> On Sep 18, 2017, at 2:25 PM, Julian Reschke <julian.reschke@xxxxxx> wrote:
>
> The problems are tools that do *not* have the metadata. That is: download the RFC and double-click it in Windows.
Legacy systems will continue to run in some non-UTF8 code page for
some time, but with any luck newer systems are moving on.
FWIW, I have exceedingly little trouble with UTF-8. On all my
systems (Debian, MacOSX, NetBSD and FreeBSD) I have:
$ env | grep '^L[AC]'
LANG=C
LC_CTYPE=en_US.UTF-8
So UTF-8 is expected by default in all data. The only time I have
to worry about other encodings is the occasional "whois" server that
returns iso-8859-1 instead of UTF-8, and I have to run the results
through iconv(1):
$ whois -h whois.registro.br registro.br |
iconv -f iso-8859-1 -t utf-8 |
perl -lne 'print if m{[^\x00-\x7f]}'
owner: Núcleo de Inf. e Coord. do Ponto BR - NIC.BR
It will take decades, but I hope that eventually most systems
will run in a UTF-8 locale by default, and exceptions will become
increasingly rare.
--
Viktor.