Ashley Anderson

Is this possibly a good “first timer” contribution? I’d love to get involved in this project but am a relative beginner in Rust and don’t want to bite off more...

No worries! Thanks for the thoughtful response. I'm not afraid to dive in but must admit I don't have experience with proc macros (yet) so I'm probably not the most...

A number of prominent Python projects use [towncrier](https://towncrier.readthedocs.io/en/latest/). The drawback I see is that it wants to tie every change to an issue/ticket number. Maybe that could be configured to...

I found some seemingly relevant discussion on BPO https://bugs.python.org/issue5945 One suggestion there is to use `PyMapping_Check() && !PySequence_Check()` for a true mapping check. Would that work here?

It seems the upstream route is blocked, but maybe we can take the same out they did and just "fix" it in documentation. For example add a warning that downcast...

Yeah that's fair I guess I wasn't sure if that was enough of a block to miss the 0.16 release this is targeted for. Also if it's a change going...

> Considering this is a wont-fix upstream, could we just implment Davids suggestion to fix this? I'm +1 on this and would be happy to implement as well if no...

I'm running into a slight snag with this implementation for Python classes implemented in Rust. For 3.10+ I think we can add the `Py_TPFLAGS_MAPPING` if defined with `#[pyclass(mapping)]` but for...

Oh wait maybe I just need to add `__subclasscheck__` edit: nope, that would have to be implemented on the metaclass 😞

> As a sanity check, I wonder if there is any evidence we can collect (or tests we can write) that this check is functionally equivalent to the slow path?...