At 15:07 19.02.2001 , you wrote:
>Perhaps this is one of the areas under revision... My experience has been
>that non-database developers don't always get this issue right away.
>
>You could model it as a dependent object of one of the entities, but cascade
as stated in my previous posting, I don't think dependent object applies if
I take the definition in the spec (life-cycle being controlled by ONE entity).
>delete behavior could be very expensive. Making it an entity in it's own
>right probably isn't a big win either.
well, that's what we do now when using EJB1.1 CMP. we model a relationship
with an entity but only use it for editing relationship attributes.
resolution of many-many relationships is mapped to a relational query using
the CMP engine (orion btw.) so there is no performance penalty. I was just
disappointed that it looks as if I had to circumvent the idea of
ejb-relationships in ejb2.0.
regards,
robert
>Cheers
>Jay Walters
>
>-----Original Message-----
>From: Robert Krueger [mailto:[EMAIL PROTECTED]]
>Sent: Monday, February 19, 2001 2:57 PM
>To: [EMAIL PROTECTED]
>Subject: Re: Modeling relationships with attributes with EJB2.0 CMP
>
>
>At 13:20 19.02.2001 , you wrote:
> >On Mon, 19 Feb 2001, Robert Krueger wrote:
> >
> > > Say I have an entity Person and an Entity Door and a relationship
> > > PersonDoorAccess with multiplicity many-many. Then things like
>lastAccessed
> > > and PIN would be modelled as an attribute of that relationship. I don't
>see
> > > any way of doing that using EJB2.0 CMP except for modeling the
>relationship
> > > PersonDoorAccess as an entity itself.
> >It sounds like a dependent of either Person or Door to me.
>
>its life cycle is not necessarily controlled by just one entity, so I think
>dependent doesn't apply here.
>
> > >
> > > Is there an obvious reason why relationship attributes don't seem to be
> > > supported? Is my modeling approach wrong?
> > >
> > > It would be great if anyone who's closer to the specifcation process
>could
> > > enlighten me where my mistake is.
> >I'm no closer to the spec. process than you, but I have heard rumors that
> >there will be substantial changes before the spec. is final.
>
>yes, don't know which source you're referring to but rickard oberg
>mentioned that on the jBoss list I think. he didn't mention what areas,
>though (probably can't without getting in trouble).
>
>regards,
>
>robert
>
>(-) Robert Krüger
>(-) SIGNAL 7 Gesellschaft für Informationstechnologie mbH
>(-) Brüder-Knauß-Str. 79 - 64285 Darmstadt,
>(-) Tel: 06151 665401, Fax: 06151 665373
>(-) [EMAIL PROTECTED], www.signal7.de
>
>===========================================================================
>To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
>of the message "signoff EJB-INTEREST". For general help, send email to
>[EMAIL PROTECTED] and include in the body of the message "help".
>
>===========================================================================
>To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
>of the message "signoff EJB-INTEREST". For general help, send email to
>[EMAIL PROTECTED] and include in the body of the message "help".
(-) Robert Krüger
(-) SIGNAL 7 Gesellschaft für Informationstechnologie mbH
(-) Brüder-Knauß-Str. 79 - 64285 Darmstadt,
(-) Tel: 06151 665401, Fax: 06151 665373
(-) [EMAIL PROTECTED], www.signal7.de
===========================================================================
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
of the message "signoff EJB-INTEREST". For general help, send email to
[EMAIL PROTECTED] and include in the body of the message "help".
