Internal VS External binding in stories.

[feret]

As noticed by hmedina, causal flow compression is not satisfying for the rules of the form: A(x~u),B() -> A(x~p),B() @ 0.0(1.)

Stories do not explain how A() and B() are connected, but they should.

Currently, KaSim makes no difference during the compression of causal flows between unary and binary application of rules. Thus it allows for applying the rule in a binary context even if the rate for the application of the rule in a binary context is 0.0.

There should be an option to select one of the following behaviors: mode 0) no difference between unary and binary applications; mode 1) we make the distinction only if one rate is permanently equal to 0 mode 2) we always make the difference.

Anyway, it requires major modification of the precomputation phase: -> how agents are connected has to be recorderd; -> causality definition has to be updated; -> when applied a rule is applied in an ambiguous context, the solver has to ensure than one potential path (in the history of the agents) is still valid.