David Hewitt comments

Results 1495 comments of


                                            David Hewitt

overflow evaluating requirement for `&pyo3::Bound<'_, _>: pyo3::IntoPyObject<'_>`

Maybe we could even make the changes `ToPyObject` immediately in 0.23. I think we'd know what we'd like the final state to be and it would avoid a weird intermediate...

overflow evaluating requirement for `&pyo3::Bound<'_, _>: pyo3::IntoPyObject<'_>`

Ah, great insights! Yes, perhaps in that case we can avoid doing anything rash here and keep things as-is for 0.23 I'll try to proceed with the `pydantic-core` update today...

overflow evaluating requirement for `&pyo3::Bound<'_, _>: pyo3::IntoPyObject<'_>`

It looks like with `-Znext-solver` it is possible to at least see where the problematic uses of `impl Trait` are (maybe ~100 of them, so it's going to be a...

overflow evaluating requirement for `&pyo3::Bound<'_, _>: pyo3::IntoPyObject<'_>`

Well, I found a different solution, which makes it more convenient to upgrade. I transformed the trait from ```rust pub trait Foo { /// Convert this `Foo` to a Python...

docs: add logo and mascot

Superseded in #4697

Update to PyO3 0.22

Sorry for the long delay here. Thanks for the PR and getting this moving! I think it's ok to remove the gil-refs methods as we'll be removing them in the...

Support complex numbers

What if we avoided using either a regex or `num_complex` entirely, and we used the Python logic to build `complex` instances? This is basically what we do already for `Decimal`,...

Support complex numbers

> Can we ignore this case for now and simply add a note in the documentation? Yes, we can xfail this case when running on PyPy. If you're willing to...

Support complex numbers

Yep, LGTM! Thanks @changhc

doesn't follow convention of serde naming

I'm not sure how I feel about this changing this one. It was originally `pythonize` and `depythonize`, and I was quite happy with that. The `_bound` suffix is intended to...