Re: design question
- From: JOG <jog@xxxxxxxxxxxxx>
- Date: Thu, 6 Nov 2008 08:39:26 -0800 (PST)
On Nov 5, 1:17 pm, "Walter Mitty" <wami...@xxxxxxxxxxx> wrote:
"Ed Prochak" <edproc...@xxxxxxxxx> wrote in message
news:98c0f37a-3594-4158-9e7a-c525b9e50df7@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
On Oct 28, 3:06 am, robu...@xxxxxxxxx wrote:
Walter Mitty wrote:
<robu...@xxxxxxxxx> wrote in message
Are you sure you don't mean "primary keys made up by composing several
foreign keys"? Just a guess on my part, since I don't really know what
you
mean.
Sorry, I was in hurry so I was not very clear. I mean some people
prefer to use surrogate (primary) keys instead of natural composite
keys and then use foreign keys to surrogates just for making joins
"faster". A bad idea in my opinion...
I favor your view. There is a time and place for surrogates, but too
many jump to using ID columns as the PK right away.
I also agree. I only use ID columns for "entity tables" not "relationship
tables". And I only use them when there are no reliable natural keys
available.
So aptly and simply put. Why this isn't at the top, printed in bold,
of texts on the subject I have no idea.
.
- Follow-Ups:
- Re: design question
- From: Walter Mitty
- Re: design question
- References:
- Re: design question
- From: Ed Prochak
- Re: design question
- From: Walter Mitty
- Re: design question
- Prev by Date: Re: Object oriented database
- Next by Date: Re: We claim that delete anomality is due to table not being in 3NF, but...
- Previous by thread: Re: design question
- Next by thread: Re: design question
- Index(es):
Relevant Pages
|