[XeTeX] New UTF-8 hyphenation patterns (& Serbian)

[Mojca Miklavec]

Hello,

Before I start with my question - everyone is invited to test the new
hyphenation patterns. They are going to replace the old patterns in
TeX Live 2008 (almost completely), so it would be really desired to
test them a bit more before release. They are also already part of
MikTeX, but we haven't heard any feedback from MikTeX users so far.
    http://tug.ctan.org/cgi-bin/ctanPackageInformation.py?id=hyph-utf8


However, nobody of us is entitled to solve this without others' opinions.

Serbian can be written either with Latin or Cyrillic. "serbian" in
language.dat was pointing to Cyrillic patterns until now, but babel
uses Latin labels for the language. I'm not sure if loading Cyrillic
patters for Latin texts is a good idea. Can anyone explain some
history behind that?

Do we need Cyrillic patterns at all? May we map Latin patterns to
Serbian instead of Cyrillic ones?

One argument against having Serbian Cyrillic patterns included by
default in language.dat is that there is no babel support for it, and
if we simply add an entry into language.dat, no Serbian user will ever
know it existed. And if someone unsatisfied with the name comes after
several years, he'll have to keep that name for the sake of "backward
compatibility" for zero users.
(To make it clear again - patterns for Serbian will be shipped in any
case, but the question is - should they also be enabled?)

It would be nice if some Serbian user would step forward and request
them, offerisg some suggestinos & testing as well.

Some more facts:
- srhyphc was done by transliterating shhyph + some minor
modifications, and the author is not accesssible.
- shhyph had the wrong licence until recently, but the author has
fixed it and also provided transliterated patterns for Cyrillic script
himself.
- now the two cyrillic files with should-be-the-same-content have tiny
differences, and we have no idea which one should be loaded (if loaded
at all).

Thanks,
    Mojca