* This helps with type checkers to know that things can be None or not (both to
silence previous warnings about "unnecessary" checks that are actually
necessary, and to ensure those checks aren't omitted by accident)
* For some reason python type checkers get confused at things like `from .Foo
import Foo` and start then treating `Foo` as either the module or the class.
* We enforce default and copy constructors where possible, this isn't very
pythonic but we can't use copy.deepcopy on our types.
* The structs that have specialised constructors e.g. FloatVector or Subresource
also have those documented though we have no way to enforce it.
* We create both a stubs folder for each specific version, and a 'latest' that
is updated by newer RD versions.
* For development, the stubs will be regenerated any time the git commit hash
changes for development builds.
* This is smaller and easier to maintain, and adds functionality like resolving
circular dependencies to be able to import the stubs as-is on python 3.6+
baldurk