Joshua Rudolph

Thanks for your time in reviewing; if you see the change as being more risk than it is worth to incorporate into pyo3, then I trust your judgment well above...

I found [this repo](https://github.com/noamtashma/owning-ref-unsoundness?tab=readme-ov-file) discussing some issues with `owning_ref`, and wow you really weren't kidding. I hadn't messed with the problematic `as_owner`/`with_owner` and similar methods since they aren't relevant to...

Thanks for linking that thread @davidhewitt, I read through it and just submitted a draft PR #4203 implementing a possible solution roughly based on my comment above.

@davidhewitt I had an idea related to this... What if we added a public `pyo3::ops` module containing traits analogous to those in `std::ops`, but defined something like this: ```rust pub...

> I'm reluctant to add more traits as we have a lot already IMO it shouldn't be an issue to add self-contained utility traits like this this that would give...

No problem at all David! I know I've been spamming you with PR's this week 😉 However, I'd actually like to put this PR on a brief hold until the...

Adding on to alex's point, you can instead use `PyRefMut` for both the argument and the return value to let you mutate the type (since `PyRefMut` dereferences to `&mut T`),...

@EricLBuehler all it really needs is a review and some minor updates to the docs, but I was hoping for some feedback from the devs before I bothered putting on...

Congratulations David! And no worries Adam, I haven't had any time for open-source work lately either. But I absolutely agree that @bschoenmaeckers deserves co-authorship on this :) I just fixed...

I fixed the first check, but I don't quite understand the second... does anyone have a clue why cargo would be failing to find a compatible `syn` version for pyo3-macros-backend?