Re: ISO/IEC 10646 as Document Character Set

Gavin Nicol (
Wed, 3 May 95 19:21:58 EDT

>Whatever my convictions about how content *should* be represented or
>labelled, the Japanese in particular have a strong track record of
>doing precisely what they want, whether or not it makes life
>difficult for anyone else.

Ack. This is true to a degree, though insufficient education plays a
large part too, as does the fact that until very recently, very little
thought was given to internationlisation at all, so the Japanese *had*
to invent their own solutions. However, there has been a marked
difference in corporate mentality in the last 5 years. Companies are
now looking outside for solutions, and are moving to standards-based

The WWW has caught many Japanese companies with their pants around
their ankles, and they have shown a great willingness to be blindly
led (or to rephrase, they are looking for packaged solutions). We are
just at the start. We can guide them, or even push them, in the right
direction *now*, before they start inventing "solutions" of their own.
In 6 months more, we'll be in an absolutely hopeless position.

>And being a commercial vendor who receives a large chunk of revenue
>from Japan, I am interested in solutions, even stopgap ones, that do
>not depend on convincing the entire Japanese part of the Internet to
>upgrade before I can sell them software. And right now there's
>nothing a content author can do to make their content more
>intelligible, since all labelling is done by the server. The same is
>true in Europe, also: Norway is still using three separate character
>encodings, as are many other countries.

I'll give all commercial vendors a hint: enforce the standards, or be
plagued with support problems from here on. The Japanese now want to
buy packaged solutions. If your packaged solutions do the right thing,
they'll do the right thing. Doing anything else shows a desire for
profit, and a lack of long-term vision.

The only other problem we face is the free software, which is being
widely deployed here. Again, though, if the base sources include
support, that support will filter over here too. This is a hint to
CERN, NCSA, and W3.