Morten Piibeleht

If the doc updates look good to you, it's good to go I'd say. Any thoughts on [checking for `isempty(DOCUMENTER_KEY_PATH)`](https://github.com/JuliaDocs/Documenter.jl/pull/1340#discussion_r445238652)?

I still think it would be a good idea to have this.

:+1: to all the points, thanks for writing them up! Implementing them is up for grabs though.

I haven't seen `DocTestSetup = :(using Future)` being used before. Does using a `quote` block change anything?

Ok, the exact issue is that after the `using`s, the `Future` name in `Main` refers to the module. That means that the `::Future`s in the type signatures in [at-docs block...

Yeah, re-using at-id makes sense to me. Currently, we're being extremely conservative about where we try to find at-ids (namely, only only within top-level headings), so generalizing it shouldn't be...

Not sure if we can do anything about this. We use the fact that source lines have a bunch of whitespace before them as a heuristic to determine if it...

Sorry for the delay. This seems fine to me to be honest. I also don't expect this to be breaking in practice -- since they don't work, I wouldn't expect...

Shouldn't the `.macros` section also go into the `.tex` section for MathJax v3? X-ref: https://docs.mathjax.org/en/v3.0/options/index.html#configuring-mathjax So maybe something like this? ```julia mathengine = MathJax3(Dict( :loader => Dict("load" => ["[tex]/physics"]), :tex...

My suggestion here would be to try to work backwards then. Take the generated HTML/JS site, look at the generated JS for this in `assets/documenter.js`, and try to figure out...