[tlug] Samba migration - code page conversion

["Patrick Niessen" <tlug.niessen@example.com>]

I need to migrate loads of Users Files from one server to another and
face the following issue:

Original configuragtion sets the client codepage to "character set =
ISO8859-15"  which is was a good setting in the past for NT4 and
Windows2000 Clients in Japanese native mode.

Now that Linux and samba have also Unicode support it would be great
to move the files to unicode so that the file names can also be read
under Linux and Windows user with English or German systems don't see
the garbage names.

Is there anyway this can be done automatically?  My idea was to map
new and old servers
on a Japanese client and then just use windows copy command. 
Unfortunately this would remove all the file data, like create date,
last modified and permissions ...

Any suggestions?

-- Patrick Niessen