<kayabanerve:matrix.org> DataHoarder[m]: It's been applied to the spend-auth transcript and generators, and I was not planning to expand that to all of helioselene.

should it use its own personalization string, to prevent possible reuse across different context (though I don't see what benefit it'd bring)

it was applied to the hash to point, which affected the generators indeed

I was going through the codebase a few days ago and saw still some raw blake2b usages

<kayabanerve:matrix.org> In general, I'd say that's unnecessary. There were specific reasons for doing so for generators and the signature.

<jpk68:matrix.org> Apologies if this is not the best channel, but why include these headers here?

<jpk68:matrix.org> github.com/j-berman/monero/blob/990…p/fcmp_pp_rust/fcmp%2B%2B.h#L39-L40

stdint.h for uint8_t, stdalign for alignof?

it doesn't include other headers from elsewhere

<jpk68:matrix.org> I'm stupid, nevermind. I thought this was a C++ header

#ifdef __cplusplus :D

it's both, C++ and C :)

<jpk68:matrix.org> The first is deprecated in C++, and the other one could be <cstdint> instead

also, it's its own compiler unit

<jpk68:matrix.org> Yeah, I didn't realize it was supposed to be compatible with C, otherwise they could be changed

<jpk68:matrix.org> I think stdalign is a C++ keyword nopw

yeah, when exporting stuff for C-header consumption (always that #ifdef __cplusplus) it needs to either be C or have the includes also gated between #ifdef __cplusplus