- BOM is now checked and correctly handled
- auto-detection of UTF-8 string if no BOM available, otherwise assume
it's an UCS-2/UTF-16 string
- try to autodetect endianess of UTF-16 string by counting zeros
- possibility to enable writing BOM for this field (disabled for now)
- for undefined charset (empty identifier), RT now assume that the
string is what glib think is the local charset (not tested), and try to
convert it to UTF-8 for display/editing
JIS is still not handled though.
9d2ccc19d3