Mylio creates duplicates at import from Camera roll

I have an Android phone with Mylio importing the photos from there.
These are imported to a folder on my main Windows PC.

Yesterday I saved 3 photos from Facebook (no metadata).
These were imported to Mylio.
I then moved the files on my phone to another folder (where they are easier grouped).
I then addes a description/caption and run face annotations on them.
These were then also synchronized to my PC.
After this, I saved the metadata to the files in Mylio (to make them available for other applications).
Today I see the photos imported again from my phone and so having duplicates. (One with changed metadata, one without metadata, same filename.)

I guess it is not the only case of Mylio creating duplicates, but it was very obvious.
I don’t know what causes this, the moving to another folder or the saving of metadata. Files were not renamed.

Please do something about this. If a photo was imported by Mylio, than keep track of it and do not import it again.

Although folder structure is not the best way to sort photos, but it is still a way to be able to sort them in a way that they can be found, also without Mylio or any other special app.
It is also “international” meaning, that it can be understood by any app.

When you say:

Do you mean moved them to another folder within the Mylio app or on the Android file system itself?

I moved them within the Mylio app on Android.

Ok, in that case, would you mind sending in a support ticket with your logs. We’ll need to try and reproduce the issue. Thanks.

Sent a support ticket.
It might have happened, before the files got synchronized anywhere.

Tried it again it was reproducible.
Photo “IMG_20201022_170649” imported today at 17:xx, synchornized everywhere, moved on PC, then changed metadata on phone, written back metadata on PC
After some time imported again on the phone and now I have a duplicate.

I guess, that “Photo - Save metadata to file” is mixing up things.

I assume that Mylio is using some kind of checksum or hash of a file to determine, if it is duplicate or not.
It clearly changes, if metadata is written back to a JPG file.

It was an idea coming up in the IMatch forum for duplicate search to use the checksum/hash of the picture part of a file and leave out metadata. This would be an elegant and relative stable solution.
In the days of non destructive editors one rarely changes a file and then moves it.