bug#34350: 27.0.50; ediff-revision broken with SVN backend + non ascii chars both in directory and in filename

[Eli Zaretskii]

> Cc: 34350@debbugs.gnu.org, monnier@iro.umontreal.ca,
>  vincent.belaiche@gmail.com
> From: Dmitry Gutov <dgutov@yandex.ru>
> Date: Sat, 9 Feb 2019 10:58:38 +0300
> 
> > coding-system-for-write
> > affects both I/O from and to a subprocess and the encoding of the
> > command-line arguments we pass to that subprocess.
> 
> Aren't there any programs that expect one encoding for their arguments, 
> yet output contents in other encodings sometimes?

It happens, yes.  Which is why by default coding-system-for-write is
not bound.  When you bind it while invoking a subprocess, you need to
know what you are doing.  If there's a conflict between the two
encodings, there are a couple of ways to deal with that.

However, in the case in point, my reading of the code indicates that
there's no need to bind coding-system-for-write at all.  It seems we
do that due to some copy/paste habit more than due to a real need.
Let's see if I succeed in convincing you and Stefan.