Daan Huybrechs

I see, I did not realize the 0.5 restriction of StaticArrays.

Overall, I'd also favour having more and simpler concrete types over having a more complicated interval type that supports everything (like a swiss army knife of intervals). Code will be...

Also, I'm beginning to wonder why being open and closed are type parameters, since it seems to lead to quite a few type-instabilities of set operations. I guess it is...

Might it be possible to "plan" the application of a composite map? Similarly to `plan_fft` and `plan_fft!`, the plan could take an example vector, from which it could deduce the...

I have yet to add tests. First I'm wondering whether this package is the right place to add this functionality and whether it's deemed useful. It is at least to...

Looking at this code, at other parts of GenericLinearAlgebra and at LinearAlgebra, I see many opportunities for better integration (and less code in total). For this PR: - reuse LinearAlgebra.floatmin2...

Being a bit more familiar with the code now, I wonder why GenericLinearAlgebra has its own Hessenberg and Schur types, even though LinearAlgebra has those too. Is there a reason...

Just pointing out [DomainIntegrals.jl](https://github.com/JuliaApproximation/DomainIntegrals.jl) here, a package for (numerically) evaluating integrals over domains, possibly with respect to a measure. The included measures are heavily biased towards approximation theory though, e.g....

That sounds good. Small comment: measures are only defined in DomainIntegrals.jl, not in DomainSets. At some point they'll probably move to a package that is yet-to-be-named, or even better, be...

Just to confirm, we like to avoid requiring domains be a subtype of `Domain`. This is not always convenient or possible, but it does lead to flexibility where it works....