I haven't added NaturalInteger (#112) yet, but I noticed one more thing: I could efficiently derive a safe 2-by-1 division algorithm from an unsafe protocol requirement. It might reduce some overhead in other algorithms. I've also considered a more elaborate Guarantee-esque (#62) approach, but I don't think it would add much value in this case.
I haven't added
NaturalInteger
(#112) yet, but I noticed one more thing: I could efficiently derive a safe 2-by-1 division algorithm from an unsafe protocol requirement. It might reduce some overhead in other algorithms. I've also considered a more elaborate Guarantee-esque (#62) approach, but I don't think it would add much value in this case.Of note, I don't believe the 2-by-1 standard library algorithm is a built-in so I could perhaps roll my own at some point.