I've long wanted to have proper C ABI support _and_ to free Nim of the C ABI
constraints.. ie right now it's a bit of a random patchwork of overlapping
features and flags and it's hard to reason about it, which is why it's also hard
to document - ie when writing about "some shared properties" of C and Nim
integers, what's lacking is a coherent and complete list of what those
properties are .. https://github.com/nim-lang/RFCs/issues/506 contains a subset
of those properties, but the idea here would be that there exists a section of
nim that _clearly_ follows C ABI rules and another that uses the Nim ABI _fully_
instead of the current half-empty solution where Nim is not compatible with C
_by default_ but also doesn't exploit the positives that being free of C has to
offer