Re: The fable of DEMETRIUS, CONSTRAINTICUS, and AUTOMATICUS

From: Laconic2 <laconic2_at_comcast.net>
Date: Wed, 20 Oct 2004 16:16:03 -0400
Message-ID: <pfudnRZe4KgQWuvcRVn-2g_at_comcast.com>


"Tony Andrews" <andrewst_at_onetel.com> wrote in message news:1098302387.996382.76820_at_z14g2000cwz.googlegroups.com...

> Yes. But I think it is important to distinguish between 2 different
> kinds of "rule":
> 1) "Salary IS a numeric value between 0 and 999999"
> 2) "Salaries are restricted (by current company policy) to a maximum of
> 100000"
>
> The first is what a DOMAIN does: it is meant to be a permanent
> restriction (like a data type definition). The second is a rule that
> is very likely to change over time - perhaps annually say. I'm not
> sure a DOMAIN change is appropriate for that. That's why I actually
> agree with Kenneth that the user should be able to change it without
> performing DDL. The way I would choose to do it would be to store the
> maximum in a table, and then have a constraint to check the salary
> value against the maximum value.

OK, fair enough.

. But having lived through the transition between "A zip code IS a five digit number" to "An (extended) zip code IS a nine digit number", I'm extremely dubious about "permanent rules". If you allow enough time, everything is subject to change. Or how about "a year is a two digit number, implicitly added to 1900"? Ouch!

"'My name is Ozymandias, King of Kings. Look on my works, ye mighty, and despair!'
Nothing beside remains."

The question is, whose responsibility is the enforcement of the rule, and what power is required to change it?

There *are* cases where the DML DDL distinction is relevant here, and I'm not going to argue that this is definitely not one of those cases.

>
Received on Wed Oct 20 2004 - 22:16:03 CEST

Original text of this message