Gentlemen!
Could we
1. cool down the debqte
No no, today is the time to sip some (or a lot of) Beaujolais nouveau
and heat up. It's cold. But, as an internationalization specialist,
I must say that people in Southern Hemisphere may cool down. :-)
2. most important, put all your highly interesting comments
in their appropriate context
One of the problem is that some of us who does not use Latin
alphabets so much strongly needs localization.
Such people have already FINISHED localization and now feeling
the need for the *REAL* INTERNATIONALIZATION.
For example, Japan defined national Chinese character standard in
1978 and used it on the Internet including e-mail and e-news for
about 10 years. Japanese whois service is also available and is
quite convenient for those people who can read Japanese. I can
say that on the Internet, Japanese localization has been quite
successful.
It should be noted that we have had strong requirements to support
US-ASCII and Chinese characters AT THE SAME TIME.
And, now, we are feeling that we need interoperale encoding
within Asia, at leat. But, unlike major European coutries, ISO 8859/1
nor ISO 10646 (UNICODE) can't let us interoperate. Thus, it is
considered to be natural to support many characters at once. It is
not so more difficult to support ASCII and national characters.
10 years ago, it was not so easy and performance could have been
a problem. But, now, we use muliple character sets daily, without
any difficulty.
But, I must also confess that Japanese localization is the Babel.
That is, different encoding methods are dominant in 1) the Internet
(ISO-2022-JP) 2) PC users (Shift JIS) and 3) UNIX users (EUC-UJIS).
But, we have learned something. That is, ISO-2022-JP and Shift JIS
can be mechanically distinguishable and may be freely mixed. So is
ISO-2022-JP and EUC-UJIS. But not Shift JIS and EUC-UJIS. That's
why we designed ISO-2022-INT-1 to be mechanically distingushable
and may be freely miixed with a lot of character sets.
Yes, LOCALE mechanism on some UNIX machines can distinguish
different localization. And, with a lot of *REAL* experience,
we know that they are quite inconvenient. Users don't want to
switch LOCALE at all.
So, already several years ago, what we need was the *REAL*
multilingual support without switching. That was the time to
distill.
On the other hand, Latin alphhabet users could have lived with
ASCII or national variant of ISO 646 (Swedesh case).
They have started localization (including InterWesternEuropeanization
by 8859/1) just recently.
So, it is not surprising to me if some people of European cluture
(including US) think suporting multiple localization is the
internationalization.
I'm interested in hearing opinions from Greek people who may not live
with ISO 8859/1.
As interEuropean communication ferments (some may rot), I think people
of European culture will feel the need for the *REAL* internationalization.
The discussion is wrongly comparing discussing 3 completely separate
issues (at least). That is worse that "apples and oranges", worse
even than "chevres et choux" (ie. goats and cabbages).
So, I think that the issue is a little less worse to be "grapes,
wine or brandy". I eat/drink all of them but prefer Chambertine
and Bas Armagnac. Don't you?
Masataka The Drunken Ohta