On 4-9-2010 21:47 Paddy wrote: > Some interesting suggestions, As far as I'm aware the Neo4j reference node > is needed to display graphs in neoclipse, That is the only reason i would > use a reference node.
Say you have some application with customers. And you want to do something for all customers. How would you acquire/iterate all customers? The approach outlined in the wiki would be to have some "customer subreference node" to which all customer-nodes are connected. So you can easily traverse each "is customer"-relation to that subreference node. http://wiki.neo4j.org/content/Design_Guide#Organizing_your_graph And the easiest access to this subreference node, is through some relationship to some starting point (i.e. the reference node). If you won't use the reference node, how would you reach that subreference node? By indexing it in some "subreferences access"-index? An alternative approach would be to index some "type"-property, where you have a "node types"-index and you'd than search for all "customer"-nodes. But, to me, that approach doesn't really seems to be the way to go in a graph database. Best regards, Arjen _______________________________________________ Neo4j mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
