Fuan

I believe this is working as intended. If you set the directory as "reference", then you are telling the program that any file found in it should be a reference....

That's what "re-prioritize" is for. It's like considering files from a certain directory as references by default, but you can still override that by marking other files as reference in...

This is why I wrote: > you should set all directories as "Normal" and use the "Re-prioritize Results" feature.

Have you tried with the "filename" scan type?

There are programs already designed to do this, most notably [diffimg](https://github.com/sandsmark/diffimg). Sure, it's a possibility, but this is also borderline feature-creep. As far as I'm concerned, there are much bigger...

I'm still not convinced that permanently swapping is a good idea as I don't see the use case.

I was tempted to work on such feature. For now I suggest using [diffimg](https://github.com/sandsmark/diffimg) to achieve this.

Not sure if I understand what you're trying to suggest. Also I'm curious to know where did you take the 15x15 grid? I had a quick look at the algorithm...

This has been a concerned of mine for a while too. At first, I thought we would need to move to a [multi-processing](https://docs.python.org/3/library/multiprocessing.html) model instead of the current multi-threading one...

Can you give the approximate number of files you added to be scanned?