Resolution:
If the adapter is  using the local repository file  then  repo.encoding value in
the local repository will be used for this adapter's operations.  

  You can specify the RV encoding for a local repository.  For local repositories,
there are no RV messages, so the repository client doesn't use the
repo.encoding property itself.  However, applications
generally do send messages, so look to this property to decide what kind of
encoding to use.

      If RepositoryManager shows no encoding then the application can
use its own default.   The default if no encoding is specified is application
defined; however, both SDK and IM use Latin-1 as the default encoding.  
SDK only supports 2 values for the encoding property with any data not
in Latin-1 needing to use UTF-8.

  The local language of the machine is completely irrelevant.  The repository
".dat" file itself will always be stored in UTF-8 format regardless of either the locale
or the "repo.encoding" property.  Files are stored on disk as a stream of
bytes.  It is up to the software application to decide how to translate
strings from memory into raw bytes on the disk.  Repository always uses
UTF-8 encoding for this.  The file system in no way imposes an encoding
format.

    For example,  when you use local repository with UTF-8 encoding specified, the
RV encoding and AE message operation for this adapter all use UTF-8 as its
encoding type.

How is the repo.encoding setting in local repository affecting the adapter message in its communication?