Re: Inserting Unicode Data.

From: Frank van Bortel <fvanbortel_at_netscape.net>
Date: Tue, 02 Nov 2004 17:12:28 +0100
Message-ID: <cm5nbc$h4l$1@news4.zwoll1.ov.home.nl>

Howard J. Rogers wrote:

> Frank van Bortel wrote:
>
>

>>Howard J. Rogers wrote:
>>[snip]
>>
>>
>>>It might also have something to do with the fact that UTF16 is the only
>>>national character set you can have in 9i.
>>>
>>>Regards
>>>HJR
>>
>>Are you sure about that? The doccu (Globalization Support A96529-01,
>>pg. 5-9, or pg 137 states you can use UTF8 as well as UTF16,
>>UTF16 being default.
>>Is this a documentation bug?
>>I have never created a 9i database with UTF8, but always with
>>the default UTF16, so I would not know first hand.

>
>
> OK, time for me to stop being so imprecise.
>
> In 9i, you can choose AL16UTF16, or UTF8. True.
>
> Note: UTF8 *not* AL32UTF8.

Should have been more precise :)
>
> But the documentation is (or should be!) quite clear on the fact that UTF8
> is a Unicode character set that adheres to an earlier Unicode standard, its
> use is thus deprecated, and it's only there for backwards compatibility.

Hmmm, not nearly as clear as you put it... Only when reading the benefits of AL16UTF16, it becomes clear there are several definitions, and AL16UTF16 supports Unicode 3.1. Backtracking, you *could* find, that UTF8 is Unicode 3.0...

The 10G manual has been improved, and states reasonably at the beginning of the same chapter (now 6, in the 9i manual, it was 5), that 10g supports V3.2 of Unicode.

Both manuals state that AL16UTF16 is the Unicode, used by MS Win2K. A pity, no comparison is made between UCS-2 and AL16UTF16... Both have better compatibility for Java and MS Windows, compared to UTF8, but UCS-2 is fixed size on top of that. I see that as an advantage. Unfortunately, based on Unicode 3.0.

> Therefore, you actually have a choice between a viable and a deprecated
> national character set... which isn't much of a choice at all, of course,
> and inevitably leads you to choosing the AL16UTF16 one for a new database.
> And that is what I was getting at, but which I didn't say.
>
> So thanks for the chance to clarify.
>
> Regards
> HJR
>

Welcome!

-- 

Regards,
Frank van Bortel

Received on Tue Nov 02 2004 - 10:12:28 CST