Mark H. Wood wrote:
For example, in an RDB with "people" data, you might fully normalize the schema so that there is one and only one record for a household's address, and then associate each person that lives at that address as separate records with a foreign key to the address record. Or, depending on needs, you might denormalize that design, so that each person's record includes its own copy of the address data.
Totally irrelevant; LDAP is not a relational database. If you need a relational model with constraints and foriegn keys, etc... use a relational database.
I think the unspoken question here is, obviously, "what are the important and useful tradeoffs in designing X.500-like schemas? What are some good ways of thinking about directory object classes?"
Yes, thank you, that's specifically what I meant.
and other references that cover issues like that for RDBs. Probably none to few cover similar issues for ldap schemas.
I wouldn't expect them too.
Books on RDBs, no. But has no one ever written a book on directories which discusses schema design issues?
-- Aaron aaron@justaaron.com