Implement One and Zero from num-traits

[ThomasdenH]

It would be useful if any Field would implement the One and Zero traits from num-traits (and perhaps even more?). One application would be the use in nalgebra, see for example the requirements to multiply two matrices there. See also this issue.

I think it would make most sense to put it behind an enabled-by-default feature. It would be possible to have a blanket implementation for Field, or alternatively an implementation could be added to the derive macro. Which would be the best way?

[tarcieri]

What exactly are you suggesting? A blanket impl won't work because One and Zero are defined in num_traits, and blanket impls can only be defined for traits defined in the same crate.

It would probably make sense for you to just write num_traits::{Zero, One} impls for your field types.

[ThomasdenH]

I meant a blanket impl for Field. That's possible since Field is defined in ff, right?

Otherwise an automatic impl included with the derive, or is that not possible?

[tarcieri]

The only thing that would be possible from the type system's perspective is a blanket impl of Field for types which impl some combination of num-traits traits, but that probably doesn't make practical sense