UTF-8 in documentation files

[Paul Eggert]

Ralf Wildenhues <address@hidden> writes:

> The name of the bug reporter is Пухальский Юрий Андреевич, but since
> UTF-8 is not supposed to appear in text documentation[2], I replaced
> it with the ASCII transcription.

Thanks.  For Autoconf we still use plain ASCII.  Autoconf tends to be
trailing-edge here, since it supports lots of old systems, so I'd
rather keep it ASCII for now.

Also, for most English readers, a Latin transcription is far more
readable, so it would be preferable even if we allowed UTF-8.  Plus,
it helps non-Cyrillic-readers to record Yury's preferred Latin
transcription ("Puhalsky" and not the more-common "Pukhalsky").

That being said, other GNU packages (e.g., gnulib, tar) have used a
bit of UTF-8 in comments and ChangeLog entries, for names like
François Pinard that are easily readable in English.  As far as I know
nobody has reported problems.  Since gnulib is used by several other
packages I expect some UTF-8 will leak into many other GNU
distributions.

In <http://lists.gnu.org/archive/html/bug-standards/2005-03/msg00004.html>
RMS said that he does not want to recommend UTF-8 yet.  That is quite
reasonable, but as far as I can tell it doesn't place a requirement on
other GNU projects to avoid UTF-8 in comments and ChangeLog entries.
I think it makes sense for projects like tar (whose maintainer is in
the Ukraine) to be guinea pigs in this matter.