>Is there a way to develop app which accepts input in many languages (serbian latin, serbian cyrilic, german, french...) at same time? Example for serbian latin and serbian cyrilic: шђчћ šđčć + german special characters etc. Program is for library.
>What code page for tables, forms? Font? Settings from control panel-regional and language options-advanced-language for nonunicode programs?
>User should enter text in different languages simply by changing keyboard input language from serbian latin to serbian cyrilic etc. Any suggestions?

Mark your character fields as nocptrans and then on display use the appropriate .FontCharSet property for each field/textbox/editbox. You won't be able to mix them in the same field, though - a textbox can have only one FontCharSet at a time, can't switch between them. For nocptrans fields the codepage of the table is irrelevant.

For indexing, however, if you're doing dbfs, tough. You'd have to pick either 1251 for your tables (if doing cyrillic) or 1250, and then use russian or czech collating sequence, which aren't perfect but are almost there. If you're using SQL or other server-side database, you can tweak your collating sequences at will - have one field in one, another in the other (and then have the SQL server claim that two strings can't be compared because they aren't in the same sequence, beware of that).

Still, you should be fine doing all in cyrillic, knowing that those who insist that their company names be in latinic do not ever use šđčćž in their names. Except Ćircom, formerly Ćira Comerc (or Comerce - I never know which wrong spelling they prefer). Within any character set you still get the 26 english characters, and that's pretty much all that these guys are using.