----- Original Message -----
From: David Schnepper <dave@...>
To: <IB-Architect@egroups.com>
Sent: Wednesday, August 02, 2000 12:19 PM
Subject: [IB-Architect] Re: Creating a case-insensitive collation set


> How would
>
> SELECT NOCASE(x) ...
>
> be different from
>
> SELECT UPPER(x) or LOWER(x) ??
>
> Dave
In that example it is probably the same. But an ORDER BY is different.

SELECT ITEM FROM TABLE1 ORDER BY ITEM NOCASE
is not the same as
SELECT UPPER(ITEM) AS FIELD1 FROM TABLE1 ORDER BY FIELD1

the former returns

ABC
abd
BCB
CBB
the latter would return
ABC
BCB
CBB
abd

There are workarounds to this issue, but "true native"case insensitive
sorting would be more efficient than placing an UPPER(FIELD) in the query.

BTW, does anyone know if other DB's such as Oracle have this feature?

Paul