23
@sailjapan:
That's right!
As the original (default) modules/indexscan/language/nederlands/ is the UTF-8 version, the user can 'fallback' to the ISO (ansi) encoding by copying the contents of the iso directory.
Traditionally, all Dutch sites used a latin1 database and ISO-8859-1(5) encoding.
Strictly, the utf directory is not necessary, but it is supplied for convenience, eg. for easy resetting a previous iso copy operation.
Or for future upgrading, eg. when the database can reliable converted from latin (=ISO site) to utf.
As it is more or less imperatif (to have the least problems), that the encoding of your site goes with the encoding of your database, I don't think it is a good idea to offer the user the choice between eg japanese-EUC or japanese-UTF
IMHO it is best to always start from the database encoding and have the sites' encoding along.
Existing sites which have EUC encoding in their database, should have also the same encoding for the site, with parameters for charset in accordance in mainfile.php and global.php
New sites are advised to have full utf (database and files). One problem here, is that most older module translations are done with (and contain thus only) the ISO (ansi) files. The user has to convert them, with NotePad++ to UTF-8 without BOM.