Many jobs are related to each other. Some jobs might be grouped because they build related source:

• they may build different branches of the same source
• produce different artifacts
• run in different environments.

Or jobs may be related because they are components of a single application, like how hudson-war and hudson-test-harness and hudson-core are all part of hudson. Or jobs may be related because they are in the same business unit, or managed by the same suit.

Hudson should support a flexible way of grouping jobs into multi-level hierarchies – that's the data part of the problem. Hudson should also provide multiple ways of viewing the hierarchy; that's the visualization part of the hierarchy.