Odd sf::st_crop() behaviour - r

Question

I want to crop one sf object by another, which I can do with sf::st_crop() or sf::st_intersection(). However, the resulting cropped object, doesn't match the object it was cropped by (see below reprex). Instead it looks like the crop happened on the top left corner and the bottom right. How do I get it to actually clip to the outline of poly1?
library(sf); library(rnaturalearth)
# polygon to clip by
poly1 <- matrix(c(-113, 23.5,
-113, -23.5,
-34, -23.5,
-34, 23.5,
-113, 23.5),
byrow=T, ncol=2) %>%
list() %>%
st_polygon() %>%
st_sfc(., crs = "epsg:4326")
countries <- ne_countries(continent = c("South America", "North America"),
returnclass = "sf") %>%
st_transform(., crs="epsg:4326")
plot(countries$geometry)
plot(poly1, add=T, border = "red")
cr1 <- st_crop(countries, poly1)
plot(cr1$geometry, add=T, col="red")
# intersection does the same
cr2 <- st_intersection(countries, poly1)
plot(cr2, add=T, col="green")
EDIT: it turns out that if I revert to a previous version of R (which also has different package versions attached), this error goes away.
Session info where issue arises:
> sessionInfo()
R version 4.1.0 (2021-05-18)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 8.1 x64 (build 9600)
Matrix products: default
locale:
[1] LC_COLLATE=English_United Kingdom.1252 LC_CTYPE=English_United Kingdom.1252 LC_MONETARY=English_United Kingdom.1252
[4] LC_NUMERIC=C LC_TIME=English_United Kingdom.1252
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] rnaturalearth_0.1.0 sf_1.0-0
loaded via a namespace (and not attached):
[1] Rcpp_1.0.7 magrittr_2.0.1 units_0.7-2 tidyselect_1.1.1 lattice_0.20-44 R6_2.5.1 rlang_0.4.11
[8] fansi_0.5.0 s2_1.0.6 dplyr_1.0.7 wk_0.4.1 tools_4.1.0 grid_4.1.0 KernSmooth_2.23-20
[15] utf8_1.2.2 e1071_1.7-7 DBI_1.1.1 rgeos_0.5-5 ellipsis_0.3.2 class_7.3-19 assertthat_0.2.1
[22] tibble_3.1.4 lifecycle_1.0.0 crayon_1.4.1 purrr_0.3.4 vctrs_0.3.8 glue_1.4.2 sp_1.4-5
[29] proxy_0.4-26 compiler_4.1.0 pillar_1.6.2 generics_0.1.0 classInt_0.4-3 pkgconfig_2.0.3
Session info where it doesn't:
> sessionInfo()
R version 3.6.3 (2020-02-29)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 8.1 x64 (build 9600)
Matrix products: default
locale:
[1] LC_COLLATE=English_United Kingdom.1252 LC_CTYPE=English_United Kingdom.1252 LC_MONETARY=English_United Kingdom.1252
[4] LC_NUMERIC=C LC_TIME=English_United Kingdom.1252
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] rnaturalearth_0.1.0 sf_0.9-7
loaded via a namespace (and not attached):
[1] Rcpp_1.0.6 pillar_1.4.6 compiler_3.6.3 plyr_1.8.6 class_7.3-15 tools_3.6.3 lifecycle_0.2.0
[8] tibble_3.0.4 gtable_0.3.0 lattice_0.20-41 pkgconfig_2.0.3 rlang_0.4.10 DBI_1.1.1 rstudioapi_0.11
[15] yaml_2.2.1 e1071_1.7-4 dplyr_1.0.0 stringr_1.4.0 rgeos_0.5-2 generics_0.1.0 vctrs_0.3.4
[22] classInt_0.4-3 grid_3.6.3 tidyselect_1.1.0 glue_1.4.2 ggnewscale_0.4.3 R6_2.4.1 sp_1.4-5
[29] purrr_0.3.4 reshape2_1.4.4 ggplot2_3.3.2.9000 magrittr_2.0.1 scales_1.1.1 ellipsis_0.3.1 units_0.6-7
[36] rsconnect_0.8.16 colorspace_1.4-1 KernSmooth_2.23-16 stringi_1.5.3 munsell_0.5.0 crayon_1.3.4

Answer 1

This turned out to be an issue with sf_1.0-0 following the move to using s2. It's fixed by running sf_use_s2(FALSE) at the start of the script.