Re: Re-design of data storage table

From: <johnbhurley_at_sbcglobal.net>
Date: Wed, 25 Feb 2009 12:19:04 -0800 (PST)
Message-ID: <bb145852-cbf6-4e6c-aa64-d6296d5af58b_at_a39g2000yqc.googlegroups.com>



On Feb 25, 12:28 pm, "rogergorden_at_....gmail.com" <rogergor..._at_gmail.com> wrote:

snip

> Avoid the EAV model like the plague that it truly is.
>
> Easy for java developers because it makes easy SQL for inserts.
> Easy for java developers because there's no hassle of datatype
> checking, (everything is a varchar2(100);
>
> Impossible for reporting due to the 7-8 self-joins just to get 1
> user's name, and address with zipcode and phone.
>
> Easy for java developers to blame the oracle DBA as he/she doesn't
> know or won't tell us the magic hidden parameter that will instantly
> speed queries up.
> Easy for PMs to blame the oracle DBA as this application is fast as
> hell on Joey Java's laptop and he has lots of rows in his database,
> thousands of them even.  So why doesn't it work on millions.  Must be
> an Oracle issue.
>
> Been there, done that.
>
> "Run like hell !"
>
> Roger Gorden

No worries just remember Paul has noted that he understands the principles of relational database design. His new design actually separates things into separate types ( whatever that means ). Hey even his demo create table statements etc even use SQL Server syntax so it's all good ( well at least from the Oracle DBA side ). Received on Wed Feb 25 2009 - 14:19:04 CST

Original text of this message