Is there a way to distinguish (in duplicate search) between files that have the sane name but different kind?
e.g. one file is a .doc file. the other one is the same file but.rtf (actually is the same file, imported…)
AFAIK DTP classifies both as document
Thanks
The icon should be different. In addition, you could enable the Kind column if you’re using List view.
so not by “creator” or filename extension…
There is a Creator column but that doesn’t necessarily mean it would have data. That’s also not reporting the creator of all file types. It’s for PDFs.
the kind column in list view distinguishes between RTF and word documents… why not in Smart group filter definition? 
Because the menu of Kinds would be enormous.
However, in searches and smart groups/smart rules, there is an Extension criterion.
just to be clear: I don’t think that two “identical” files (a file Test.doc imported in DTP and a file Test.rtf drawn from the first one -the .doc is unchangeable in DTP…) are duplicates, this is misleading and dangerous (it is possible to trash important files).
I rather prefer that the duplicates are real duplicates, ie with the same extension…
See preferences General > General.
From help:
Stricter recognition of duplicates: Check to have DEVONthink mark files as duplicates consider the file type and size, as well as the content, when determining duplicate files.
I apologise but let’s start from scratch please…
I have a file Test.doc imported in DTP and a file Test.rtf drawn from the first one -the .doc is unchangeable in DTP…
if i don’t modify the latter file, these are qualified as duplicates, ?
If I do my work on the .rtf file (let’s say annotate it) they should not be still duplicate, correct?
Thanks
I have a file Test.doc imported in DTP and a file Test.rtf drawn from the first one -the .doc is unchangeable in DTP…
if i don’t modify the latter file, these are qualified as duplicates, ?
Correct.
If I do my work on the .rtf file (let’s say annotate it) they should not be still duplicate, correct?
Adding an annotation file wouldn’t change the duplicate state.
Modifying the content of either file can change the duplicate state.
And as @pete31 pointed out, using the stricter recognition will change the duplicate state in this case since the file type would be different.
Why annotate a file doesn’t change its status?
When the status modifies?
Because an annotation file is a separate file.
The content must change for the duplicate state to change.