Re: In an RDBMS, what does "Data" mean?

• Post for FREE via your newsreader at post.usenet.com ****

"Laconic2" <laconic2_at_comcast.net> wrote in message news:brOdnUAWKbT1yFLdRVn-tA_at_comcast.com...
>
> "x" <x-false_at_yahoo.com> wrote in message news:40cf3513_at_post.usenet.com...
> > > > What is the difference between a simple, sound design of a star
schema
> > and
> > > a
> > > > fully normalized relational database ?
> > >
> > > Is this a rhetorical question or do you really want to know?
> >
> > I really want to know. (is this so strange ?)
>
> OK. Sorry to be so difficult, but I've been burned more than once in this
> forum for giving a straightforward answer to a rhetorical or sarcastic
> question.

I try to put warning signs when I make rhetorical or sarcastic questions. Some of those questions, even the ones that seems "stupid" are not what they appear to be.
But nobody is perfect and fools are so ingenious :-)

Counting propositions instead of employees is an example. Difference between Datalog/Prolog and Relational is similar.

> A star schema serves different purposes than a fully normalized design.
> Depending on what you are trying to do, it could be the right way to go,
> or it could be an unfortunate design decision.

As I understood Codd's papers, a fully normalized design was not intended for users/applications schema.

> A star schema is still in 1NF, so there's no issue there. Many of the
> dimension tables have a simple key, so they are automatically in 2NF. The
> fact tables typically aren't even in 2NF. But for 3NF and beyond, there
is
> simply no attempt to normalize at all. You pay the price in redundancy,
> and you pay the price in difficulty to load.

One could design a star schema as a user/application schema. There will be a drop in performance ?
Why ?

> But you gain enormous power and simplicity when you want to take the same
> data, and slice it and dice it every which way, or drill down for more
> detail, etc. etc. There is even a tool, from the vendor of SQR (I
forget
> who) that does all kinds of OLAP games, but has no MDDB of it's own.
> Instead, you plug it into your star schema, and you're ready to rock and
> roll!

I've heard that MD model is (mathematically) equivalent to relational model, but I cannot remember where.
I also heard that a star schema is a properly designed schema. See http://www.dbdebunk.com/page/page/622790.htm - P.S. at the bottom.

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

• Usenet.com - The #1 Usenet Newsgroup Service on The Planet! *** http://www.usenet.com Unlimited Download - 19 Seperate Servers - 90,000 groups - Uncensored -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=