Re: Need for Localized Look of Shared Japanese/Chinese Gliphs??

From: Adrian Havill (
Date: Wed Jul 15 1998 - 04:39:00 EDT

John H. Jenkins wrote:
> You're going to get a gazillion answers to this question, but--

If the question keeps coming up and everybody keeps answering it, maybe
there should be a "FAQ" linked off of the Unicode Homepage. Especially a
Japanese (and maybe Chinese and Korean) versions-- see below.

> The protoypical example is the "bone" radical, U+9AA8, which is drawn one
> way in Japan/Taiwan and another way in the PRC.

Yes, I see this example cited over and over again (as recently as this
month) in Japanese magazines (esp those doing features on Tron) under
"problems with Unicode," using indicating that if you use Unicode, the below
text is "impossible to display":

"The kanji U+9AA8, in Chinese, is written as U+9AA8."

Obviously, the reason this FUD gets repeated over and over is because the
_average_ Japanese-speaking computer user has no access to _correct_
information regarding Unicode.

Just look at what the Japanese reader gets when they use web search engines
for "Unicode":

<URL:> 0 hits
<URL:> 0 hits
<URL:> many hits

(the first hit in Japanese sums up the tone of most articles is an article
based on FUD, which includes the misinfo "America says, 'they're the same
[Han] character so just accept it.'" ("Amerika houmen kara wa 'Onaji kanji
nan da kara gaman shite yo.")

I'll omit the rest of the search hits and the hits related to products, Java
tools, etc. However, I could find only 1 _positive_ article (excluding "how
to use Unicode with Java/Win32" type articles) about Unicode in Japanese,
and only after careful searching.

There is one interesting collection of essays that takes apart that various
critical articles regarding Japanese and Unicode (and JIS) at:


If there are any more like this, please let me know. :-)

On a personal note, I've brought this up before and the some on the Unicode
list said that this is a "non-issue," as the major manufacturers have
already or are implementing new I18N systems using Unicode, so one may
safely ignore the mis-information out there.

I disagree, however, and say that the misinformation regarding Unicode DOES
still affect it's acceptability (at least in Japan).

I'm in charge of development for a medium scale project (90,000 lines of C)
that uses Unicode internally to deal with data that's in Chinese, Japanese,
and English

At my last status report, my supervisor (who is Japanese) discovered that I
was representing plain text data (there is no need for locale info for our
requirements) using Unicode.

He balked, and recited the usual FUD verbatim. The infamous AERA (a very
visible Time/Newsweek-ish Japanese weekly magazine) article "Ima, Nihongo ga
abunai!" ("Japanese is in Danger!") that the infamous article that has
given Unicode a bad name in Japan) lives on.

I had to spend the next hour pleading with him, arguing that the rewriting
to use legacy sets would push the product deadline back (that line always
works wonders with upper-management) and telling him that Unicode is used
internally only; the input and output is still in legacy JIS, KS, and GB
character sets, so the end-users would never notice-- only the software
engineers that browse the source.

His incredible response can be summarized as, "ok, but whatever you do,
don't mention it to [the client] during the formal presentation, because
he'll say the same thing I just said-- and an argument coming from an
American engineer won't help, because Japanese think that a) America rammed
Unicode down Japan's throat and b) the [client that's paying the bill] isn't
an engineer and doesn't understand Unicode, but they DO have pride regarding
the Japanese language and Kanji and they have read the same things I've read
about Unicode, which is written in layman's terms and designed to scare the
hell out of Japanese that are already worried that the language is being
corrupted and bastardized by the younger-techo-foreign-word using

So, I boss will let me use Unicode, but _only_ if I keep it a secret from
the client. Of course, the engineers at the client company will know, as
they will look at the source code, but I'm hoping that because they're
engineers, they'll be more knowledgable regarding Unicode. And I hope
they'll keep it a secret from the boss. (;_;) And if he asks, I'll say "I
didn't want to waste your valuable time with technical details that the end
users will never notice (which is a true statement)."

Ironically, the client is probably using Unicode without even knowing it (if
they're using Windows NT, MS Office, XML, VRML 2.0, Java, etc).

I have noticed that many products that use Unicode in Japan never "brag"
about using Unicode. They quietly implement it, mentioning it in the "notes
for developers" section. Maybe this is because most end-users don't get very
excited about character sets, but I also suspect that using Unicode is NOT
considered to be a "plus" in the eyes of ignorant/misinformed Japanese IT

So, the misinformation about Unicode in the Japanese IT world is so strong
that if I want to use Unicode in a project written for a client, I have to
keep it hush-hush about it, as if I'm using a dirty kludge.

Adrian Havill <URL:>
Engineering Division, Systems Section
an in-the-closet Unicode application developer

This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:40 EDT