I'm using a localized form (specifically, Serbian) of
U+0431 "CYRILLIC SMALL LETTER BE". I think Latex does not kern it the right way, behaving as if it was adjusting space to the standard version rather than to the localized one.
Gonna make you understand this lofty issue through a very simple visual sample: in the attached .pdf you'll find just two lines with each two letters! OK?

Now, the first one, the capital (
U+0423 "CYRILLIC CAPITAL LETTER U") doesn't concern us much, I put it in here just because it displays what a blatant example of bad kerning.
Are you following me

All right

So let's concentrate on the second letter: you'll have noticed that while the first letter appears identical on both lines, the second looks different! But it still bears the same Unicode code & name
U+0431 "CYRILLIC SMALL LETTER BE", that's because
the first is the localized Serbian glyph, the second is the standard (Russian) glyph.
The bottom line of all this complicated talk, it's that since kerning is all right in the second line (with standard-Russian glyph), while patently
wrong in the first (with the localized-Serbian glyph),
I got the impression that Latex persists on kerning "having in mind" the standard-russian form rather than the localized-serbian one.
Hope I didn't scare you guys with such a long subject.

And thank you all in advance!
