, this means folder z4 gets deleted along with it.
If I hit Ctrl-D to change folders, mp3tag does exactly the best thing it can possibly do under those circumstances: z4 is gone, so it looks for the parent, z3. z3 is also gone, so it also tries its parent, z2, which still exists, so the folder selection dialog box then starts at the z2 folder. That's great and it's exactly what I would want it to do given the situation.
Assume the same scenario again, where you have C:\z1\z2\z3\z4 on disk, and mp3tag is showing the content of z4.
If I shut down mp3tag and delete z3, I'd expect the same behavior: The folder no longer exists, so look at the parent folder hierarchy "from the bottom up" until you find something that exists, and select that. But no, when restarting mp3tag, if the last folder selected no longer exists, it doesn't start searching across the parents of the last folder that was used; rather, at startup, it defaults to wherever mp3tag itself is installed (C:\Program Files (x86)\Mp3tag), which is almost certainly always a "bad" folder to start with.
Obviously, you've worked out the logic to search across the folder hierarchy when a given folder no longer exists (use the "closest" parent). It would be nice if mp3tag could re-use that same logic to determine what folder to display at startup, when the last one it used no longer exists, rather than defaulting to the installation folder.
It's not a huge thing, but given you clearly already have the code to do this (and it works well), it ought to be trivial to invoke the same code at startup to decide what folder to use initially.
Just something I come across every once in a while.
Thanks.