[idn] A question about the display of decoded IDNs

[Yves Arrouye <yves@realnames.com>]

Hi,

Pardon me (and please point me to the right place!) if this topic has been
discussed. I searched in the titles of the mailing list archives, but didn't
seem to find a related thread, nor is the IDNRA draft addressing this issue.
 
As I was generating an XHTML document with a bunch of ACE-encoded strings
and their decoded IDNs, I wondered how applications in Asia will be able to
pick the correct glyphs for display of some characters?

My understanding is that a result of the Han unification process in Unicode
is that some characters with culturally different glyphs are shared between
Chinese and Japanese (and Korean?). In HTML and XML, people use a language
attribute to let the rendering software pick the glyphs appropriate to a
given language.

In IDN, I haven't seen such tagging in the ACE encodings I've looked at
(namely RACE and LACE; pardon me if this has been addressed in other ACEs).
As a matter of fact, the ACE is always being said to be applied to a UTF-16
buffer, with no mention of language tagging (wether the name uses a single
language or many). Which means that there may be an upcoming display problem
for IDNs.

Will IDN support the upcoming tag characters U+E0000-U+E007F when Unicode
3.1 will be released? Won't such tagging be costly and make the names even
shorter (a tag would use something like three code points, say U+E0001
LANGUAGE TAG followed by a two-letter language code, e.g. 6 code units, and
force a plane change and back)? If .com still prevails with IDNs, one won't
certainly be able to make (risky) assumptions on the language based on the
TLD of the domain name.

YA
--
My opinions do not necessarily reflect my company's.
The opposite is also true..