Ricardo Fernández Serrata

I feel like this is an insult to Zig devs 💀, lol

Now we need Fortnite and GTA V

"Finally, no more random BSoDs!" (he doesn't know drivers are written in C 💀)

> previous bound checking It's for this (and many other reasons) why I believe this lint should be in `restriction`, not `pedantic`

I recommend reviewing any other code that uses `==` and/or `assert_eq!`, to check if it should be replaced by [`total_cmp`](https://doc.rust-lang.org/core/primitive.f64.html#method.total_cmp). `total_cmp` also distinguishes between signed `NAN`s, which may be useful...

> `.to_bits()` I agree, but that would only apply to other cases (such as performance-sensitive code), not here. The spec imposes no requirements about `fabs` return bit pattern

> There is only a single bit pattern for 0 or -0. Yes, but some platforms use "flush-to-zero" for subnormal numbers. I'm not entirely sure how that mechanism works 🤔

Not quite closed, only in the sense that nobody is allowed to copy or modify it. It's publicly read-only. Although, it's a bit more complicated: The author has implicitly granted...

@sinisimattia What do you mean?

I totally expected this to eventually happen, lmao