The SVN version of Amarok from today (19.12.2005) does not recognize correctly the tags that are with Cyrillic characters (encoded in utf-8 and ID3v2 tags).
I enclose a picture.
The problem is that when doing \"Rescan collection\" it catches these tags in a certain wrong way (pic.1). When you do \"Rescan\" again (without doing anything else with the tags,files or anything) the tags are entirely different and wrong again (pic.2). But when you play a song with a broken tag the OSD is showing the right info (pic.3).
All the tags in the collection are UNICODE (and they are ID3v2 tags) , engine is xine (but the same problem exists with all the other engines).
Another problem is that in the collection window, the sorting of names in Cyrillic is broken.
One more thing is that when you search (in the search bar of the collection) with Cyrillic characters it is Case Sensitive Search (and if the characters are Latin it is Case Insensitive).

Post edited by: PhobosK, at: 2005/12/19 16:54

PhobosK wrote:

Another problem is that in the collection window, the sorting of names in Cyrillic is broken.
One more thing is that when you search (in the search bar of the collection) with Cyrillic characters it is Case Sensitive Search (and if the characters are Latin it is Case Insensitive).

This is a known bug in SQLite.

Here are the missing pictures from the above post (in pdf format) [file name=Amarok.pdf size=32974]http://amarok.kde.org/components/com_simpleboard/uploaded/files/Amarok.pdf[/file]

Forgot to say that I use (and compiled Amarok) only with SQLite.

Post edited by: PhobosK, at: 2005/12/19 17:09