At 12:21 PM 2/02/2004 +0000, you wrote:

>Hi all,
> Just to clarify, all the HTML and XML data will be encoded in UTF-8.
>
>Now I understand that FireBird supports UTF-8, but do I have to
>setup/change anything for this to be the native string format?

Yes, unless UNICODE_FSS is already specified as the DEFAULT CHARACTER
SET. If all your character fields and blobs are going to be UNICODE_FSS
then this is the only sensible thing to do.

The client has to connect with lc_ctype UNICODE_FSS in the database parameters.

>Are there any problems associated with just filling up my DB with
>UTF-8 data, performance wise or any other wise?

Lots, if you don't have the client and the database set up as described above.

Indexes are limited to 84 characters (or less, if you make compound indexes).

String limit of 32767 bytes (char) or 32765 (varchar) mean (in theory)
UNICODE_FSS limits of 10922 and 10921 characters respectively.

Functions involving uppercasing and lowercasing don't work - there are no
language-specific collation sequences.


Trim your replies, please.

^hb