--On 21. mars 2002 16:47 +0859 Masataka Ohta
<mohta(_at_)necom830(_dot_)hpcl(_dot_)titech(_dot_)ac(_dot_)jp> wrote:
However, as Harald and I agreed, such products need local
context information OOB to work properly.
apologies for indicating too strong an agreement....
we agree that for many functions that users would like software to perform
(sorting being an obvious example), context information is needed to work
properly.
Then, the problem is that some local contexts are mutually exclusive.
There is no instance of Unicode-based-local-character-set
internationally useful.
I assert that for many functions (such as non-high-quality text display) in
many locales, having information in Unicode-based character sets are useful
without OOB information. Thus, Unicode is internationally useful - but we
are deluding ourselves if we claim that it solves ALL problems.
Some product using unicode may be usable in multiple local context,
if the context information is given OOB, and the product is valuable
in international market.
However, unicode, still, is not usable in international context unless
OOB information, which, thanks to poor charset reviewing, essentaily is
yet another charset designaiton, is attached.
We disagree on this point.
I assert that the fact that charset does not give the required information
is by DESIGN of the charset mechanism, not by "poor review".
As a citizen of a continent where > 10 nations use the same character set
(ISO 8859-1), but where locale requirements vary across the continent, I
believe that separation of those concerns was a correct engineering
decision.
Just to make it clearer where Ohta and I agree, and where we disagree.....
Harald