<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body text="#000000" bgcolor="#FFFFFF">
<div class="moz-cite-prefix">Hello Maik,<br>
<br>
Sorry it took me so long to answer, but I needed all this time to
understand the implications of your question and to find possible
solutions. Not that it was very complicated but lots of other
matters took the precedence.<br>
<br>
First, tell me if I am right with the following guess: <br>
<ol>
<li>dK reads the tags from an photo P, with <span
style="white-space: pre-wrap">"A/A1/X"</span></li>
<li>in the tags database, dK finds two "X": "<span
style="white-space: pre-wrap">A/A1/X" or to "B/B1/X</span>". </li>
<li>dK correctly writes <span style="white-space: pre-wrap">"A/A1/X" in P, but b</span>ecause
dK handles also DC, which does not support hierarchical tags,
dK writes also "x" .<br>
</li>
<li>dK re-reads the tags in the photo, finds the "x" tag and
adds "x" to the tags database.</li>
</ol>
Up to step 3, I don't think anything is wrong. Step 4 creates the
duplicate in the database. <br>
<br>
In a way, dK is correct, the tag "X" is at root level in the DC,
non-hierarchical system. But storing tags in a non-hierarchical
way is only there to allow tags to be read by software which don't
understand hierarchical tags.<br>
<br>
Would disabling any non-hierarchical system be a workaround? I
unchecked Xmp.dc.subject, Iptc.Application2.Keywords and
Exif.Image.XPKeywords and checked all other tags systems, and
indeed I can now have duplicates. <br>
<br>
I can see several ways to improve things, but none would be
complete. So maybe stating that users should either uncheck those
three systems or avoid duplicates is the best. I suggest that the
tags system check list should indicate whether each tag system is
hierarchical or not, this would make configuration more clear, and
few users are aware of those subtleties.<br>
<br>
-- </div>
<pre class="moz-signature" cols="72">Frédéric Da Vitoria
</pre>
On 11/12/2024 21:08, Maik Qualmann wrote:<br>
<blockquote type="cite"
cite="mid:1898083.OvTRdeHEbu@linux.fritz.box">
<pre wrap="" class="moz-quote-pre">Another explanation of the tags: digiKam finds a tag X without a hierarchy in
the uncleaned metadata of an image.
Where should digiKam assign it now, to A/A1/X or to B/B1/X. digiKam can't
decide, so it creates the tag X in the root.
What would be the right solution for you?
Maik
</pre>
</blockquote>
<br>
<pre class="moz-signature" cols="72"></pre>
</body>
</html>