John-Alan Simmons

I can certainly open a PR as I have a branch I'm using for this. @Stebalien at least from my POV it's not about trusting crypto.Rand, but intentionally wanting deterministic...

Just want to add that currently the libp2p crypto generation of Ed25519 keys, which uses `crypto/ed25519` under the hood. However, the [`crypto/ed25519.GenerateKey`](https://cs.opensource.google/go/go/+/refs/tags/go1.22.3:src/crypto/ed25519/ed25519.go;l=138) which also takes in a `rand io.Reader` and...

> https://gitlab.com/yawning/secp256k1-voi which doesn't even support reading from a user-specified random source In this example, the library doesn't specify it because its a lower level utility, the `dcrec/secp256k1` and the...

> It does not provide such a method because the absolutely overwhelming majority of the time, the system cryptographic entropy source is the right answer and the API is intentionally...

Only done an quick look, nothing immediate code wise. But can you create a follow-up issue to change the `null` behavior for Numeric comparison on the Go connor implementation too....

I'd rather leave my implementation of Fuckfuck as a "fuck" only type of language (If thats even a thing). I'd look into what it would be like to use variations...

New [Language-Spec](https://github.com/jsimnz/fuckfuck/blob/master/doc/LANGUAGE-SPEC.md). Will branch, change, and merge latter.