Re: sql tables

From: Roy Hann <specially_at_processed.almost.meat>
Date: Thu, 11 Oct 2007 21:05:49 +0100
Message-ID: <LJKdnSs3w_-6H5PanZ2dnUVZ8taknZ2d_at_pipex.net>


"David Cressey" <cressey73_at_verizon.net> wrote in message news:OVuPi.10972$br2.6118_at_trndny03...
>
> "Roy Hann" <specially_at_processed.almost.meat> wrote in message
> news:ds2dnQF3Rrfm0pPaRVnyigA_at_pipex.net...
>> "David Cressey" <cressey73_at_verizon.net> wrote in message
>> news:2RrPi.10966$br2.10003_at_trndny03...
>> >
>> > All of the "major" SQL DBMS products permit storing more than one
>> > identical
>> > row in a table. However, they provide several ways the database
>> > manager
>> > can
>> > protect the database from that event. The simplest is to declare a
>> > primary
>> > key for the table.
>>
>> Sadly given that it is now considered "best practice" to blindly and
>> automatically attach an entirely spurious unique "primary key" value to
>> every row in a table, that would be entirely futile.
>
> Why is is sad? What's wrong with declaring a primary key? Or do mean
> something else by "attach"?

I mean precisely that: attach. Construct a tuple that represents your proposition, and then *attach* a synthetic value that is chosen precisely so that it is unique (usually the next sequential number in practice), and call that your primary key. And that makes me sad.

Roy Received on Thu Oct 11 2007 - 22:05:49 CEST

Original text of this message