Structural version control for canonical S-expressions.

Saman means together in Icelandic.

Initially this the garbage collector for cxp (traces from or counts refs to names defined here in saman).
Then in order to do garbage collection of named data as well we will extend it with the logic programming.
Then in order to automate the garbage collection of named data we will extend it with probabilistic assessments.

The aspiration here is for this to be the economic engine of the peer (todo: vegur).

A name will either dereference to a context for understanding the following data or to a
namespace which indexes competing definitions for the name.

Nested S-expressions are used to index into the namespace lattice to find a particular 
definition. A path in the lattice can be found with weighted logic programming and the
weights of the properties can be kept track of with an object-property graph.

Properties are metadata. They are related using logic programming.

Each user will have their perspective calculated based on the trust in sources of metadata.

The event structure across the network can be a DAG, consensus is always probabilistic.