Hello Martin, Tuesday, October 26, 2004, 10:09:58 PM, you wrote:
>> I do not have any problem on this, but maybe a tagged value will users >> let decide if they want to map it in one table or not. MW> In the IBM Sanfrancisco framework in addition to Entity there was a MW> class of objects ( I forget the actual name) which could only existed as MW> part of an Entity. You could follow this strategy. If the associated MW> component was an Entity it would be in its own table, if "non-entity" MW> stereotype it would be a composition. This also relates to the Jira MW> issues http://team.andromda.org:8080/jira/browse/HIB-10 and MW> http://team.andromda.org:8080/jira/browse/HIB-16 Yeah, these things were named "Dependent"s. I once worked in a project where they were used - the only disadvantage was that the modelers frequently asked themselves: "Should I model this as an entity or as a dependent?". On the other hand, you may call this an advantage because the modeler makes a concious decision, then. What do the others think about dependents? I find them nice because using a separate stereotype, we could also use a different metafacade and could easily attach OCL constraints to dependents, for example: "a dependent must not own a collection". Chad, how would we write this constraint in OCL? What do the others think: should we "depend" or not? :-) Cheers... Matthias --- Matthias Bohlen Internet: http://www.mbohlen.de/ [EMAIL PROTECTED] ------------------------------------------------------- This SF.Net email is sponsored by: Sybase ASE Linux Express Edition - download now for FREE LinuxWorld Reader's Choice Award Winner for best database on Linux. http://ads.osdn.com/?ad_id=5588&alloc_id=12065&op=click _______________________________________________ Andromda-user mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/andromda-user
