Cory Fields

> compiles regardless of the BUILD_BITCOIN_INTERNAL macro. Now it is gated with the DISABLE_OPTIMIZED_SHA256 one. Is it intentional? Thanks for pointing this out. Yes it was intentional as I believe...

Added a brief description of crypto_base.

Concept ACK on avoiding exposing iterators externally. Exposing references instead seems like a bit of a lateral move though, and I think it's going to take a good bit of...

Nice use of tidy! Concept ACK, assuming `::contains()` exists for common stdlibs by now. Waiting to see what c-i thinks.

@achow101/ @ajtowns Looking again, I agree the order of operations here probably looks a little strange. For context, most of this discussion came out of the discussion about porting this...

> Yeah, that's exactly what I was expecting fwiw. Done.

Sorry, yes, I meant to factor out those changes into a separate PR.

Thanks @theStack and @hebasto for the quick review. @fanquake sounds good, will push a commit to do the switch once we see some benchmarks.

I can reproduce a slight slowdown here. I added a bench that demonstrates the difference, then the original commit, then the removal. Both libc++ and libstdc++ implement this in terms...

@maflcko I'm thinking we should just take the c++20 code and assume it will be more performant going forward. That makes sense to me anyway. I'm going to go ahead...