Viewing fact / rule chains relative to event 0

[alex-dixon]

Each event 0 is an action or insert unconditional. We should be able to trace the effects of this by building "chains" (paths of a DAG) of events. E.g.

action -> ruleA -> fact2 -> ruleB -> fact3
    \                               \
      \                               \  -> ruleC -> fact4
        \  -> ruleD -> fact5

It seems possible to start visualizing chains from any rule insertion so this doesn't have to be limited to actions.

[alex-dixon]

The general concept here might be the next most useful and interesting thing we could do.

Going off the comment above, actions need not be privileged -- we should be able to trace a causal chain for any fact backwards and forwards. Facts need not be privileged either -- we can analyze rules the same way.

I doubt I'd be happy with anything short of an interactive node-edge graph for this. In the past, confusion and challenges have arisen when considering how to implement a node-edge graph that can depict slices of time, how to visualize facts that may have existed "ages" prior to the current focused view.

My least favorite part of this challenge is the graph visualization itself. Neo4J has an open source graph visualization atop d3, cytoscape appears to be a highly configurable framework specifically for node-edge graphs.