Re: Hashes from composite keys?

From: Cimode <cimode_at_hotmail.com>
Date: Tue, 27 Jul 2010 11:56:29 -0700 (PDT)
Message-ID: <f334cc9d-76c1-46f6-9b19-71b5e12c562c_at_a30g2000vba.googlegroups.com>


On 27 juil, 02:04, Daniel Pitts
<newsgroup.spamfil..._at_virtualinfinity.net> wrote:
> On 7/26/2010 4:03 PM, Bob Badour wrote:
>
>
>
> > Daniel Pitts wrote:
>
> >> On 7/23/2010 1:14 PM, Karsten Wutzke wrote:
>
> >>> Hello,
>
> >>> what are the best practices for generating hash codes from composite
> >>> keys? I need to mimic something like what a composite index does. In
> >>> fact, it's for mapping between relational keys and object IDs.
>
> >>> Can anyone point me into the right direction please?
>
> >> Yes, don't use Hashing, but use a composite key. Aggregate the keys
> >> into one value (eg, the BIT pattern of that one key value is the
> >> concatination of the bit patterns of all the other values.)
>
> > That's rather dumb advice.
>
> Really? The OP said "I need to mimic something like what a composite
> index does"  Well, I suggested a simplified (yet correct) version of
> what a composite index does.
Unless a constraining mechanism insures a 1:1 cardinality between the surrogate key physical implementation and candidate key physical implementation, creating a physical pointer reference with *any* technique is *still* a waste of time.

Additionally, it is reasonable to assume that hashing could, at best, produce an unreliable implementation of a poor concatenated key which makes totally useless.

The physical is an implementation of the logical, not the way around. Received on Tue Jul 27 2010 - 20:56:29 CEST

Original text of this message