Re: Database in depth, by C.J. Date

On Mon, 13 Jun 2005, Mladen Gogala wrote:

> in my opinion, the author did shy away too much from using mathematics.
> Cary Millsap did not make such mistake. I don't want to keep this
> esteemed audience under suspense any longer: yes, I am a mathematician,
> with college in mathematics. The book is good, but it needs more math.

No sour grapes intended, but none of that mathematical theory has ever been of any use to me. I've never worked in an environment that afforded enough time to profess a theory - I always had to base my statements on observed results.

The thing that makes this theory analysis possible is the company where an employee works - some companies bring in enough money so that they can handle the overhead associated with employees spending half their day analyzing theoretical ideas, but the vast majority of companies employing people (DBAs, in this example) have an overwhelming amount of work to be done and any theoretical ideas have to be done on an employee's personal time.

Honestly, I think the use of theoretical math in a world of absolutes is a dodge used by mathematicians to force their way of doing things when their assertions cannot be proven by empirical data. If you're fortunate to work in a company that allows you to do that, enjoy it... because it isn't the real world.

Cary's book (particularly Method R) was a smash hit where I work (5,000+ employees) but the section on M/M/m Queueing Model might as well not even have been printed for as little attention it has received.

I apologize in advance if I have hurt anyone's feelings - no ill will intended - but like Bruce Hornsby said, "That's just the way it is."

regards.

--
http://www.freelists.org/webpage/oracle-l