Re: Generalised approach to storing address details

> Even though I work with what some call a hierarchical database, it is
> not considered a good practice to model strong entities in a
> hierarchical relationship with each other. At the higher level, the
> database represents a di-graph, but with trees on the nodes. Those
> trees represent a single entity, with weak entities and attributes
> subordinate to this entity. This approach to modeling data is seen
> throughout the industry, with enough success to consider it an industry
> "best practice" I suspect. It is hard to get relational database folks
> to identify the model with a name other than "hierarchical" however.

I am not familiar with di-graphs, strong/weak entities, trees on nodes. What is the advantage of modelling such things in non-RMDBs? Could you give a small example? Received on Wed Dec 13 2006 - 22:36:25 CET