[illumos-Developer] webrev: POSIX style localedef & multibyte encoding support

[Garrett D'Amore]

On Mon, 2010-10-04 at 16:46 +0200, Joerg Schilling wrote:
> Alan Coopersmith <alan.coopersmith at oracle.com> wrote:
> 
> > > A quick grep of my mailbox reveals thousands of:
> > > Content-type: text/plain; CHARSET=US-ASCII; format=flowed
> > > 
> > > I've no idea what a mail reader might do if that become an unknown
> > > locale.
> >
> > Probably not even notice, since charset & locale are two different
> > concepts, and I don't think Solaris has ever shipped a US-ASCII
> > locale, since it doesn't match the locale name scheme - it would
> > be something like en_US.ASCII to include region, language & charset.
> >
> > On Solaris, "locale -a" does not list US-ASCII, and both of these
> > just show fallback to C:
> > 	env LANG="US-ASCII" locale
> > 	env LANG="bob" locale
> 
> US-ASCII is not a locale but a coding (see above, there is "CHARSET=US-ASCII"). 
> You see it with iconv -l

Right.  I had introduced en_US.US-ASCII (and a bunch of other
en_*.US-ASCII) locales with illumos a while ago.  This change undoes
that addition -- there is no value in them, and maintaining them is more
trouble than its worth.

People can use en_*.UTF-8 or en_*.ISO8859-* and still get all the same
functionality, *plus* have support for extended characters.  (For UTF-8
its pretty much support for *all* characters... for ISO8859 its just the
usual Latin-1 or Latin-9 characters.

	- Garrett