feature/simple_curl_2

This saves another 3% per iteration.

Looking at the generated assembly and comparing it with equivalent C, it doesn't look as if there's much more potential for optimisations here on our end. Our Rust code is already pretty much identical to the C one.

Using copy_from_slice and unsafe Vec::set_len instead of extend_from_slice has no impact.

Here's to hoping that Rust will improve 🎉!