Hi Vladimir,

You write:

 > there already exists an example of such
 > character: \CYRpalochka, which is invariant to
 > upper/lowercase conversions.

Interesting.  Then maybe \CYRapos is the right
solution? However, that will place a stubborn road-
block in the way of the typographer who wants
distinct lowercase and uppercase Cyrillic
apostrophes. Maybe there are none today (??). But
what about the furure?  The \CYRAPOS \cyrapos
solution seems more standard and more powerful
(less limiting).

 > is the apos character (letter modifier) used in
 > other (non-cyrillic) languages? i think that yes,
 > since it is defined in non-cyrillic unicode area.

I hope others will chime in here. Most of the uses
in English, French and German that I know may well
be non-letter and covered by the vague:

0027;quotesingle;APOSTROPHE

I would just note that Unicode is *not* a model of
clarity here since in English quotesingle and
APOSTROPHE are disjoint meanings with an accidental
coincidence in the ASCII and Unicode norms:
Octet "27=39 has several meanings.

TeX can do better here in the Cyrillic world where
TeX is young and flexible.

You suggest at a couple of points that the concept
of a 'trans-script' letter is a good idea and in
particular that "apos" should be regarded as one.

To support this thesis you rightly note that
many many *non*-letters are common to the Cyrillic
and Latin scripts.

As for 'trans-script' *letters*, one might reasonably
claim that the (hard) hyphen is one.  (I agree
that it is 'trans-script' but I do *not* agree that
it is a *letter* even though it is found in
dictionary listings.

The notion of a trans-script *letter* seems
vvvvvery doubtful to me.  One could gaily claim
that the letters A,E,K,M,O,T,X are obviously common
to Cyrillic and Latin and that in consequence CYRA,
CYRE, ... should somehow be eliminated (:=/

Is it not dangerous to allow even one letter in the
Cyrillic script fall out of the 100% control of Cyrillic
users??

           Cheers

                Laurent S.