XYplorer Beta Club

I'm finding that the tools to select duplicates are kind of lacking...

As of now there's:
Select all first
Select all non-first

Which are near useless because XY doesn't sort the groups properly.

So I'll have

Group1:
Folder1/copy1
Folder2/copy1

Group2:
Folder2/copy2
Folder1/copy2

Which makes both options useless since neither one allows me to bulk select all the copies in folder 2...
So first off XY should sort groups by folders so that these two already built in selection options actually work in more cases.


Second other particularly handy options would be:
Select all duplicates with the shortest name
Select all duplicates with the longest names

Select all duplicates in the deepest folder
Select all duplicates in the shallowest folder

Maybe not the best names, but hopefully they get the meaning across.

I think those 5 options would cover all the most common way's people wind up with duplicates and offer quick fixes.

Thanks for the ideas. I hardly ever use the dupes finder myself, so I would like to invite other dupes finder users to this thread for some discussion of needs and desires.

After a search for duplicates it looks like this in Total Commander.
The Select by folder allows what the poster requested.

Give me a set of sample files where the order of the dupes is as chaotic as you say. I don't see it here.

Here you go:

1) Try to sort the list by the Dupes column. (By which column is it sorted in the screenshot?)

2) Show a screenshot of your Dupes tab. (What kind of dupes are we talking about?)

Here's the tab, I uploaded a video of sorting the dups col as a zip since it wouldn't let me attach an .mp4
The column sorts but never so that all the same folders end up in the same place in all groups.

In the 1st screen shot it's sorted by whatever the default is, so I'm guessing dup groups.

Alright, I could reproduce and fix it. Next beta.

BTW, tick this, it makes understanding and debugging easier:
Configuration | General | Sort and Rename | Sort | Show implicit secondary sort order arrow

BTW:

I don't know when this started, but surprisingly this is currently reported as my thumbnail provider for png files: Apparently this is the Google Drive File Stream thumbnail handler. Never told or allowed Google to take over this job. And it has serious side effects. At least on my system the dupe finder by image was broken for PNG files! Fixed in next beta, but WTF!

Responding to the invitation to other dupes finder users for some discussion of needs and desires:

Here is the brief summary of what I wrote below:

• Show duplicate files grouped together, but don't filter out the non-duplicates.

• Retain the shading of grouped files even when the files are sorted by date, rather than by dupes number.

Here is the long version:

For about 10 years, before I became employed and learned about version control systems, I saved snapshots of my development folders at different points in time. Those were saved on floppy disks for safe storage, frequently multiple times. All of those floppy disks have now been copied to my hard disk. The result is that I have about 600 files that are composed of 390 dupes in 138 groups, and about 200 more that are not duplicated. I would now like to impose some sort of order on that mess, along the lines of putting those files into a version control system. The problem I have with the dupes filter is that I can either see all the files that are duplicated, or all the files that are not duplicated. I can't see all the files with duplicates marked.

What I need is a way to show all the files that match the search criteria, and mark the files that are duplicates, but don't filter out the non-duplicates. This would look like a regular dupes filtered list, except that files that have no duplicates are each in a "group" containing only one file.

I also need to sort these files by date. The problem with that is that the shading that makes grouping stand out disappears. Most, if not all, of my duplicate files have matching dates, so when the files are sorted by date, the dupes group numbers are still clumped together, but the shading is gone. It would be really nice if the files could be sorted by date, but retain the shading that highlights the grouping.

I realize that that last sentence can create an ambiguity: If some duplicates have non-matching dates, where should those files be located in the date-sorted list? Should the list be strictly ordered by date, causing groups with non-identical dates to be broken into fragments? (I'll call this "sorting order A".) Or should groups remain intact, causing the files to no longer be strictly ordered by date? (I'll call this "sorting order B".) Both could be useful. The issue, then, is how does the user select one of the two choices?

My first answer is that since my files don't (or usually never) have non-matching dates in groups of duplicates, if XYPlorer had my first two wishes--marking duplicates without filtering, and retaining group shading when sorting by date--I would be thrilled, and perfectly content with the easy solution: strict date ordering, with possible fragmenting of groups (sorting order A).

If both sorting orders were to be supported, the easiest implementation is probably to have a checkbox in the options: When sorting duplicates by date, keep duplicates grouped, regardless of date. (But if there is a way that's easier for the user than going into the options dialog, that would be better.) The sorting algorithm that would achieve this (sorting order B) is pretty easy:

• Start by ordering the files by date.

• Traverse the files in order of date. For each file that has duplicates, move those duplicates up to be grouped with this file.

• Once there are no more duplicates for a given file, sort that group by date.

• Skip over the files that were moved, and continue with the next file in the list.

That would result in all of the first files in each group being strictly ordered by date, and all of the subsequent files in each group sorted by date within that group.

These additions would make the dupes system much more useful to me.

Try to make your point with fewer words (and maybe more screenshots), and explain why it would be a benefit to many other users.