Re: [idn] case preservation

["ben" <ben@cc-www.com>]

From: "Martin Duerst" <duerst@w3.org>

> This is a good point. But please note that you don't actually have
> to know whether a character is simplified or traditional. If you
> work on a simplified Chinese system or with a simplified Chinese
> input method, you will just get simplified Chinese characters
> as the first choice from you input method. Same thing the other
> way round for traditional Chinese.
>
You *do* need to know whether the characters of the CDN are
traditional or simplified.  This is analogous to the 'l','1' issue.
One must be aware enough to recognize that the 'l' is the 'lowercase
L' and not the 'number one'.

> This means that if somebody wants to have a domain name usable
> both is China,Singapore,.. and in Taiwan,Hong Kong,..., they
> have to do two registrations. Of course, there are cases where
> there are more registrations needed (e.g. for simplified
> characters used in Taiwan), but these are exceptions.

Yes, this is certainly one of the ways to achieve TC<->SC equivalence.
Again, this is analogous to someone registering both the '1.TLD' and
the 'l.TLD' (or maybe even 'I.TLC') to cover all bases.

>
> Also, for most characters where there is actually a difference
> (there are many cases where there is none, and which are
> therefore unified), it is quite easy to tell whether they
> are simplified or not. The main reason for this is that
> most simplifications are very systematic. Of course, because
> there are so many Chinese characters, there are always some
> that you don't know. That is something that applies to any
> other aspect of Chinese characters.
>

For people who use exclusively TC *or* exclusively SC in their
everyday lives,  it is easier.  However, for people who use them
interchangably, it is much harder.  But in any case, every user must
atleast make the conscious effort to be aware of whether a CND  is SC
or TC much the same as a user must make the conscious effort to be
aware whether 'l.TLD' is a 'number one' or a 'lowercase-L'.

Now... ofcourse, whether this justifies having TC<->SC conversion as a
standard is obviously not up to me to decide... but up to this WG.
(As I have humbly learned from my experience with my "Supreme CDN
system"- although supreme, is not interoperable with others.)
However, I do hope that this analogy using 'l','1','O','0' will
atleast offer much more understanding as to why TC<->SC is such a
critical issue for the Chinese culture.

Thanks,
Ben Chan