Encoding problem

[Sven Bretfeld <sven.bretfeld@relwi.unibe.ch>]

Hi to everybody

I still have my problem with the encoding of emails. Some days ago I
asked the list already but I haven't got an answer. Now it becomes
more and more molesting and I didn't find any solution yet.

Sometimes when I reply to an email citing parts of the original, the
encoding of my reply changes to charset=iso-2022-jp without Emacs
telling me. This destroys German umlauts of my own and the cited
text. I haven't found any reason or system. It seemingly doesn't
depend on an unknown charset of the original message or any MIME
contained. I'm using VM.

I have two questions:

1. Does anybody know the reason for this behaviour and a way to stop
   it? There must be any option like vm-always-reply-with-charset.

2. How can I manually check which charset I'm using in the present
   text if Emacs doesn't give any clue like -t or u in the status
   bar. And how can I manually change the charset to, say, ISO-8859-1?

Thanks for your help.

Sven