Re: Multiplicity, Change and MV

In RM, the proper method of allowing an attribute to have multiple values is to have a separate table. If the separate table is not included in the initial schema, adding it later usually requires schema changes, data migration and script/query/code updates.

MV allows an attribute to have multiple values without schema changes. MV inherently allows each "cell" to have multiple values. However MV's approach has a fundamental flaw. That flaw is that it uses non-data independent references to link things.

The data model that I have suffers neither of the above. It allows each "attribute" to have 0 to many "values" (without ever having to design a schema) AND uses data independent references to link things.

If you would like explore the consequences of data-dependent references, I can extend the your example to illustrate the point. Received on Wed Apr 12 2006 - 00:26:58 CEST