[Nmh-workers] Deprecating MM_CHARSET

[Ken Hornstein]

We agreed (in January of 2012) to deprecate MM_CHARSET.  That would mean
all nmh programs would get the character set information via the standard
locale environment variables, which I assume is still non-controversial.

I am wondering if we can assume that nl_langinfo() exists on all
platforms that we care about.  It's part of POSIX.  I ask because if it
doesn't exist I'm not sure how we get the current character set.  My
feeling is in this day and age we should assume it exists.

Also ... I am wondering if we should remove the ability to disable
the locale code (--disable-locale).  We haven't been very consistent
on #ifdefing everything more recently when it comes code that depends
on the locale, and I don't see any value in disabling it.  I think we
should simply remove the ability to disable locale support (this is
distinct from support for the multibyte POSIX functions, which would
still be conditional on those functions existing).

--Ken