Daan Huybrechs

> I agree with having DomainSetsCore.jl. I thought we may have `AbstractInterval{S}

I took a [stab](https://github.com/JuliaApproximation/DomainSetsCore.jl/blob/main/src/DomainSetsCore.jl) at `DomainSetsCore.jl`. The first iteration has `AbstractDomain` and `Domain{T}

There was already an issue for DomainSetsCore.jl with some discussion: #136. I'll move there.

Coming back to the issue at hand: so far I've sticked to `Domain{T}` as a supertype, because I don't know how to build DomainSets otherwise. The `T` is quite pervasive....

Meanwhile, a lot of changes happened over in DomainSets (unmerged, but in [this PR](https://github.com/JuliaApproximation/DomainSets.jl/pull/141)), which are beneficial independently. Three possibilities to move forward are: - we incorporate the changes in...

Just pointing this is implemented in DomainSets using broadcast, and it is frequently useful, e.g. for affine maps: ```julia julia> using IntervalSets, DomainSets julia> 2 .* (1..2) .+ 5 7.0...

> @daanhb interesting, thanks for pointing it out! Not great that it involves heavy piracy though... Also, multiplication doesn't really work with Unitful it seems. Good catch re Unitful, it...

Meanwhile DomainSets is updated to fix the promotions going on: ```julia julia> using DomainSets, Unitful julia> (1..3) .* 3u"m" 3 m .. 9 m julia> (2u"m"..3u"m") .* 4.0 8.0 m...

> I totally agree with this approach, but mostly for enduser-facing packages or when pirating some rarely used function. I didn't know of heavy piracy of common basic function done...

> Tried `DomainSets#master`, and `(1..2)*1u"m"` works indeed. But `(1..2)u"m"` still doesn't, even though it's also just a multiplication operation. Thanks for chasing this, there is still a promotion issue. Multiplication...