Aurelia Molzer

If it is common, I imagine a specific implementation on `ImageBuffer` would be more optimized in any case as it could be implemented to be [cache-oblivious](https://en.wikipedia.org/wiki/Cache-oblivious_algorithm) which the pseudo-code is...

Similar to https://github.com/image-rs/image/issues/1013, I don't think this is an optimization that will work on `impl GenericImage` bounds with useful ergonomics. We can consider it for a less type-y buffer. Sidenote:...

We could host it, you only shouldn't expect that to imply a volume of outside contributions by itself (although I'd hope so). I've also been looking for a better safely...

@okaneco Regarding the design, I appreciate how complementary release of [`slice::as_chunks`](https://doc.rust-lang.org/stable/std/primitive.slice.html#method.as_chunks) is with regards to that API design. Combined we can construct most unaligned iteration with SIMD loads from any...

Thank you. On a general note, before you worry too much about it, I'm fine with the idea of bumping MSRV from the current rather ancient `1.34` to something that...

Note that more of the implementation is unsound: - tuples have padding, which violates the requirements for `as_bytes(&self)` that is accessible via the `AsContiguousBytes` provided default implementations for `[T]::as_bytes{,mut}` as...

I agree on performance, absolutely, but come to the opposite conclusion. At the moment, the `GenericImage` trait is in a double bind where it tries to _both_ serve performance and...

...

I don't think I can reproduce this on the current head commit. Here's running the output of running the explode example:  It was (by bisect) fixed in ed26c15d99eb20cfcb2b2b9ea74339e4c044af06 although...

I can do the backport over the weekend. @fintelia can you ship 0.13.2 at https://github.com/image-rs/image-gif/commit/e2990d5b49536791cfaa2bc89db98a2eb6e1a923 where it was originally intended? I'll then branch a backport from there and integrate the...