Creating Efficient Schemas for Dependency Lookups

        I have been tasked to create a DB that tracks all
        inter and intra dependencies of software my project
        uses.  There is approx. 65 titles from about 40 vendors.  
        As all those product versions evolve,  the need to 
        maintain sanity is crucial. Thus, the database.  

        Version "X" of product "Y"  depends on or is not 
        compatible with version "A" of product "B". is the
        classic information to smoothly maintain.  It can all 
        turn into a plate of spaghetti pretty quickly.

        Does anyone have suggestions and / or examples
        of efficient schema designs and algorithms to
        efficiently create a system that tracks many
        and multi level dependencies of items ?


        Many thanks in advance.
        Terry    terry.hampton_at_convergys.com