Re: Another attempt at plain language

I believe that overly strict criteria are being applied to variants, in
general. If we applied the same sort of criteria to languages, scripts, or
regions, think of what a mess it would be! The fact that one can
generatively produce "es-US" or "ru-US"—without going through a long
process of registering "specific and well-defined" variants—is of great
benefit.

The same is true of general variant subtags, like 'fonipa' or 'plain'. It
would be pretty darn'd useless if every single language + fonipa had to be
specifically defined, and use a gratuitously different subtag for the same
general semantic. Think of if you couldn't say "fr-fonipa", but instead had
to have "fr-ipafr15b" for some specific expert's formulation for French,
then register "it-ipait666" for some other's Italian formulation, etc. It
is much better to have a general subtag like fonipa, and then—*if needed*—allow
registration of another tag that is more specific.

The same would be true of 'plain'. It could have a pretty clear general
semantic, and be applied with any tag. Only if people really, really wanted
very specific variants would it be necessary to add them as well. See also
John Cowan's comments.

Mark <https://google.com/+MarkDavis>

*— Il meglio è l’inimico del bene —*

On Fri, Sep 4, 2015 at 10:12 AM, Michael Everson <everson@evertype.com>
wrote:

> Specific and well-defined ones would be few and far between, I’d think.
>
> > On 4 Sep 2015, at 07:01, Martin J. Dürst <duerst@it.aoyama.ac.jp> wrote:
> >
> > I would agree, and guess that many on this list would agree, that a
> specific and well-defined CNL would be in scope (mostly through a specific
> variant subtag registration) for BCP 47.
>
> Michael Everson * http://www.evertype.com/
>
> _______________________________________________
> Ietf-languages mailing list
> Ietf-languages@alvestrand.no
> http://www.alvestrand.no/mailman/listinfo/ietf-languages
>