Jan-Erik Rediger

> Following up on this, there are a couple of hiccups with the Python 3.12 package upgrades, I will resolve those and hopefully once that is in place this will...

Yeah, I completely forgot to upload it to the channel. Next time I try to be faster on that. :) I'd say we should link to both.

Maybe with things like [`pretty-assertions`](https://crates.io/crates/pretty_assertions) or directly using a diff crate. I play around and see what works with not too much overhead.

fwiw, I really only did the fork because I wanted to get it out of the door and write my blog post. I won't maintain much more beyond that (I...

One way this is called is through this: https://github.com/mozilla/uniffi-rs/blob/cd38ccea8236df7d93aff336c325a3a8e524af5d/uniffi_core/src/ffi_converter_traits.rs#L459 which gets called from e.g. this: https://github.com/mozilla/uniffi-rs/blob/main/uniffi_core/src/ffi/rustfuture/mod.rs#L75 which ultimately ends up being a FFI function. So the `handle` should indeed only...

This should work once I have https://bugzilla.mozilla.org/show_bug.cgi?id=1971960 landed

I assume you mean 0.25, etc. (there's no `v2.x` yet)

I cannot reproduce this locally. Code inline below. Tested with 0.25, 0.26 and 0.27. All compiles, targetting the host architecture. The proc macros correctly use `::core::*`. Given that's ... the...

> There are some places we use `core::` where using `::core::` would probably solve this - eg, [here](https://github.com/mozilla/uniffi-rs/blob/e83fc4eae267b772e69f7eab720d088a7b47fcea/uniffi_meta/src/types.rs#L149) Ah, somehow that is the one instance I overlooked. Yes, changing that...

Turns out this is biting us now. We're working around that for our own builds, but currently this makes the shipped PyPi packages fail on install, as we need a...