Re: Table design question

Bob,

I'm sorry if I appear to be talking past people. That is not my intent. In fact, I thought I did address the specific statement you reference when I wrote:

"I think everyone agrees that there is a penalty involved with enforcing certain constraints to ensure quality of the system, but that is not what is at issue. People argued, and still argue, that association by value as a performance liability for the same reasons - to justify their rational for not accepting something."

The remainder of my end of the discussion was primarily in response to the following major elements of the thread.

Mike Sherrill wrote the following:
> > (where the enumeration of values is unavoidable) *might* map to
> > something like this for a primitive data type
> > CREATE DOMAIN User_ID AS INTEGER;
> > CONSTRAINT user_id_values
> > CHECK (VALUE IN (1, 112, 314, ... 9997));
Tony/Andrewst, for whom I have the greatest respect for, responded with the following:
> The only reason you are trying to enumerate "six thousand
> non-consecutive integers" is because you are under the mistaken
> impression that the set of users who actually use a system is a
> domain, when it isn't.

• and -

> I agree that SQL database don't support domains very well, but the
> example being discussed here is NOT a domain.

Nonetheless, I probably was stating a point or points that didn't have to be made. Moreover, my math was atrocious! No one called me on it though.

Regards,

Dan

"Bob Badour" <bbadour_at_golden.net> wrote in message news:bICdnfytNMUOnLXdRVn-tw_at_golden.net...
> "D Guntermann" <guntermann_at_hotmail.com> wrote in message
> news:Hstwz4.59I_at_news.boeing.com...
> > "Tony" <andrewst_at_onetel.net.uk> wrote in message
> > news:c0e3f26e.0402090330.1ba1c048_at_posting.google.com...
> > >
> > > Of course, it is conceivable that an organisation might
> > > decide in advance that all user_id values must be taken from the
> > > enumerated set {1, 112, 314, ... 9997}, but if so then of course they
> > > pay the penalty in the complexity of enforcing the constraint.

>

>