On 2025-03-15  20:13, Stefan Monnier wrote:

> That would be the poor man's choice, but given you already wrote the
> code to do it better, I wouldn't make that choice.

Ok, kept the other choice, implemented your and Eli's feedback,
added tests, both for the new function and for unloading a feature
that defines a variable alias.

Please see the attached patch.

Thanks