Re: UTF-8 text files

[Ralf Wildenhues]

Hi Alexandre,

* Alexandre Duret-Lutz wrote on Mon, Mar 21, 2005 at 10:15:37AM CET:
> >>> "Ralf" == Ralf Wildenhues <address@hidden> writes:
> 
>  Ralf> Didn't add a thank you note to the last ChangeLog entry yet --
>  Ralf> are the Libtool developers alright with UTF-8 encoded ChangeLog*,
>  Ralf> THANKS, (maybe eventually AUTHORS) files?
> 
> Automake has utf-8 ChangeLog and THANKS file since a few months.
> Gettext started a few years ago (IIRC it even uses unicode
> characters in C comments).

IIRC there are compilers which don't like putting it in C source.
Don't remember, though, whether comments only were ok.

> I think it is safe to use utf-8 today as long as it is not in
> (portion of) files that get installed in user packages.

OK, good.

>  Ralf> Any emacs-foo I'd need to add at the top or bottom?
> 
> ;; Variables:
> ;; coding: utf-8
> ;; End:
> 
> at the end, or 
> 
> -*- coding: utf-8 -*-
> 
> at the beginning.  Or nothing and rely on autodetection.

Beginning is not so nice, because that makes merging patches a little
more clumsy.  At the very end is not so nice for Libtool, because that's
where the $Revision$ comes, used for some rebuilding stuff.

I'll just rely on autodetection and see who breaks it first.  :)
(probably that's going to be me, since I have a non-UTF8-ready system I
sometimes work on.)

Regards,
Ralf