From: Ivica Bogosavljevic (ivica84@ptt.yu)
Date: Sun Oct 20 2002 - 09:51:03 EDT
  To view this message right, you must have support for ISO-8859-2 and 
ISO-8859-5.
When one translates a program to Serbian (Cyrillic), he doesen't 
translate all the text. Some labels, for instance AbiWord, stay in 
Latin. But when you are translating to Latin, the same labels are also 
in Latin.
Now when I would write a script to translate from Latin to Cyrillic, I 
would get something like this:
Informacije o AbiWord-u                        Информације о АбиВорд-у
The part of string which contains word AbiWord shouldn't soppoused to be 
translated, but it is.
The correct translation would look like this:
Informacije o AbiWord-u                        Информације о AbiWord-у
But, when script translates from Cyrillic to Latin, you get something 
like this:
Информације о AbiWord-у                Informacije o AbiWord-u
You get a good translation to serbian (latin).
The script should do this:
change the language coding from ISO-8859-5 to ISO-8859-2 and change some 
letters (the complete list of letters is included as an attachment).
This archive was generated by hypermail 2.1.4 : Sun Oct 20 2002 - 09:58:36 EDT