interaction design – Are non-existing files in a list of recently used files to be deleted without prompting the user?

Say, you have this common thing of "recently used files" or "frequently used files" in your application.

It can easily happen that the user moves or deletes a file that he has opened recently / frequently. So what happens now if the user clicks on this file in your recently / frequently used list …?

  1. They should obviously have returns / reviews as the file is no longer there.
  2. Now here is the question: Should the application also automatically delete the file it does not find in the list?


Some ideas that I have about it:

  • (+) Pro: The entry is absolutely useless because a user can never open such a file anymore. It takes precious space where it stays.
  • (-) Contra: In some cases, a file can only be temporarily deleted. While on Android, this can for example This is never the case, it can happen with network shares (think FTP or samba, for example) on desktop systems. In such a case, the user can simply be disconnected and a file can be opened later.
    On the other hand, an app can put these items on the blacklist. (looking at the file path)
  • (+) Prompt: A user may also be prompted to indicate whether the file should be deleted. However, if the file is really gone, this may not be useful because it's just another click and perhaps obvious?
    • (-) Against prompt: Confirmation prompts are generally not considered the best UX model, so a "cancel" feature would make sense here. But how can this be accomplished in this case without completely confusing the user? (note that we must also 1. warn the user that the file could not be found …)
  • (-) Only manual removal: we could not automate this at all, and let the user just click on an "x" or next to each entry, so that they can delete it manually. However, in my humble opinion, this is pretty lame in that, although the feature may be independently useful / necessary anyway, in case the file is gone, we can be totally intelligent as well. application and at least suggest to the user to delete it automatically. Why should users first interpret our error message, thought and then react appropriately by manually deleting the (correct) entry? (I guess it's "difficult" and I doubt that many users do it, which could again "accumulate" "broken entries" in our list.)