wapnik/tree
The application consists of trees with mixed structures separating different aspects of the code, and a runtime capable of executing simple algorithms on tree nodes, where the logic and construction elements of the application resides. The trees are a representation of an unambiguous categorization describing a link between those aspects for a particular piece of code. The algorithms are also capable of writing configuration related to the code, thus treating most of the trees like data. The major goal of this is to make it possible to modify and evolve the application in commonly described manners by the runtime itself. It's also easy to use a VCS to have the state of the evolution fully under control.