(at first,my english skill is poor.)
in japan, we use multiple charset.
so like this.
client may use shift_jis encoding.
but maybe blog uses utf-8 encoding.
so the strings are not displayed properly.
then shitf_jis to utf-8 conversin needed.
so japanese xmlrpc client only supports utf-8 encoding.
of cause, ideally, blog client should detect the encoding of response,and convert it to the native encoding of the client.
but if then, problem remain. still, it can be differnt between blog charset and client charset.
so we do like this.
client (some encoding to utf-8)==>>utf-8===>>
===>>utf-8===>>server(utf-8 to some blog encoding).
and retuen.
client(utf8 to native charset)<<===utf-8 <<==server(some charset to utf-8)