Object-oriented Databases
Andreas Korth <[email protected]>
| Newsgroups | gmane.comp.web.webobjects.general |
|---|---|
| Message-ID | <[email protected]> |
Hi, I was always fascinated by the idea of using an OODBMS (db4objects, for example), instead of having to deal with the issues of object-relational mapping. Almost all product vendors claim, that using an OODBMS outperforms an ORM / RDMBS stack and I'm tempted to believe them. On the db4objects website (http://www.db4o.com) I came across the following statement: "The major reason for using relational databases today is legacy." Now that sounds perfectly reasonable to me. But since I haven't done a single project in which I had to deal with a legacy database, I think it's perfectly legitimate to question the use of ORM. The major case against OODBMS has always been the lack of performance and maturity. But since the concept is around for more than a decade now and a variety of products have evolved, I seriously doubt that this is still valid. Actually, I can well imagine that, at least for some application domains, it has turned against RDBMS. I would really like to hear if there are real-world experiences with OODBMS - no matter which particular product. I'm especially interested in the implications on EOF. Would an OODBMS render it completely obsolete? Could EOF probably serve as a bridge (transition) between the two worlds? Would it make any sense to replace the access layer with an OODBMS-enabled one? I'm eager to hear your thoughts! Cheers, Andreas