Posted by bob.chatman@gmail.com on 09/30/07 18:09

On Sep 30, 5:45 am, "Sanders Kaufman" <bu...@kaufman.net> wrote:
> "Jerry Stuckle" <jstuck...@attglobal.net> wrote in message
>
> news:WeCdneLtDcJVNmPbnZ2dnUVZ_q_inZ2d@comcast.com...
>
> > Sanders Kaufman wrote:
> >> I'm woring on a similar situation.
> >> I've got a userid, username, emailaddress and password.
>
> >> But I also want fields for login_cookie, last_login, and parent_user.
> >> I don't really WANT to create whole nother table to track that stuff, but
> >> good design dictates that I should.
>
> > And what "good design" is that? Definitely not normalization.
>
> Definitely normalization.
>
>
>
> >> The question becomes - if I break atomicity for expedience's sake, what
> >> will be the consequences?
>
> > --
> > ==================
> > Remove the "x" from my email address
> > Jerry Stuckle
> > JDS Computer Training Corp.
> > jstuck...@attglobal.net
> > ==================

Has anyone got any benchmarks or profiling data to back up their
claims? i have never noticed any performance increases based on
setting up a database to be set up in the, for lack of a better label,
"Everything belonging to the user in the user table" form versus the
"Spread it out and call it normalized" form. There is of course merit
to both options, but i think that it would be better to stop
speculating and start supporting the information.

I am working on a paper on this stuff so if i try to back something up
with "Sanders Kaufman" or "Bruno Barros" said it wont come across very
well.

[Back to original message]