Hi, thanks for your suggestion.
I have not played with converting into/out of base64. I may give that a try at a later date, however it seems that would add many extra steps/workload.
Plus, you still have the basic issue of how to find the original file since its saved to a random user ID (for anonymous visitor/user A).
More generally, is duplicating then removing a file really the ‘best practice’ on changing a file’s ownership/user ID? Conceptually, if you knew the user ID of the original file, I guess you would have to use a backend workflow to actually do this, otherwise you’d run into privacy rule issues (one user should not be able to delete the file of another user!).