Tony Arcieri

@paulmillr that’s true (for encryption, decryption in a SIV mode can still be 1-pass), but we’re talking about ChaCha20Poly1305 here...

If anyone would like to try wiring it up, `chacha20poly1305` v0.4 now has a [generic `ChaChaPoly1305` type](https://docs.rs/chacha20poly1305/0.4.1/chacha20poly1305/struct.ChaChaPoly1305.html) which should theoretically be usable with the [`ChaCha20` implementation in the `c2-chacha` crate](https://docs.rs/c2-chacha/0.2.3/c2_chacha/type.ChaCha20.html)....

Also note that the `chacha20` dependency in `chacha20poly1305` is now optional if `c2-chacha` ends up working out.

STREAM is "embarrassingly parallel" so pick any parallelization strategy you want

@str4d a few options for additional improvements: - `asm` implementations of ChaCha20 and/or Poly1305 - [Pipelining ChaCha20 and Poly1305 via XMM registers](https://github.com/RustCrypto/traits/issues/289#issuecomment-721182415)

In the past we’ve talked about phasing out `cargo-audit` in favor of `cargo-deny` (Edit: I see you mentioned that now) I’m curious why you need to run both and the...

This crate has been unmaintained for nearly 6 years. See #467

Yeah, `unsafe` is for memory safety, and changes the semantics of the language (in scary ways that would only make things worse for a crypto context). However, there is something...

@burdges I think the simpler solution is for those who are enthusiastic about "legacy" crypto to make their own crates for algorithms they are willing to support, as opposed to...

> So categorically labeling those primitives as unsafe and tossing them out/hiding them behind flags seems wrong I think if you were to take a straw poll among cryptographers as...