RE: For the SQL gurus: Is this practical? Is it even possible?

["Joe Pluta" <joepluta@xxxxxxxxxxxxxxxxx>]

> From: James Lampert
>
> I'm truly amazed that there even IS a solution that doesn't involve
> having multiple open cursors into QUUX/QUUUX.
>
> Again, I'm not really looking for an example; just an idea of whether
> this concept would remain practical by the time it was expanded to the
> size of the real application (with thousands of records in each file),
> or would degenerate into a total mess.

It looks to me that what you've got is a "keyless" database.  That is, the
unique key to each file is just that: a unique number.  Entity relationships
are created entirely via cross-reference file and actual field values are
never used as primary keys.

The benefit of this database architecture is that any fields in the database
(including those that what we might originally have considered to be "key"
fields) can be changed arbitrarily without having to update any associated
records.

The downfall is what you've found: the code to gather related data can get
quite convoluted.  Not to mention the fact that the actual relationships
between files are disguised; it's darned near impossible for an analysis
tool to get a good picture of your data relationships.

However, some people swear by the design, and as far as I know, as long as
you have the right indexes they run fine.

Joe