FW: views on views on views

From: Mark W. Farnham <mwf_at_rsiz.com>
Date: Thu, 26 Mar 2009 20:46:35 -0400
Message-ID: <FB7A134554604B58BC227AB98A44388A_at_rsiz.com>

So my question remains then, what do you mean by

"That's it. Any data manipulation/calculation/display should be in the application layer."  

Clearly you've decided that a summarizing and grouping calculation belongs in the database. Or did I get that wrong too? I'm really having quite a bit of difficulty understanding the meaning you intended to convey. I thought I understood you to write:  

"But I believe the database should store raw data. It should guarantee contstraint and referential integrity. That's it. Any data manipulation/calculation/display should be in the application layer."  

Any way you frame it, your response "Of Course not" contradicts that.  

I guess I'm wondering if my question caused you to change your belief or if I still just misunderstand you.  

How about joins? Do you contend that the component tables representing the relations should be loaded into the application in their entirety and then the application should perform the join? I'm really very curious about the methodology you advocate. Perhaps I am clouded by my belief that the relational model is sufficient to represent any finite set of data and that data manipulations and calculations should take place in the RDBMS engine.  

I'm afraid I don't have a complete and clean vision of where display functionality should reside, but as the display media and tools evolve I believe that is the most ephemeral function of applications anyway.  

Please understand that I *think* you have this all wrong, BUT I am genuinely curious about your reasoning.  

Regards,  

mwf      

You responded:  

On Thu, 26 Mar 2009 12:55:44 -0400 mwf_at_rsiz.com wrote:

> That is my understanding of what you wrote. Please let me know if I

> got it wrong, because I'm stumped about what else you could mean, by
"That's it.

> ..." Oh, and I definitely disagree, if that is indeed your meaning.

>
 

Of course not.  

--

Lyndon Tiu

 

 

 

I actually wrote:

 

 

-----Original Message-----
From: Mark W. Farnham [mailto:mwf_at_rsiz.com] 
Sent: Thursday, March 26, 2009 12:56 PM
To: 'ltiu_at_alumni.sfu.ca'; 'oracle-l_at_freelists.org'
Subject: RE: views on views on views

 

I want to make sure I understand your meaning.

 

Do you mean to say that if you have a few hundred million rows that you want
to total by say, oh, a hundred categories, then you would chose to return
the millions of rows to the application rather than using an aggregate
function and a group by to return the one hundred result rows?

 

That is my understanding of what you wrote. Please let me know if I got it
wrong, because I'm stumped about what else you could mean, by "That's it.
..." Oh, and I definitely disagree, if that is indeed your meaning.

 

Regards,

 

mwf

 

-----Original Message-----

From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org]
On Behalf Of Lyndon Tiu

Sent: Thursday, March 26, 2009 11:36 AM

To: oracle-l_at_freelists.org

Subject: Re: views on views on views

 

 

My blood boils too! Been there, done that.

 

One of the solutions that we implemented is to move the logic of the views
in the database layer (views, PL/SQL, etc.) to the application layer (Java,
C++, etc.).

 

It depends what your school of thought is. But I believe the database should
store raw data. It should guarantee contstraint and referential integrity.
That's it. Any data manipulation/calculation/display should be in the
application layer.

 

--

Lyndon Tiu

--

http://www.freelists.org/webpage/oracle-l

 

 




--
http://www.freelists.org/webpage/oracle-l

Received on Thu Mar 26 2009 - 19:46:35 CDT