Re: Data Redundancy

From: Marshall Spight <marshall.spight_at_gmail.com>
Date: 17 Feb 2006 07:20:17 -0800
Message-ID: <1140189617.069848.121030_at_z14g2000cwz.googlegroups.com>


x wrote:
>
> Step 1 for "declarative":
> Try to unambiguously specify what the constraints are using a declarative
> language
>
> Step 1 for "procedural":
> Try to unambiguously specify what the constraints are using a procedural
> language
>
> Step 2 for "procedural":
> Verify if your specification is really what you wanted
>
> Step 2 for "declarative":
> Verify if your specification is really what you wanted

This description ignores a whole host of difficulties associated with procedural constraint implementation that aren't present in the declarative case. Your phrasing is valid but glosses over the differences.

With declarative constraints, you have to figure out their logical specification, and you can simply tell the system to execute the logical specification. With procedural, you still have to figure out the logical specification, even if you are less aware of doing so because this step is often mixed in with the procedural implementation. Just because you mix them doesn't mean it's not there.

Yes, in both approaches, you have to get the spec right. (As I explicitly acknowledged.) This step is necessary and fundamental.

Marshall Received on Fri Feb 17 2006 - 16:20:17 CET

Original text of this message