Re: PostgreSQL and Unicode

hackers:

I got following mail:

> I was just looking over the PostgreSQL 7.0 docs and noticed that there
> doesn't seem to be any new features for Unicode support. I wanted to verify
> if this is true?
>
> Right now we have a database that must support many different languages.
> This works ok when we use UTF8 but the problem is that we do not know how
> many characters the text will be. I was hoping that PG7.0 would support true
> Unicode (2 byte) instead of just UTF8. Do you know if there is any plan to
> support plain Unicode?

I think supporting "true Unicode (2 byte)" (probably that means UCS-2)
is not that easy since it includes '\0'. We need to fix at least:

the parser
libpq
psql
all client programs ...

Another idea might be doing a conversion between UTF-8 and UCS-2
somewhere between frontend and backend. However we still need to fix:

libpq
psql
all client programs ...

in this case. Any idea?

By the way, does anobody know what's wrong with UTF-8? In my
understanding UTF-8 and UCS-2 are logically identical.
--
Tatsuo Ishii