A proposed solution for descriptions
mark.davis at icu-project.org
Tue Jun 20 06:07:37 CEST 2006
I agree with Addison. I'll go a bit further. The goal is to have
enough of a description to distinguish the language from others.
That's the goal. It should not be a goal to have a gazillion
languages, nor does the representation need to capture all the
possible variations in spelling or punctuation.
On 6/19/06, Addison Phillips <addison at yahoo-inc.com> wrote:
> > >
> > > I don't support changing any letters in the registry. Personally:
> > > I think we're best off if we just replicate whatever the
> > ISO MA sends
> > > to us. Even when different ISO MAs send us something
> > different for the
> > > "same thing" (as with N'Ko). If "we" have a problem with it---tell
> > > *them* and let them work it out.
> > Well and good, but searching for "N'Ko" in the registry with either
> > grep or Google will find nothing at all. I don't think that's
> > an acceptable state of affairs.
> So my weasel words come into play here: I don't support *changing* any
> *letters* in the registry. Apostrophe vs. right-single-quote I don't have
> such a problem with (and I +1ed a proposed record yesterday to that effect).
> Admittedly, right there I cite N'Ko. In that case, I think putting the ASCII
> apostrophe into one description will not cause the planet to spin off its
> axis. We should be vaguely smart about this, not so rigid that we cannot
> make good decisions. But changing Cote d'etc. we should forget. And I don't
> believe that it would be *that* bad if temporarily we put the curly quote in
> the registry. If someone feels gipped, they can propose an additional
> description for registration (problem solved). But as a general policy:
> 1. Register whatever ISO 639 (etc.) send us, exactly as they send it
> (possibly unbundling multiple names, ala the Old Slavonics)
> 2. Register additional descriptions via the consensus process (that's what
> it is for), separately.
> Frankly, I'll be happiest when we can go to a pure text representation. Most
> user agents, tools, search engines, etc. can deal with searching plain text
> given the encoding.
> Ietf-languages mailing list
> Ietf-languages at alvestrand.no
More information about the Ietf-languages