Teun van den Brand

It does in the current dev version :) Not sure how recent the `ggplot2_3.5.1.9000` from your session info is, but this is a reprex with a recent-ish ggplot2 dev version....

It seems to work now with recent versions of ggforce and ggplot2 ``` r library(ggforce) #> Loading required package: ggplot2 ggplot(mtcars) + geom_point(aes(disp, mpg)) + facet_col(~gear) ```  Created on...

Currently interacts with {patchwork} like so: ``` r devtools::load_all("~/packages/ggplot2") #> ℹ Loading ggplot2 library(patchwork) p But can preserve space when the plots are freed. It'd be neat if we could...

The nice thing about making this a Facet method is that extensions can implement their own if their gtable structure differs

I'm unsure why the coverage CI is failing

TODO: check errors in scales EDIT: It says: "Discrete values supplied to continuous scale.", which is a little bit off but not terrible.

Also incorporated #6090 as it is thematically related in the sense of "ggplot2 should give vctrs treatment to input". ``` r devtools::load_all("~/packages/ggplot2") #> ℹ Loading ggplot2 p x .value .label.foo...

The answer to 'could it' is 'yes', but perhaps the bigger question is 'should it' (and if so, what should it be)?

So the same argument could be made for a linetype scale, and I also wouldn't know what best represents a missing linetype. And if we'd accept this for all discrete...

The problem with setting a non-logical `na.value` for any scale that can use multiple types is that it will run into issues with type casting. For example, if we use...