Re: The Practical Benefits of the Relational Model

From: Leandro Guimarães Faria Corsetti Dutra <lgcdutra_at_terra.com.br>
Date: Mon, 04 Nov 2002 15:14:34 +0100
Message-ID: <aq5vc9$6trji$2@ID-148886.news.dfncis.de>

Paul Vernon wrote:

> "Leandro Guimarães Faria Corsetti Dutra" wrote in message
> news:aq5dh5$6il6e$1_at_ID-148886.news.dfncis.de...
>

>> Paul Vernon wrote:
>> 
>>> "Now suppose the row  is to be inserted into this database."
>> 
>> Now read the following phrases and the next paragraph or two.

>
> Yes, yes. Note the word 'presumably' and the words 'just as easily'

And?…

>> You can now understand that the row is indeed to be inserted not
>> simply into the database, but into a specific relation.

>
> No. The point is that Chris & David are not saying that. The user is
> simply wanting to add a new proposition to the database, but without
> 'Orthogonal Design', the DBMS cannot always know what to do with the
> new proposition, because some meaning is hidden.

Hm. That's not what I understood. Perhaps I'm reading too much, but I think not.

What I did read is that the DBMS should be able to reject specific updates based on some kind of representation of external rules, and those are constraints.

>> It is just a matter of  how to make known to both system and other 
>> users the constraints that would give meaning to the relations
>> apart from simple meaning, thus avoiding incorrect data entry.

>
> If so, how do you propose to 'make known to both system and other
> users' such constraints?

Still waiting for the second part of the article!

But seriously, the best approximation I know of are the domain, column, relation and database constraints.

-- 
  _
/ \ Leandro Guimarães Faria Corsetti Dutra        +41 (21) 216 15 93
\ / http://homepage.mac.com./leandrod/        fax +41 (21) 216 19 04
  X  http://tutoriald.sourceforge.net./      Orange Communications CH
/ \ Campanha fita ASCII, contra correio HTML      +41 (21) 644 23 01

Received on Mon Nov 04 2002 - 08:14:34 CST