[Bug-AUCTeX] Re: 11.81; *output* buffer is set to UTF-8

[stefan pofahl]

David Kastrup <dak <at> gnu.org> writes:

> 
> Stefan Pofahl <stefan <at> pofahl.de> writes:
> 
> > David Kastrup <dak <at> gnu.org> writes: 
> >  
> >>  
> >> Peter Dyballa <Peter_Dyballa <at> Web.DE> writes: 
> >>  
> >> > Am 06.11.2005 um 10:09 schrieb David Kastrup: 
> >> > 
> >> >> The question is whether the _characters_ appearing in the buffer are 
> >> >> what you would have expected. 
> >> >> 
> >> > 
> >> > There are few problems before answering this ... 
> >> > 
> >> > I need a mathematical environment with ö or ä -- an ü would do it too, 
> >> > I presume. Or an ß instead of β? Well, I have no idea whether this 
> >> > nonsense is supported by TeX. 
> > I had also the problem with the wrong saving of the file, 
> > this is what I wrote into my init.el: 
> >  
> > ------------------------------------------------------------------------- 
> > ;;; tell (x)-emacs what coding-system it should use under `LaTeX-mode':  
> > (add-hook 'LaTeX-mode-hook 
> >         (lambda() 
> >            (set-buffer-file-coding-system 'raw-text) 
> >        ;(setq coding-system-for-read 'raw-text) 
> >        ;(setq buffer-file-coding-system-for-read 'raw-text) 
> >        ;(setq file-coding-system-alist 'raw-text) 
> >            (setq file-coding-system-alist  
> >               (cons '("\\.\\(?:tex\\)\\'" . raw-text) 
> > 'process-coding-system-alist)) 
> >            )) 
> > ------------------------------------------------------------------------- 
> 
> I have rarely seen any advice more disastrously wrong.
> 
> First you set the `file-coding-system-alist' in `LaTeX-mode-hook': at
> this point of time, one file using LaTeX-mode has already been loaded.
> Then you destroy the variable  `file-coding-system-alist' completely
> (and globally) every time LaTeX-mode is entered with a nonsensical
> setting concocted from the nonsensical coding system `raw-text' for
> TeX files, and what likely was supposed to be the contents of the
> completely unrelated `process-coding-system-alist', but actually
> amounts to the symbol, rather, making `file-coding-system-alist'
> contain an invalid value.
> 
> And then you set the buffer file coding system to 'raw-text', too.
> `raw-text' is a unibyte format for byte streams that are not
> interpreted as characters.
> 
> In civilized countries with active Emacs worship, actions like this
> could cause an ostracism, and indeed this would explain why it took
> you more than a year to read this report and reply to it.
> 
> Just for the record: what version of Emacs/XEmacs were you trying to
> butcher into submission in this manner?  XEmacs 21.4 nomule?  Because
> that is pretty much the only system I can think of where the
> consequences might not lead to disaster rather soon and obviously.
> 
Hallo David,

thank you for your quick response :-)
I'm using the actual debian distribution (sarge). 
I tried to make an update of auctex from within xemacs(-mule),
but the response was, I'm up to date.
First I was not shure, why I can not work with umlauts,
and than I started a research and found this discussion here.
As non-elisper I'm very proud how far I get, 
belive it or not, know it works :-)
(even if it is a crude/disastrous peace of code).
Please tell me what should I do?
What is the reason? Do I have to install another version
of auctex? Or can I put something in my init.el.
(At the moment I can not check the version of auctex and
xemacs, I'm sitting at another machine at the moment,
but I will check and report).

Best regards,

Stefan