nan checking in all potentially non-terminating loops

[Joshuaalbert]

Is your feature request related to a problem? Please describe. I believe I have caught all possible places where a NaN could arise and thus cause some non-intuitive non-terminating behaviour. However, as presented in the thesis of Joe Armstrong where Erlang was introduced "Making reliable distributed systems in the presence of software errors", we should strive for isolation of the responsibility of erroneous behaviour handling. To this end, we must expect that there are errors in the code and introduce checks to catch even the places where I think things are working correctly and nudge the program back on the correct path.

Describe the solution you'd like NaN checks for all potentially non-terminating loops and a correction of state when detected.

Additional context If done correctly this will not introduce a significant overhead.