| Re: [higgins-dev] IdAS Attribute data model |
|
Let me rephrase what I think you said with an example.
Using the original case, a DigitalSubject could be presented which has an eyeColor attribute having two values (blue and green). It could also have an eyeColorFromSource1 attribute with a value of blue, and an eyeColorFromSource2 attribute with a value of green. Interrogating applications only need to read the eyeColor attribute. Applications wanting to update, can work with the *From* attributes.
I've probably got it all messed up. Also remember that this one example (of source) is just one example of attribute (or value) metadata. However we solve this should be applicable to other kinds of metadata as well.
To answer your followup question: The IdAS consumer interacts with a Context. As far as it knows, all values come from that context. The context provider has to maintain the relationship between the IDigitalSubject and IAttribute instances presented to the IdAS consumer and their actual source(s).
Jim
>>> "Marc Boorshtein" <mboorshtein@xxxxxxxxx> 3/27/07 5:51 PM >>> This is a very similar issue that ldap virtual directories have in dealing with "joined" data. To stick with the analogy Mary has blue eyes and blond hair. Blue eyes may come from one source and brown hair may come from another. To the application Mary just has blue eyes and brown hair, it doesn't much care where the data is coming from. There are instances however where the source is an issue (primarily in updating the user's data). For this reason a few virtual directories (Oracle VirtualDirectory and MyVirtualDirectory off the top of my head) include an attribute that identifies where data came from if the application or the system needs it. Then between the knowledge of the source information and generally some type of configuration that maps attribute x to source a and attribute y to source b the virtual directory can make a decision as to where each attribute comes from when needed. There are still issues if values for the same attribute come from different sources, but these are generally manageable through mapping and are pretty rare. One thing I've yet to understand about the IdAS is how in fact does it build relationships between sources? In a virtual directory there is generally a "router" that traverses the DIT. I understand the IdAS is not an ldap directory but I don't follow how the relationships are made, or is that the responsibility of a higher up process? Marc _______________________________________________ higgins-dev mailing list higgins-dev@xxxxxxxxxxx https://dev.eclipse.org/mailman/listinfo/higgins-dev |