Re: Vertical inheritance support proposal

From: Mike Kienenberger (mkienen..laska.net)
Date: Wed Jun 15 2005 - 14:15:11 EDT

  • Next message: Andrus Adamchik: "Re: Vertical inheritance support proposal"

    Andrus Adamchik <andru..bjectstyle.org> wrote:
    > I am confused. I think inheritance is inheritance, so the Java class
    > hierarch will be the same no matter whether "verti" or "singti" is used to
    > map it to the DB layout. There won't be any "composition" inside
    > DataObject. It will be a regular subclass.
    >
    > I don't think I follow your example of "Organization -> SecGroup ->
    > SecRole" (is this really inheritance, "role" is not "organization"
    > subclass?).

    Yeah, it's probably not the best example, but it was what I was given to
    working with. If I'd been the one setting up and naming the tables, it'd
    be SecOrganizationRole, SecGroupRole, and SecRole. SecRole represents a
    privilege set. SecGroup adds hierarchy. Organization is just a
    specialized SecGroup. Let's stick with PERSON/CUSTOMER[/INTERNAL_CUSTOMER]
    since it's a clearer example.

    > So let's take KODO example at
    >
    http://www.solarmetric.com/Software/Documentation/3.3.3/docs/img/inheritance-
    vertical.png
    >
    > Customer is a subclass of Person, but it is still a single object.

    Yes, it's a single object, but it has to represent two java tables (more
    with deeper inheritance hierarchies). To my understanding, that has to be
    handled with two separate DataObjects, one for the PERSON table, and one for
    the CUSTOMER table.



    This archive was generated by hypermail 2.0.0 : Wed Jun 15 2005 - 14:12:32 EDT