Dana S Emery writes:
keith moore > [is ISO 639/1967 sufficient]
It must minimally distinguish of C/J/K/T (traditional, or taiwanese
chinese),
are all 4 provided distinct codes?
No, there is no Taiwanese/traditional Chinese code
Then, the only politically safe was is to use ISO celebrated distinction
of characters (or, if you are an ISO-terminology-purist, glyphs).
It should be noted that, the combination
language_name/country_name
is also unsafe and unstable. For example, what, do you think, is
chinese/hongkong
now? Should the Hongkong be considered as a valid country name? How
will it be after Hongkong is returned to China?
DIS 10646-1.2 distinguishes Hans as:
G: mainland china
T: taiwan
J: japan
K: korea
So, we should have:
charset=iso-10646-g-*
charset=iso-10646-t-*
charset=iso-10646-j-*
charset=iso-10646-k-*
where "*" is replaced with Devanagari variations. As DIS 10646-1.2 cites
ISCII as its source of Devanagari characters, Devanagari distiction
should be done according to Indean standard. Are there any Indean
standard which lists names of Indean languages in Latin alphabets?
By the way, the technical contents of ISO 639 (1988) is available
in dkuug.dk:i18n/ISO_639
I found its content is overly simple.
For example, as for the name of the language, "Chinese" is too much
broad. There are Cantoniese, Pekingese and so on.
I also found that Ainu, Japanese minority language, is missing.
Though those defects do not affect the capability to render
simple/plain text as simple/plain text, if you are expecting
the language information more, for example, spell checking, ISO 639
is of no use.
Thus, to have a new header:
Conteent-language:
for general use other than simple text rendering is, overkill.
Masataka Ohta