Re: A possible issue with theory

"JohnD" <john_nospam_at_tech-testing.com> wrote in message news:EBf4d.23602$WZ.6198_at_newssvr24.news.prodigy.com...
> Hi,
>
> I'm new to the list and have a question regarding an implementation of a
> database schema design. I'm looking for any information I can find that
> confirms what I think, or tells me I'm completely wrong - and hopefully
> where to look for additional answers.
>
> It's a very simple scenario:
>
> An account has a list of products which may have any number of attributes.
>
> My design would be to have a product table that contains a foreign key
> to an account table. Another table, say ProductAttributes_xref, relates
> a many-to-many relationship between a product and its attributes. The
> product's attributes would be stored a separate table - ProductAttributes.
>
> Therefore, give a product key and a desired attribute, you could find
> its description by selecting from the ProductAttribute table joining
> Product and ProductAttribute through the cross reference table.
>
> However, the current design is quite different and to retrieve the
> attribute description I have to do this:
>
> Given a product key, I select a value from a table. The value comes
> back as a negative number, which I must convert to positive (obviously
> multiplying by -1). I then convert this positive number to a string
> where I concatenate it onto the end of a literal to produce the name of
> another table that I have to query. I query this table to select a key
> that allows me to query another table to get the description.
>
> My first reaction was shock in horror at the design, however I admit
> that I don't know everything. Is it possible that the design, given the
> reality I face, is based on some advance theory, or an implementation of
> the relational model that I'm not familiar with?
>
> Thanks for any and all responses and/or suggestions.
>
> John

It's what you get when you let a programmer design a database. Received on Thu Sep 23 2004 - 17:18:37 CEST