Bug 416759

Summary: in duplicate tab, it should be possible to have reference image always at first/last
Product: [Applications] digikam Reporter: Daniel <daniel-other+kdebug>
Component: Searches-SimilarityAssignee: Digikam Developers <digikam-bugs-null>
Status: RESOLVED WORKSFORME    
Severity: normal CC: metzpinguin
Priority: NOR    
Version First Reported In: unspecified   
Target Milestone: ---   
Platform: Other   
OS: Linux   
Latest Commit: Version Fixed/Implemented In: 7.0.0
Sentry Crash Report:

Description Daniel 2020-01-25 17:01:09 UTC
SUMMARY
Currently, the images aren't sorted by anything it seems (at least changing in View -> Sort Items didn't change anything). You can also watch that sometimes the reference image is the first, sometimes the last entry. 

It would be useful to have the reference image always as the last entry, because then you can quickly delete duplicates by pressing delete on the keyboard until only the reference image is left.
Comment 1 Maik Qualmann 2020-01-25 17:24:10 UTC
Sorting the items works (tested) just like in the album view. I find it bad to push the determined reference image to the end.

Maik
Comment 2 Daniel 2020-01-25 17:33:57 UTC
(In reply to Maik Qualmann from comment #1)
> Sorting the items works (tested) just like in the album view. 

Indeed, they are sorted (all sorting options didn't change the order for me, but they are correctly sorted), but that's not what I indicated here.

> I find it bad
> to push the determined reference image to the end.


I have had about 8000 duplicates (in a 40.000 image collection), and it was really annoying when the reference image jumped to first image (as it was at the end most, but not all, of the times.)

I think a "sort by reference image and name" would be reasonable.
Comment 3 Maik Qualmann 2020-01-25 17:37:19 UTC
Sorting the images by similarity already exists. I think that's what you want.

Maik
Comment 4 Daniel 2020-01-25 19:10:23 UTC
(In reply to Maik Qualmann from comment #3)
> Sorting the images by similarity already exists. I think that's what you
> want.

I think too (but now I have no duplicates left to test :) )