Benoît Legat

We have such concept of domain for `x` in https://github.com/JuliaAlgebra/SemialgebraicSets.jl but it wouldn't work like that. Note that this can be achieved as follows: ```julia julia> @polyvar x (x,) julia>...

We try to avoid defining methods just to make a workflow work as it might break another workflow. Here, defining `isless` between a variable and a polynomial variable would be...

Thanks for letting me know, this was fixed in https://github.com/JuliaAlgebra/DynamicPolynomials.jl/pull/146 but I forgot to make a new release

Thanks for reporting this bug. I'm afk right now so a stacktrace could help

Indeed, as a workaround, you can try `u(u => ones(3,3))` which assumes that all variables are substituted which is the case here

We could make it work now with the variable order field

Yes, we don't automatically promote to `BigInt`, we might want to use checked operations

No, the error is expected, things will get worse after the assertion so it's best the program stops there. See the detailed explanation in https://discourse.julialang.org/t/error-in-dynamicpolynomials/89196/2

There is no feature like that that I know of. This is the kind of features that could be built on top of the MultivariatePolynomials API

Good catch ! Looks intriguing indeed, I am curious to see the result of the discussion on the julia repo.