utf8 wrongly encoded chars in draft messages

[Uwe Brauer]

Hello

I am using Ma Gnus v0.10 with Xemacs 21.5.34 Mule without problems.

Now I started to use GNU Emacs 24 with gnus as well, and since I started
I have problems with the coding in the draft message for example I see
symbols like

¡

Which I can decode via (decode-coding-region start end 'utf-8 nil))

To the correct

¡

However when I use C-c C-d the coding is again wrong.

I am not sure who is here the culprit. I tried to play around with
(setq message-draft-coding-system 'utf8)

But this did not help. Any help is welcome.


thanks

Uwe Brauer

[Adam Sjøgren]

Post by Uwe Brauer
I am not sure who is here the culprit.

Maybe something in your configuration?

For what it is worth, it works for me (GNU Emacs 24.4.1, Ma Gnus v0.14).

The draft does not seem to have a header indicating charset yet, maybe
you (or I) have some defaults set for encoding that are unexpected?


Best regards,

Adam

[Uwe Brauer]

Post by Adam Sjøgren

Post by Uwe Brauer
I am not sure who is here the culprit.

Maybe something in your configuration?

Right! The following (stupid) setting

(add-hook 'gnus-message-setup-hook 'my-set-utf8-on)


(defun my-set-utf8-on ()
(interactive)
(set-buffer-file-coding-system 'utf-8 nil))

Caused the problem. I removed the function and now everything works as
expected.

Sorry for the noise.

Uwe