Javadoc
Sets the
java.io.OutputStream that receives the changelog.
This takes
java.io.Writer and not
java.io.OutputStream because the changelog is a textual format,
and therefore it is a stream of chars, not bytes. (If the latter, then we'd be unable to handle
multiple encodings correctly)
According to man git-commit, the "encoding" header specifies the encoding of the commit message,
and git CLIs will try to translate encoding back to UTF-8. In any case, it is the implementation's
responsibility to correctly handle the encoding