Hello, Change http://svn.apache.org/viewvc?view=revision&revision=1801052 tries, in case no charset is specified for the response, to determine a default one based on the content language if such is present. For en and fr languages the ISO-8859-1 charset is used as default.
However, this is done for all content types, not only for text media types. Is this intentional? For example, for some media types (particularly application/json) this could lead to incorrect representation of characters. Kind regards, Lazar
