Re: Exceeds maximum index number

Tech Tip: Click here to run a free scan for Windows Errors and optimize PC performance

From: Tim Ferguson (FergusonTG_at_softhome.net)
Date: 03/04/04 

Date: Thu, 04 Mar 2004 08:33:04 -0800

rsginca@aol.com.NO.Spam (RSGinCA) wrote in
news:20040304004208.09510.00000908@mb-m12.aol.com:

> Hmmmm.... That's what I would describe as taking a quote out of
> context.

Unfortunately, it's quite a common sentiment around Access circles and
arises from the fact that this one program allows you to draw lines in the
"relationships" window and pretend they are relationships, with some kind
of optional "enforce relational integrity" property. It's a wrong-headed
attempt to make Access an end-user tool rather than a developer platform
and is just not a feature of real dbms systems.

> DBMSes have been around for quite a while, and for a long time none of
> them provided any automated capability of enforcing referential
> integrity, and none provided any automated capability for cascading
> updates or deletes.

If we are talking about relational databases, they _all_ provided the first
because that is inherent in the notion of R. The second is not part of the
R model at all, but is provided by many systems as a (dubious) convenience
to the developer.

What you are saying was true of previous systems like network and
hierarchical dbs, and that's why they died so fast when R came out. I am
old enough to remember Fabian Pascal's outbursts on Compuserve as the old
guard went through the pain of moving.
 
> If you actually read the full message, it would have been clear that
> what I meant by "working" had absolutely nothing
>
> to do with things like "having tests without patients".

Apparently so. But for me, that kind of error means it's not working.
 
>
> I wasn't the author of that quote,

If I thought you were, I would not have been nearly so rude :-)

> however it appears to me that what
> he was saying was that the relationships exist whether or not you use
> that "Ralationships" window to draw the little lines between the
> tables.

No: it's saying that relationships exist whether you switch on RI or not.
 
> I don't think that I would consider a relationship to be a
> 'constraint'.

So what else use is it? If a relationship does not protect the data it
doesn't actually do anything at all. In Access, a line drawn in the
relationships window without RI enforced, does nothing except provide a
default join in the query design grid. Big Deal.

> It's a
> statement of fact about the data.

No: it's an instruction to the db engine about what operations are legal
and what are not. "Statements of fact" are important in the documentation
and the data dictionary etc. but the db engine won't care about them at
all.

> The fact is that the original poster of this thread has designed a
> database and, apparently because of the limitation of MS Access, he is
> being prevented from implementing the definition of that database.

True: see my other response, having done some testing.

> I
> have no reason to believe that there is anything wrong with that
> design. (Perhaps you can provide some input.)

As has been suggested, it's a bit of an odd design and I guess there
probably is a way of getting the links down to fewer than 31; but of course
I don't know his data and there is no theoretical reason why he should have
to.

Bit long: sorry!
All the best

Tim F

Quantcast