Re: "We don't do triggers"

"Mike Sherrill" <MSherrillnonono_at_compuserve.com> wrote in message news:mig9svkqrsqhn30gkmkua54345v9fiabov_at_4ax.com...
| On Tue, 25 Nov 2003 08:35:23 -0500, "mcstock"
| <mcstockspamplug_at_spamdamenquery.com> wrote:
|
| >if you're referring to declarative constraints, that's a given
|
| I'm pretty sure that most of the people posting to this thread don't
| regard it as a given.
|
| Looks like it's time to clarify who means what when they're talking
| about "business logic" and "integrity constraints". To me, it looks
| like some people regard them as identical, and some people regard them
| as different.
|
| --
| Mike Sherrill
| Information Management Systems

took me awhile to remember the context of my comment...

is was in response to your observation that "integrity constraints cannot be subverted (or sidestepped)" did not appear on lists comparing advantages J2EE and PL/SQL

i specified declarative constraints, referring to those you'll find in the USER_CONSTRAINTS data dictionary view -- and they can not be subverted or sidestepped no matter where the SQL comes from, whether J2EE, PL/SQL, VB, MS-Access, SQL*Plus, etc. etc. etc.. hence, my comparison to datatype checking -- if a declare a column to be DATE, it doesn't matter where the SQL comes from, Oracle will not let a non-date value in the column

however, taken out of context, you seem to read my comments differently -- using declarative constraints is definitely not a given for too many designers/developers

• mcs