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

"Dawn M. Wolthuis" <dwolt_at_tincat-group.com> wrote in message news:canb6q$rab$1_at_news.netins.net...

> Is the star join a relational concept? I heard someone suggest that
> fact-dimension tables with star schema is bad design, but I forget the
> rationale for that and they seem to be very effective.

Star schemas are created primarily for performance reasons, because SQL DBMSs are so bad. They're typically denormalized extracts / transformations of normalized schemas, and thus can be regarded as large views. I think they're 2NF but not 3+NF. In any event, you wouldn't want to update one of them, because not only are they denormalized enough that you'd need to update N other rows, but expressing the constraints as triggers in a SQL database, as a derivation of the real integrity rules in the source normalized schema, would be ugly (to say the least). Thus the ETL (extract-transform-load) as basically a big function over the original database.

> and that's fine behind closed doors. I'm thankful that there is less of
> that in public on this list than there was when I started (I'm not sure
now
> what to do with the balls I had to grow at that time, but pleased that I
no
> longer need them to chat here ;-)

I won't ask where they go when you're not using them, or whether you're still able to regrow them at will like the gender-changing frogs whose DNA provided the catalyst for the dino-crisis in Jurassic Park... oh wait, I guess I just begged the question I was too demure to ask directly. :-\

I do wonder whatever happened to Bob... maybe he's reading, maybe not. I can't say I miss the abuse, but do think he knew his stuff relationally.

• Eric