Re: [Ltru] Macrolanguage usage

I think that would be fine.

On Fri, May 16, 2008 at 12:03 PM, Randy Presuhn <
randy_presuhn@mindspring.com> wrote:

> Hi -
>
> As a technical contributor...
>
> > From: "Mark Davis" <mark.davis@icu-project.org>
> > To: "Doug Ewell" <doug@ewellic.org>
> > Cc: "LTRU Working Group" <ltru@ietf.org>
> > Sent: Friday, May 16, 2008 11:08 AM
> > Subject: Re: [Ltru] Macrolanguage usage
> ...
> > The only thing that the text is saying is that when there is a
> predominant
> > language, for backwards compatibility -- AND with some careful caveats --
> > the above is broadened. This is *not* with respect to permissibility, but
> > with respect to recommended practice.
> >
> >    - It is *permissible* to tag Mandarin text with "und" or "zh" or
> "cmn".
> >       - It is a *good idea* to tag/lookup with "cmn"*, except that for
> >       backwards compatibility, "zh" may be better for many
> implementations.*
> >    - It is *permissible* to tag Standard Arabic text with "und" or "ar"
> or
> >    "arb".
> >       - It is a *good idea* to tag/lookup with "arb"*, except that for
> >       backwards compatibility, "ar" may be better for many
> implementations.*
> >
> > That lets people like Google, which has a very serious and important
> issue
> > with backwards compatibility, support the vast amount of
> Mandarin/Arabic/...
> > that is tagged with "zh", and continue to tag Mandarin content with "zh".
> It
> > also lets people like Shawn's group at Microsoft, having no backwards
> > compatibility issues with "cmn" and "zh", to shift over immediately.
> >
> > It is also orthogonal to the recommendation for non-predominant
> encompassed
> > languages, where we would tag with the more specific tags: "yue",
> > "yue-Hant-HK", "yue-Hant-US", "abh-AF", and so on, once they are
> available.
> ...
>
> I'd re-frame it in slightly different terms, as an example of the "tag
> wisely"
> principle, rather than as anything that might be understood as a
> recommendation.
> "For some applications, it is possible that there will be cases where it
> could
> be advantageous to deliberately use a tag which is imprecise.  For example,
> due to the peculiar situation of written Chinese, an application with large
> quantities of legacy textual data already tagged with 'zh' might countinue
> to use this over-broad tag even for new data which would more precisely be
> tagged with 'cmn'."
>
> Randy
>
>
> _______________________________________________
> Ltru mailing list
> Ltru@ietf.org
> https://www.ietf.org/mailman/listinfo/ltru
>

-- 
Mark