Feedback about an experiment to embed Python interpreter with FFI API

[Mark Raynsford]

On 2022-01-04T11:09:37 +0000
Maurizio Cimadamore <maurizio.cimadamore at oracle.com> wrote:
>
> I think a better way, which we have thought about (but have not got to 
> yet) is to think about native access as a permission that can be granted 
> to a module to other modules.

I think this is how languages like Safe Haskell [0] handle this kind of
thing now. If a function wants to perform unsafe IO, or break
referential transparency, then it has to be in a module that declares
that it does those things. It's then the responsibility of the
developer aggregating all of the modules into an application to decide
if they can live with the application code having those capabilities or
not.

[0] https://wiki.haskell.org/Safe_Haskell

-- 
Mark Raynsford | https://www.io7m.com