ISO 639 - New item approved - N'Ko
Doug Ewell
dewell at adelphia.net
Fri Jun 9 16:55:25 CEST 2006
Mark Crispin <mrc at CAC dot Washington dot EDU> wrote:
>> I would agree with Mark's rather ambitious vision IF that was the
>> kind of registry we were maintaining. But the Language Subtag
>> Registry is a registry of codes, not names. The names are just
>> informative.
>
> In that case, stipulate that the names are in English and restricted
> to ASCII. It's not politically correct, but is the ONLY form that is
> globally useful.
They're not restricted to ASCII. Look at the debate we're having over
non-ASCII apostrophes.
The actual characters that make up the Registry text file are restricted
to ASCII, but they include XML-style escape sequences that represent
non-ASCII characters. (I think Mark understands this; this is largely
for the record.) In the current registry there are only 13 of these (so
far, pending the outcome of the apostrophe debate), but with the
addition of ISO 639-3-based subtags there will be well over 400.
I would dearly love to allow non-ASCII characters directly in the
Registry, in UTF-8, but apparently in 2006 this is still considered
problematic for some users. Hence the escape sequences.
The "English-only" nature of the existing descriptions is not absolute.
It's much like looking through a gazetteer where the most well-known
places are in English, but most of the others are in their native
language. The English for "Nyamwezi" is "Nyamwezi."
If you're saying that tihs is how it should be, not how it is, then we
should take that as a vote in favor of U+0027 in the Great Apostrophe
Debate.
I do stand corrected with regard to my "revisionist" claim that multiple
Description fields were not intended to hold non-English names. That is
one of their purposes; there are others. The point I was trying to make
was that I do not believe multiple apostrophe styles, a distinction that
will be lost on 98% of Registry users and will not even be visible to
some, are one of those purposes.
> My "ambitious vision" was intended to be in descending priority order,
> and subsetable.
I understand this, and think it is an very sensible approach for a
registry of names. I'm just not convinced we have that kind of
registry. I'm willing to be convinced along your lines, but definitely
NOT along the lines that were once presented to us, that in order to be
usable the Registry must contain the names of all languages (and other
entities) *translated into all languages*. That is neither necessary
nor feasible.
> To be honest, if you don't limit it to the English/ASCII name, and
> possibly a second native-language/Unicode name, you effectively have
> no limit.
I think Addison's point was that we indeed have no limit as far as the
"law" is concerned, but we do have a limit as far as our implementation
of it is concerned.
--
Doug Ewell
Fullerton, California, USA
http://users.adelphia.net/~dewell/
More information about the Ietf-languages
mailing list