Results 1 to 12 of 12
  1. #1
    Join Date
    Jul 2003
    Location
    Australia
    Posts
    217

    Normalization and OOP

    OOP and normalization have similarity.

    Normalization solves data maintenance problems by putting each type of data in only 1 place -- table.
    OOP solves code maintenance problems by putting each object-related procedure in only 1 place -- object. OOP is built upon the concept of modularity, but more radical.

    Am I right ?

  2. #2
    Join Date
    Apr 2004
    Location
    Toronto, Canada
    Posts
    249
    What do you mean by:
    OOP is built upon the concept of modularity, but more radical.
    How is OOP more radical than data modelling? What has modularity got do with data modelling? or OOP, for that matter? Modularity is simply a property of good software design, not of OOP or data modelling, or anything else.

    Actually, I would argue the converse, that data modelling results in more modular structures than OO design. The higher level abstractions of OOP are not necessarily modular. They are analogous to views that include several concepts.
    For example, a data model that results in a customer entity contains information only about the customer. Whereas a Customer class often includes related concepts such as a list of addresses, or a list of contact numbers, etc.

    Does OOP really solve the problem of code maintenance? If so, how come OOP projects are just as hard to maintain as projects built on other paradigms?

    Ravi

  3. #3
    Join Date
    Dec 2004
    Posts
    54

    Welll

    Hi Rajiravi,
    Kind of yes and no. There's a lot to Object Oriented Design. It supports concepts not found in relational modeling.

    Objects focus more on behavior, yes thats true, but the relationships or aggregations are very different in nature. The imply a lot more than ER relationships.

    Vmusic

  4. #4
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    710
    Perhaps clearly stating what ER modeling concepts are and what OOP concepts are would answer the question. You would be able to compare and contrast instead of wave your arms.

  5. #5
    Join Date
    Apr 2004
    Location
    Toronto, Canada
    Posts
    249
    Vmusic, the original poster (OP) stated that OOD was more modular than normalization. I pointed out, with a simple example, why I feel that is not the case.

    When the OP said that OOP was more radical, I asked what exactly he meant by that. I haven't got a response to that.

    If you want to discuss the relative merits of OOP and Normalization, that is fine. But do provide some examples of why you feel that OOP/OOD is so much different or better than normalization/relational theory. In particular, could you please provide examples for this statement:
    There's a lot to Object Oriented Design. It supports concepts not found in relational modeling.
    Ravi

  6. #6
    Join Date
    Sep 2002
    Location
    UK
    Posts
    5,171
    Quote Originally Posted by rajiravi
    Vmusic, the original poster (OP) stated that OOD was more modular than normalization. I pointed out, with a simple example, why I feel that is not the case.
    Actually what he said was: "OOP is built upon the concept of modularity, but more radical."

    I take that to mean "more radical than standard code modularity" rather than "more radical than normalization", which seems non-contentious. He then seems to be pondering whether OOP is to code as normalization is to data.

    I don't see an attack on the relational model there!

  7. #7
    Join Date
    Aug 2004
    Posts
    330
    So what's more modular? Conceptual radicality or object-oriented normalization?

  8. #8
    Join Date
    Mar 2001
    Location
    Lexington, KY
    Posts
    606
    urquel poses a very good question.

    Personally I think conceptual-oriented modularity provides radically normalized objects.
    Thanks,

    Matt

  9. #9
    Join Date
    Apr 2004
    Location
    Toronto, Canada
    Posts
    249
    MattR's and urquel's comments are really funny.

    I agree that
    conceptual-oriented modularity provides radically normalized objects

  10. #10
    Join Date
    Jul 2003
    Location
    Australia
    Posts
    217
    Quote Originally Posted by andrewst
    Actually what he said was: "OOP is built upon the concept of modularity, but more radical."
    I take that to mean "more radical than standard code modularity" rather than "more radical than normalization"...
    Andrewst, yes, that's what I meant.
    OOP is going one step further by grouping related Procedures and Variables within an object. Do you agree ?

    Do you agree with my point that OOP and normalization have similar underlying principle , as I have stated in my initial post ?

  11. #11
    Join Date
    Sep 2002
    Location
    UK
    Posts
    5,171
    I'm not sure that they have the same underlying principle - other than that they are both intended to improve the quality of the stuff they manage, i.e. data (normalisation) and code (OOP).

    Where problems arise is when OOP enthusiasts try to extend the remit of OOP into database design as a replacement for normalisation. AFAIK no one ever advocates the opposite: applying normalisation principles to code! Whatever that might mean...

  12. #12
    Join Date
    Dec 2003
    Location
    Canada
    Posts
    710
    http://www.geocities.com/tablizer/core1.htm

    Core Differences between OOP and Procedural Relational

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •