Re: Data modeling for a multi-company product

From: dawn <dawnwolthuis_at_gmail.com>
Date: 25 Jan 2007 15:10:30 -0800
Message-ID: <1169766629.968914.200230_at_a75g2000cwd.googlegroups.com>


On Jan 25, 12:54 pm, "Neo" <neo55..._at_hotmail.com> wrote:
> > While I see no need for exceptions in an initial design, there
> > could be exceptions when refactoring the database schema as changes are
> > needed. I have not been in such a position, but understand from others
> > that it is difficult to manage a SQL-DBMS schema over time without
> > introducing nullable attributes.

> If ones hasn't lived thru refactoring of db schemas and related
> applications, then one is missing some critical experiences related to
> dbs.

Absolutely! I have "lived through" such, and have been hands-on with several different DBMS's and file systems, but my hands-on, heads-down-coding, production-development years were in the late-70's and 80's with IMS and VSAM files, for example, and although I have designed SQL-DBMS schema, I have not been a hands-on developer for any subsequent (production) refactoring. I've been "management" of one form or another since the late 80's, which might account for some brain damage^h^h^h^h^h^h^h^h^h^h^h^h^h differences in perspective. Cheers! --dawn

> This is where systematic approaches begin to express their
> advantages. Much experience in this area is what drove me to a method
> of representing things that is NULL-less and least impacted by meeting
> new data requirements. It turns out that methodology is based, not on
> RM, but directly on sets or what I call thingyz :)
Received on Fri Jan 26 2007 - 00:10:30 CET

Original text of this message