Re: character set for Russian and French reading Google News

[Katsumi Yamaoka <yamaoka@jpl.org>]

soichi wrote:
> I have added
>  (setq mm-text-html-renderer 'w3m)
> into .gnus.el, but no change.  I believe that emacs-w3m is the latest.

You might have the corrupted nnrss files[1] that were made before
you've set `mm-text-html-renderer' to `w3m'.  How about re-creating
those groups (i.e. delete groups by typing `C-u G DEL' in the group
buffer one by one, and do `G R' again)?

> More precisely, the following happens.
> When I choose an entry of the feed in the summary buffer, Emacs asks,

> "Message contains characters with unknown encoding."
> "use ASCII as charset?"

> if I choose "y" for it, the fonts become a mess. "n" does pretty
> much the same.

[...]

So far I've never faced that problem, though the one I checked was
only that Russian feed in question.  Sorry for the lazy response.

[1] Those files exist in the directory ~/News/rss/ (the default)
and each name consists of the nnrss group name and ".el".  For
instance, "В мире – Новости Google.el" or somthing
is for the group:
http://news.google.co.jp/news?pz=1&cf=all&ned=ru_ru&hl=ru&topic=w&output=rss