Re: Data Constraints Vs Application Constraints

From: FrankHamersley <>
Date: Mon, 14 Mar 2005 12:22:24 GMT
Message-ID: <4mfZd.197556$>

-CELKO- wrote:

>>> I note that for all these "advances" there is still not an apparent
>>> change in the % of failed IT projects..<<

> Wrong. When the applications do all the constraints, there is no data
> integrity.

A very bold statement - must be the era or something? I'd accept it if you said ...

	When the applications do all the constraints,
	data integrity may become compromised by
	unsanctioned actions.

But you didn't!

In my book (unwritten (sorry (about the pun) (about the lithp*))) data has integrity if it right now is entirely consistent with what its creator sought to represent...i.e.. it is a temporal state.

How that might be achieved or protected against degradation is another thing altogether. To suggest that only embedded constraints can deliver this is a nonsense.

Sure there can be benefits in using constraints but they are not the only or even complete way to systems development salvation.

Preaching that your particular God is the only true God (and damnation to all non believers) does not attract a very discerning crowd of followers in my experience.

> Read Ralph Kimball, Ron Ross or anyone involved with DW or
> data quality --DM Review is a good place to start.

Have had a browse - too much material to find specifics in the time allowed (oops - the DW pun was inadvertent).


>>>..unless someone can refer me to a well prepared and contrary article
>>> on that subject. <<

> Look up my three-part article on this insanity in
Tried that - don't think I found it. There was one 3 part article for noob's. Can you post a link to the specific entry point?

Also (back on topic) can you point to an industry source that shows a decline in the number of stuff ups and ideally sheets that glory home to the use of ultra constrained DBMS schemas. I know I am being a bit cheeky but the original intent was to draw out a weighting for these benefits that are accepted by many as conventional wisdom on this subject.

Cheers, Frank.

  • I always loved the name of that language - lisp that is - gawd I can't help myself!
Received on Mon Mar 14 2005 - 13:22:24 CET

Original text of this message