On Wed, Jul 6, 2016 at 3:28 PM, John Dennis <jdennis@redhat.com> wrote:

+1 for UTF-8

Take it from someone who has done more than his fair share of solving internationalization and character encoding issues that UTF-8 is the current defacto standard today across a wide range of technologies. It's almost universally agreed that specifying ISO-8859-1 as a default was a regrettable mistake from an earlier era before internationalization matured. There is an old proverb "no matter how far down the wrong road you've traveled it's never too late to turn back", that could be the motto for ISO-8859-1 :-)

I share your pain, but most of my problems have been due to applications not following standards or by expecting the system encoding to be something it's not and not allowing you to specify the correct encoding.

I also agree with you that UTF-8 is the encoding to use, UNLESS it's common practice/a de-facto standard to be using something else which it is for .properties files. If everybody presumes they are ISO-8859-1 then we should not confuse people by defaulting to something else.

That being said I think using a header to specify the non-default encoding, eg. UTF-8 is a good compromise.

Best regards,

Thomas