add story about nalgebra

[lcnr]

No description provided.

[oli-obk]

Would this work if the type alias were expanded manually here?

[lcnr]

no, we still only see ConstKind::Unevaluated in the self type

[oli-obk]

How is this case a problem then? For the later cases with exhaustive impls?

[lcnr]

because you can't write that impl

#![feature(adt_const_params, generic_const_exprs)]

struct Foo<const N: Option<usize>>;

impl<const N: usize> Foo<{ Some(N) }> {}

results in

error[E0207]: the const parameter `N` is not constrained by the impl trait, self type, or predicates
 --> src/lib.rs:5:12
  |
5 | impl<const N: usize> Foo<{ Some(N) }> {}
  |            ^ unconstrained const parameter
  |
  = note: expressions using a const parameter must map each value to a distinct output value
  = note: proving the result of expressions other than the parameter are unique is not supported

[oli-obk]

OK, thanks! That's what I was hoping to be the issue, but it wasn't clear to me from the text. To make it a bit simpler, can you a) expand the type alias and b) explain that a generic constant isn't what we actually want, as that is the notion I got from your comment about "constructors being injective"