Tommy Hofmann

If the random thing/sampler is iteratable, this comes for free: ``` julia> struct A end julia> Base.iterate(::A, state = 1) = rand(ZZ, 1:1000), 1 julia> for p in Iterators.take(Iterators.filter(is_prime, A()),...

Do we actually endorse `zeros` or `ones` anywhere for anything? I kind of dislike them and it feels very wrong that `ones(Rational, 3, 3)` works at all. It creates a...

No, not possible. Even the roots over `PadicFields` have some issues. I will change the error messages to "Not implemented yet" or something like that.

You can overload `evaluate(...)` for a specific type if you want a different behavior. But this helps only till someone finds the next wrong result. All the code for commutative...

@YueRen can you confirm that the polynomials we want are the same as implemented here: https://github.com/sagemath/sage/pull/38536/files#diff-fe963133a558d88dbe26710f804ba7682f40f80e7e35b3999543fe13423b8862?

My point would be that they disallow `2*x` and only allow `R(2)*x`.

> The only thing that is currently holding back the 1.3.1 release is this PR (and the associated changelog PR produced by this script). Since there are still open issues...

Some breadcrumbs: - That we have in Oscar `exponents` but not `exponent_vectors` was noticed here https://github.com/Nemocas/AbstractAlgebra.jl/issues/1229 - So yes, probably we can do this `exponents` alias. - Maybe we should...

I am not sure if the "draft" status means that more comments are welcome, but here is a minor comment. The function signatures are sometimes annotated with an explicit `convert`...

Maybe something like this: ``` julia> F = free_group(3); x, y, z = gens(F); julia> w = (x*y*z)^100; julia> R = matrix_algebra(Nemo.Native.GF(2), 200); julia> X, Y, Z = rand(R, 3);...