Re: problem with m:n relationship/inheritance

From: Andrus Adamchik (andru..bjectstyle.org)
Date: Tue Feb 10 2009 - 02:51:21 EST

  • Next message: Andrey Razumovsky: "Re: Cloning a CayenneDataObject"

    Hi Steve,

    This information was useful (among other things in motivating me to
    look at it rather sooner than later).

    Andrus

    On Feb 8, 2009, at 4:34 PM, Stephen Winnall wrote:

    > I was getting verification exceptions (note: "was"; I don't any
    > more) referring to "runtimeRelationship0", which I think is what you
    > meant by auto-reverse relationships. I had these in two different
    > cases.
    >
    > The initial case was the m:n mapping. In the meantime I don't think
    > it has specifically to do with m:n mappings either. The two object
    > entities involved in said mapping are subclasses, so I think it had
    > to do with inheritance, which is what CAY-1009 seems to be about. I
    > fixed this one by removing the "mandatory" option on the appropriate
    > attribute, as suggested in CAY-1009.
    >
    > The second case was another use of inheritance. It went away after I
    > went with the modeller through the entire inheritance tree and
    > removed relationships which weren't relevant from the root class. I
    > think that was my ignorance. The difficulty that I had was in
    > working out what "runtimeRelationship0" referred to. It would be
    > more helpful for the naive and the ignorant like me if the exception
    > text referred to the "auto-reverse relationship of X".
    >
    > I'm conscious that this won't help you much and that I am something
    > of a bull in a china-shop at the moment. But the more I use Cayenne,
    > the more I like it. And I appreciate your support, which I find
    > exemplary.
    >
    > Steve
    >
    >
    > On 8 Feb 2009, at 14:41, Andrus Adamchik wrote:
    >
    >> I am not 100% sure your problem is related to CAY-1009. IIRC
    >> CAY-1009 describes a very special case that did not work. In most
    >> cases n:m relationships work just fine. So can you describe your
    >> problems in more detail? Anything special in your mapping that
    >> makes you think it is the same problem as CAY-1009?
    >>
    >> Andrus
    >>
    >>
    >> On Feb 8, 2009, at 3:03 AM, Stephen Winnall wrote:
    >>
    >>> I've got this error again in another part of my schema. Since I
    >>> don't really understand what the cause is, I go through the
    >>> attributes of the DB entity switching off the "mandatory" option
    >>> and seeing if the problem goes away. Is there an easier way to
    >>> recognise which attribute is causing the problem? Is it a problem
    >>> with the DB entity or with the object entity?
    >>>
    >>> Steve
    >>>
    >>> On 4 Feb 2009, at 09:42, Andrus Adamchik wrote:
    >>>
    >>>> Yes, the issues with auto-reverse relationships need to be fixed
    >>>> rather sooner than later, and to the best of my knowledge nobody
    >>>> is working on it right now. But this is certainly high on the list.
    >>>>
    >>>> Andrus
    >>>>
    >>>> On Feb 3, 2009, at 11:39 PM, Stephen Winnall wrote:
    >>>>
    >>>>> Is CAY-1009 (Bogus runtime relationships can mess up commit) due
    >>>>> for resolution any time soon? I think I've run into it whilst
    >>>>> modelling an m:n relationship. The work-around (making the
    >>>>> attributes in the intersect entity non-mandatory) works, but it
    >>>>> would be nice to do it properly.
    >>>>>
    >>>>> Steve
    >>>>>
    >>>>>
    >>>>
    >>>
    >>>
    >>
    >
    >



    This archive was generated by hypermail 2.0.0 : Tue Feb 10 2009 - 02:51:55 EST