🚀 go-pugleaf

kyle_jones@wonderworks.com (Kyle Jones) wrote in message news:<ua1fpnpdlohv40@corp.supernews.com>...
> Emmanuel Beffara <emmanuel@beffara.org> wrote:
>  > I am encountering a strange behaviour with VM (my version is 7.01 on
>  > Emacs version 21.1): when I recieve mail encoded in charset
>  > iso-8859-15, it displays without any problem. When I want to answer to
>  > it (quoting the original message), I get the correct characters in the
>  > input buffer. However, when sending the answer, it gets re-encoded in
>  > charset iso-2022-jp. Not only does this disturb the people I answer to
>  > (most people in Europe don't need support for Japanese encodings,
>  > after all), but also VM itself is unable to read correctly the copy of
>  > the message it sends.
>  > 
>  > Would anyone know how to fix this?
> 
> I don't think it's fixable without hacking the code.  I think the
> kludge fix is to edit vm-determine-proper-charset and replace
> iso-2022-jp with whatever charset you want to use.  UTF-7 and UTF-8 are
> better choices for messages that contain multiple character sets,
> but Emacs doesn't have full support for those coding systems yet.

I had the same problem recently. Actually, for me it also appears when
I write some German language message with non-ASCII characters. The
reason appears to be that with Emacs 21 / SuSE Linux 8.0 the default
character set is now iso-8859-15 rather than iso-8859-1.

The following fix worked for me (at least with Emacs 21 and VM 7.03):
Add '(latin-iso8859-15 "iso-8859-15") to
vm-mime-mule-charset-to-charset-alist.

I suggest to add this in the distribution.

Hans