David Widmann

I would suggest implementing `TensorCore.tensor` for product measures and reexport both `TensorCore.tensor` and its alias `TensorCore.:⊗`. This would allow to construct product measures both with the unicode operator `⊗` and...

I am more used to the notation ⊗ than × and agree with the explanation given by @mschauer above, therefore personally I would rather implement `TensorCore.:⊗` than `LinearAlgebra.:×`. Apart from...

The problem was fixed more than a year ago and should not occur with StatsFuns >= 0.9.9. The issue seems to be that some dependency pulls in StatsFuns 0.9.7: https://github.com/oxfordcontrol/COSMO.jl/actions/runs/2897741705/jobs/4609456933#step:4:377

Bump @tkf :slightly_smiling_face:

I don't think a separate SpecialFunctions package is the best way to go - I assume since SpecialFunctions is such a core part of the Julia ecosystem in most cases...

And yeah, it's already hard finding maintainers and reviewers for SpecialFunctions, so I assume it would become even more difficult if some focus on one, and some on the other...

To reemphasize (I think I mentioned it above), I think it would be very helpful if the specialized subpackages would preserve the git history of the files extracted from SpecialFunctions....

I wonder if possibly maintainers of Distances who didn't want to add a dependency on ChainRulesCore might be fine with adding it as a weak dependency?

To be honest, ReverseDiff has other issues and ForwardDiff is not always an option. Actually, Zygote is developed much more actively (https://github.com/FluxML/Zygote.jl/commits/master) than ReverseDiff (https://github.com/JuliaDiff/ReverseDiff.jl/commits/master) or ForwardDiff (https://github.com/JuliaDiff/ForwardDiff.jl/commits/master) (there hasn't...

I'm a bit worried that this might cause performance regressions, or at least is suboptimal, based on the fixes that were needed in Distributions: https://github.com/JuliaStats/Distributions.jl/pull/1492