very very serious drawback: when we know p(p(1,2), 0) = 1, then if you ask p(1,2) = 1, kernel compares p(1,2) = p(p(1,2),0), then 1 = p(1,2) is checked again

[malloc-realloc]

TODO: There is a fundamental problem with the main loop in cmpFc and fcEqualSpec, which may lead to circular reasoning. For example, given

$$ \text{know } p(p(1,2), 0) = 1, $$

we now ask whether

$$ p(1,2) = 1. $$

The current approach compares

$$ p(1,2) = p(p(1,2), 0), $$

and at this point it fails: since we compare element by element, we end up back at

$$ 1 = p(1,2), $$

creating a circular dependency.

[malloc-realloc]

fixing this bug might lead to 1000% efficiency improvement, because I will reorganize calling dependecny and hierarchy to fix this bug, which might make the code base much much faster