Compilation on Windows latest + R3.6 seems to fail

[mgirlich]

See the dbplyr run: https://github.com/tidyverse/dbplyr/actions/runs/6664449009/job/18112087560?pr=1374

[krlmlr]

@jeroen: any idea?

[jeroen]

I'd recommend to remove 3.6-windows from your testing matrix. I am no longer actively testing/building for R-3.6 on Win because it used a very old toolchain, and nobody really uses that anymore. We now need to support so many different toolchains on Windows for different versions of R, that it is no longer feasibly to support the very legacy things.

[mgirlich]

Which version of R would you recommend for Windows?

[jeroen]

Which version of R would you recommend for Windows?

I'd recommend testing on Windows with at least R-4.1 and R-4.2 and R-release. That implicitly covers R-4.0 because it uses the same toolchain as R-4.1.

Soon we may need to add clang toolchains but those are not yet ready.

[krlmlr]

R 3.6 removed.

[github-actions[bot]]

This old thread has been automatically locked. If you think you have found something related to this, please open a new issue and link to this old issue if necessary.