tips for starting esnx with the correct codepage

Hi guys,

I am having trouble understanding what's goingon on two of our Linux servers. One is running Ubuntu 14.04, the other 16.04.

We start our headless runtime image with the following export right before the start of esnx:

export LANG=de_DE.iso88591

During startup of the image, the result of (AbtCodePageConverter currentCodePage) is logged into a file.

The image on the 14.04 server logs the correct codepage, but the 16.04 logs 'UTF-8'. I've checked that the language-pack-de is installed on both and locale -a listst the locale as available on both machines. Yet one works and the other doesn't - using the same image, VM and startup script.

Any ideas?

Joachim