Marijn Suijten

Assuming we didn't yet race on #793, is it safe to start working on factoring out the false-positives from `unsafe-default-safety.not-bounds-affecting = true` and set them to `.unsafe = false`? ---...

That happened in https://github.com/rust-mobile/ndk/pull/66, likely because the `enum` variants were listed in opposite order when compared to the NDK header files.

New Rust lints will be fixed on `master` later today after 1.91 releases, there's nothing you'll have to do for this to be merged :+1:

Thanks @tannergooding, that makes sense. However, what is the way to rename the struct - for example making it anonymous like you proposed - in `ClangSharp` without modifying the original...

@tannergooding thanks; no I haven't been able to submit a PR as I'm not up-to-speed with working on/in this project yet. Even if this is fixed, we'd be blocked one...

> Yes, it's false for all the configs, even if i request for all avaialble pixel_formats with empty attrs in ChoosePixelFormatARB. It's still false for each of them Note that...

> I don't think this is worse than using a type that has no lifetimes at all. Ideally, of course, you should have bindings that accurately annotate lifetimes, but I...

It is possible with bindgen's [`ParseCallbacks::item_name()`](https://docs.rs/bindgen/latest/bindgen/callbacks/trait.ParseCallbacks.html#method.item_name), but a `.strip_prefix("Vk")` is a way too large hammer. Replacing that with something that returns `ash::vk::PrefixStrippedType` is not allowed, because that's "not a valid...

Note that we're also addressing this as part of #80 🙂 This however by checking for the length being zero, since that's the predicate that indicates "don't read any bytes...

We have an extension wrapper and `ash-window` support in the pipeline, but the extension was only added in Vulkan `1.4.316` while our generated code is currently """stuck""" at https://github.com/ash-rs/ash/pull/951. It...