The mismatch is very old... I think that the API is the most right source of info, I assume that all of us are happy with many-to-many associations for communities (also with themselves), collections & items. I'm not sure if we want keep a many-to-many for item-bundle-bitstream. In my opinion these are only one-to-many and we should change the API (and db) in this way. I know that the Jim work about versioning uses the many-to-many association for item & bundle to share some of them between different versions, I hope that we can change this to share only the physical location of the bitstreams. In this way should be more simple to keep different versions that change only metadata of a bitstream.
Community * - * Community Community * - * Collection Collection * - * Item Item 1 - * Bundle Bundle 1 - * Bitstrea Best, Andrea daniele.ninfo ha scritto: > Hi all, > > I'm writing to underline some differences between the data model shown in the > dspace site > ( > http://www.dspace.org/index.php?option=com_content&task=view&id=149#data_model > ) and the actual code present in the /trunk ( > http://dspace-sandbox.googlecode.com/svn/mirror/dspace/trunk/ ). > > There are differences in the associations between objects of the model: some > one-to-many associations in the model are replaced in the code by > many-to-many ones. I found 3 examples: > > - In the model, a Collection can be owned by only one Community, whereas in > the code a Collection can be owned by many Community. > - In the model, a Bundle can be owned by only one Item, and in the code it > can be owned by more then one Item > - In the model, a Bitstream can be owned by only one Bundle, and in the code > by more then one. > > In the database schema, all the associations are many-to-many, there are no > restrictions. > I also looked at Jim's DAO-prototype ( > http://dspace-sandbox.googlecode.com/svn/branches/dao-prototype/ ), and he > reflects the database schema, using only many-to-many associations. > > I'm working on a prototype to introduce Hibernate, and i'd base my work on > Jim's choice, but i'd like to have some opinions about it, what should we do? > update the data model associations? follow the model and update the code in > /trunk? what kind of associations should be kept and what dropped? > > My mentor Andrea will write his opinions replying to this email, to let > everyone know them. We both are interested in having other opinions to choose > the best way to go ahead. > > Cheers, > Daniele > > > ------------------------------------------------------------------------- > This SF.net email is sponsored by: Microsoft > Defy all challenges. Microsoft(R) Visual Studio 2005. > http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ > _______________________________________________ > DSpace-tech mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/dspace-tech > > > -- Dott. Andrea Bollini Responsabile tecnico sviluppo e formazione applicativi JAVA Sezione Servizi per le Biblioteche e l'Editoria Elettronica CILEA, http://www.cilea.it tel. +39 06-59292831 cel. +39 348-8277525 ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2005. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/ _______________________________________________ DSpace-tech mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dspace-tech
