Re: Base Normal Form

From: Kenneth Downs <knode.wants.this_at_see.sigblock>
Date: Fri, 08 Jul 2005 12:29:39 -0400
Message-Id: <3em2q2-t5f.ln1_at_pluto.downsfam.net>


David Cressey wrote:

>
> "Jan Hidders" <jan.hidders_at_REMOVETHIS.pandora.be> wrote in message
> news:Lfeze.139531$Jn.7313539_at_phobos.telenet-ops.be...
>

>> I think that would be very confusing. First, his table actually has
>> ordering, functions don't.

>
> It's not clear to me that a table has ordering, if I understand you
> correctly.
>
> The columns of a table can be referred to by name. While the names can be
> sorted alphabetically, it's not clear that columns have an inherent order.
>

I'm probably going off on a tangent here, but this is a great example of why a theory has to put into a context.

To wit, relational theory gave us a significant advance over prior art by requiring that access to columns be based on name only, with the programmer not knowing or caring about column positions on disk relative to each other. So the programmer no longer had to parse a record to get the last name at position 35, he just refers to "Last_name".

But while it is an advance for programmer, it does not eliminate the need to present columns in a certain order when dealing with human users. People generally don't like "first_name", "city", "zip", "last_name", "state".

-- 
Kenneth Downs
Secure Data Software, Inc.
(Ken)nneth_at_(Sec)ure(Dat)a(.com)
Received on Fri Jul 08 2005 - 18:29:39 CEST

Original text of this message