Re: Yet another OO question...

On Mar 26, 10:09 pm, "Chris Uppal" <chris.up...@xxxxxxxxxxxxxxxxxxx
THIS.org> wrote:
Lew wrote:
Are you suggesting that Boyce-Codd, 4th and 5th normal forms (BCNF, 4NF
and 5NF) are at "a level of absurdity"? I assume that you agree that at
least 3rd normal form (3NF) is /de rigueur/ for relational database
design.

Maybe not (although I'll note that anything which claims to be inherently good,
all things being equal, /and/ which comes in numbered variants, obviously has
some serious problems). But the parallel is not misleading.

Cute, but you don't seem to know what you are talking about.

Talk about good OO, and about good programming in general, and you are likely
to get hit with half a hundred precepts which you "should" follow -- and the
idea that you are writing good code, and that /as a result/ the code displays
<such and such> characteristics is apt to get lost in the wash. No longer are
short methods (etc) the result of a design which has been carefully
micro-factored for readability, but they /define/ what it is to be "good".

Apples and oranges. The Relational Model has a solid mathematical
foundation, and Normalization is built on RM. Normalization has stood
the test of time. OO has no formal basis, and its design techniques
are ad-hoc (best practices.)

Something similar can happen with DB normalisation (especially if there is
someone around who can remember how the various normal forms are defined).
Instead of being /symptomatic/ of a well designed data model, they come to be
seen as definining a good data model.

Normalization is a strategy for designing databases whose goals
include: accurate representation of the entities involved, robustness,
redundancy elimination, easing future changes, ... Do you see a
problem with achieving these goals?

Don't confuse the tools with the end results, don't confuse tactics with goals.

You seem to see proper database design as a trap and prefer a seat-of-
the-pants approach. That may be true for OO designs but not
Normalization.

--
Lee Fesperman, FFE Software, Inc. (http://www.firstsql.com)
==============================================================
* The Ultimate DBMS is here!
* FirstSQL/J Object/Relational DBMS (http://www.firstsql.com)

.

Relevant Pages

  • Re: Help writing SQL statement in PHP script
    ... IMPOSSIBLE to determine if the database is normalized or not from the ... normalized - AND STILL HAVE DUPLICATES IN THESE COLUMNS. ... It would be nice if you knew something about database normalization ... There is nothing flawed about a design which has three ...
    (comp.lang.php)
  • Re: Database Design Problem
    ... Design is a hard thing to do if you want a good design that is. ... because of bad database design. ... There are lots of normalization rules ... and why I include the customer PO in the OrderDetails is because ...
    (microsoft.public.sqlserver.programming)
  • Re: Newbie question about db normalization theory: redundant keys OK?
    ... information (my understanding of what a normalized database is). ... Or, is there ever a time that you want redundant keys (that is, the ... The answer is yes, there are times when a design is a good one, ... For each normalization form, ...
    (comp.databases.theory)
  • RE: Northwind Template
    ... hand you'd learn more about database design as you go along if you start from ... I would not create a column for your structured CustomerID. ... entire database, but I will give it a try. ... example of its lack of proper normalization can be found in its Customers ...
    (microsoft.public.access.gettingstarted)
  • Re: Conceptual design advice for relational database
    ... violates the first rule of normalization. ... Get a copy of "Database Design for Mere Mortals" by Michael Hernandez. ... Access Database Samples: www.rogersaccesslibrary.com ... Want answers to your Access questions in your Email? ...
    (microsoft.public.access.tablesdbdesign)