Robin Freyler

> That means if D is Ix2, the item is (usize, usize). That's awesome! > This is maybe not so straightforward, you could leave this to me if you want....

I have created a prototype implementation of a generic, malloc avoiding `into_si_with_offset` trait implementation which can be found [here](https://gist.github.com/Robbepop/b69ccc7217094c156ffbba587948d12e). This makes it possible to map a pair of `shape: Dim`...

I haven't had quite enough time to work on this the last few days. It is nice that you also worked on another path there which looks really promising! Maybe...

Sounds very nice so far - especially the improved integration with rayon! The question that I am asking myself is: Do you still need a `Chunks` and `Windows` iterator implementation...

Thank you for your response. Glad that you have already taken it into consideration. Now I understand the big picture a bit better. > This larger road map admittedly is...

Done https://github.com/WebAssembly/meetings/issues/481 Thanks for the info!

Hi and thanks for your PR! Looks interesting! Usually for bigger changes it is nice to have an issue up before a PR emerges to allow for discussion beforehand. I...

Thanks for your elaborate explanations! > That's... weird. Maybe the generics are adding additional indirection to the trait methods? Probably inlining should fix that. May I know specifically which benchmarks...

> @Robbepop did you have the chance to check the latest changes? Sure but what does @jedel1043 think about the state of this PR? > If that's the plan, then...

That's a decent request and it may make sense to adjust the `string_interner` crate in order to support this use case. It is more a question about a proper API...