Version:            (using KDE Devel)
Installed from:    Compiled sources

KDE applications are not doing "the right thing" regarding Unicode strings.

Unicode defines canonically equivalent sequences of characters.
For example these are equivalent:
ẹ́ <U+0065 LATIN SMALL LETTER E + U+0323 COMBINING DOT BELOW + U+0301
COMBINING ACUTE ACCENT>
ẹ́ <U+0065 LATIN SMALL LETTER E + U+0301 COMBINING ACUTE ACCENT + U+0323
COMBINING DOT BELOW>
ẹ́ <U+1EB9 LATIN SMALL LETTER E WITH DOT BELOW + U+0301 COMBINING ACUTE
ACCENT>

Applications should normalize strings before comparing them. This needs to be done in sorting, searching, opening/creating/saving file.

For example if a file has one string sequence and a query is a canonically equivalent string, they should match.
Or two canonically equivalent strings should be sorted as the same string.
Creating files or directories, or other meaningful entity, should behave as if canonically equivalent strings were the same, i.e. warn the user of a conflict of name and offer to overwrite. Also, one path should access the equivalent existing one (unless both exists).

QString::normalized(QString::NormalizationForm_D) or (QString::NormalizationForm_C) should be used before comparing strings.

It might be wise to have a policy normalizing all created filenames to NFC, for better compatibility with legacy systems.

For sorting QString::NormalizationForm_KC should probably be used.

Please read http://www.w3.org/TR/charmod-norm/#sec-NormalizationMotivation for motivation and http://www.unicode.org/reports/tr15/ for more info.