[Digikam-devel] [Bug 298547] digikam crashes, if pics of collections are changed externally

[Axel Krebs]

https://bugs.kde.org/show_bug.cgi?id=298547

Axel Krebs <axel.krebs at t-online.de> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |VERIFIED
         Resolution|DUPLICATE                   |REMIND

--- Comment #3 from Axel Krebs <axel.krebs at t-online.de> ---
I am _not_ quite sure, if this is a duplicate of bug #280142.

Further more, if so, it would be still open (=unsolved).

The relavance of the reportet bug is tremendeous for me:

digiKam version 2.5.0
Bilder: 
BMP: 1
GIF: 100
JP2: 16
JPG: 101070
PGF: 2
PNG: 1245
RAW-CR2: 632
RAW-CRW: 15760
RAW-DNG: 6
RAW-NEF: 55508
TIFF: 1442
XCF: 1
Gesamt: 175783
: 
Videos: 
AVI: 73
MOV: 9
MPEG: 2
Gesamt: 84
: 
Gesamtzahl der Einträge: 175867
Alben: 2580
Stichwörter: 44
Datenbanktreiber: QSQLITE

About 3000 faces - all have gone after crash: loss of at least 2 days
About 2582 subfolder - all folder previews have gone after crash: weeks of
work?
Writing all metadata: takes ~ 8 hours
Writing all thumbnails: takes ~ 12 hours
Writing all fingerprints: takes ~ 18 hours

All gone, once again...

These experiences make me doubt about the metadata storing concept heavily. 

Instead of keeping _several_ databases (db, thumbnail-db, faces?) with
_uncertain_ (as my experiences indicate) connecting to _single_ pictures, I
suggest to use one unique number per pic (I support here another persons
suggestion) to connect to database.

As different formats seem to be supported to a different extend by digiKam,
this could resolve this problematic situation to a far extent. Writing metadata
into Canon CRW takes about 5 times more time than to write into jpg!!

By the way: I always search for dublettes intensively when reoporting a bug.
But I don´t see similar ones. Is there a bug in KDE bug-reporting system?

Please let me know about progresses!

Axel

-- 
You are receiving this mail because:
You are the assignee for the bug.